1
MONSOON DEPENDENT ECOSYSTEMS: IMPLICATIONS OF THE VERTICAL
DISTRIBUTION OF SOIL MOISTURE ON LAND SURFACE-ATMOSPHERE
INTERACTIONS
by
Zulia M. Sánchez-Mejía
_______________________
A Dissertation Submitted to the Faculty of the
SCHOOL OF NATURAL RESOURCES AND THE ENVIRONMENT
In Partial Fulfillment of the Requirements
For the Degree of
DOCTOR IN PHILOSOPHY
WITH A MAJOR IN WATERSHED MANAGEMENT AND ECOHYDROLOGY
In the Graduate College
THE UNIVERSITY OF ARIZONA
2013
2
THE UNIVERSITY OF ARIZONA
GRADUATE COLLEGE
As members of the Dissertation Committee, we certify that we have read the dissertation
prepared by Zulia Mayari Sánchez-Mejía, titled Monsoon Dependent Ecosystems:
Implications of the vertical distribution of soil moisture on land surface-atmosphere
interactions and recommend that it be accepted as fulfilling the dissertation requirement
for the Degree of Doctor of Philosophy.
_______________________________________________________________________
Shirley A. Papuga
Date: 5/31/2013
_______________________________________________________________________
Francina Dominguez
Date: 5/31/2013
_______________________________________________________________________
Xubin Zeng
Date: 5/31/2013
_______________________________________________________________________
Steve Archer
Date: 5/31/2013
_______________________________________________________________________
Willem van Leeuwen
Date: 5/31/2013
Final approval and acceptance of this dissertation is contingent upon the candidate’s
submission of the final copies of the dissertation to the Graduate College.
I hereby certify that I have read this dissertation prepared under my direction and
recommend that it be accepted as fulfilling the dissertation requirement.
________________________________________________ Date:___________________
Dissertation Director: Shirley A. Papuga
3
STATEMENT BY AUTHOR
This dissertation has been submitted in partial fulfillment of the requirements for an
advanced degree at the University of Arizona and is deposited in the University Library
to be made available to borrowers under rules of the Library.
Brief quotations from this dissertation are allowable without special permission, provided
that an accurate acknowledgement of the source is made. Requests for permission for
extended quotation from or reproduction of this manuscript in whole or in part may be
granted by the head of the major department or the Dean of the Graduate College when in
his or her judgment the proposed use of the material is in the interests of scholarship. In
all other instances, however, permission must be obtained from the author.
SIGNED: Zulia M. Sánchez-Mejía
4
ACKNOWLEDGEMENTS
This research was possible thanks to the doctoral studies fellowship by CONACYTMexico (Consejo Nacional de Ciencia y Tecnología, National Council for Science and
Technology). Additional support was provided by the University of Arizona College of
Agriculture and Life Sciences (CALS), The Arizona University System Technology and
Research Initiative Fund (TRIF), The University of Arizona Office of the Vice President
for Research (VPR), SAHRA (Sustainability of Semi-Arid Hydrology and Riparian area)
under the STC Program of the National Science Foundation (NSF), NSF CAREER
Award # EAR-1255013, the Springfield Scholarship for Graduate Research in
Rangelands, the University of Arizona Global Change Interdisciplinary Program
Dissertation Improvement Grant, the William A. Calder III PhD Scholarship for Mexican
graduate students focused in conservation, the Kel M. Fox Watershed Scholarship, and
the Institute of the Environment travel support grant.
This research would not have been possible without the guidance and mentoring of my
major advisor Shirley Papuga and the rest of my doctoral committee: Francina
Dominguez, Xubin Zeng, Steve Archer, Willem van Leeuwen. I am especially thankful
for their encouragement, patience, and sharing of knowledge. Other professors that had a
large impact in my doctoral studies included Peter F. Ffolliott, Valerie Trouet, Katherine
Morrissey, Paul Brooks, and Jaime Garatuza.
I am very grateful to Russell Scott, Jim Shuttleworth, and Michael Crimmins for
providing field equipment that was necessary for the completion of this research.
I also am grateful to those volunteers who shared their time and made the hot sunny days
in the field fun and pleasant, including Retta Bruegger, Carrie Presnall, Russell Lyon,
Jose Arizpe, Alex Arizpe, and Xavier Zapata.
Significant progress in my language and critical thinking skills is thanks to the Papuga
Lab: Ami Kidder, Andy Neal, Bhaskar Mitra, Daniel Bunting, Evan Kipnis, Jessica
Swetish, Joseph Miller, Krystine Nelson, Lori Lovell, Maria Pilar Cendrero Mateo, and
Zack Guido.
I owe special thanks to all the support from Katie Hirschboeck, Lesa Langan-Du Berry,
Katie Hughes, Ashley Stewart, Amber, Khuyen Huynh, Chiyo Yamashita-Gill, Andy
Honaman, and Bryce.
I will like to thank all the members of the NRGSO (Natural Resources Graduate Student
Organization) for organizing talks, seminars, HH, and for sharing the good and not so
good moments of graduate school.
5
Last but not least, I thank my sanity and emotional wellbeing to my parents Edgar and
Lulú, my siblings Ana and Camilo, my boyfriend Alexis Arizpe, Familia Arizpe, Familia
Nuñez, Familia Sanchez, Familia Mejia, and so many friends that have been there
through this process.
6
DEDICATION
To creosotebush…
7
TABLE OF CONTENTS
ABSTRACT ............................................................................................................... 11
CHAPTER 1: INTRODUCTION .............................................................................. 13
1.1 Background ..................................................................................................... 13
1.2 Research Objectives ........................................................................................ 18
1.3 Dissertation Format ......................................................................................... 21
CHAPTER 2: PRESENT STUDY............................................................................. 23
2.1 APPENDIX A: OBSERVATION OF A TWO-LAYER SOIL
MOISTURE INFLUENCE ON SURFACE ENERGY DYNAMICS AND
PLANETARY BOUNDARY LAYER CHARACTERISTICS IN A
SEMIARID SHRUBLAND .................................................................................. 23
2.2 APPENDIX B: QUANTIFYING THE INFLUENCE OF DEEP SOIL
MOISTURE ON ECOSYSTEM ALBEDO: THE ROLE OF
VEGETATION ..................................................................................................... 24
2.3 APPENDIX C: EMPIRICAL RELATIONSHIPS BETWEEN SOIL
MOISTURE, ALBEDO, AND THE PLANETARY BOUNDARY LAYER
HEIGHT: A TWO-LAYER BUCKET MODEL APPROACH............................ 25
2.4 Future Research Opportunities ........................................................................ 26
2.4.1 Land surface- atmosphere interactions: dryland ecohydrology
modeling.............................................................................................................. 26
2.4.2 The shrubby “gray” –zone of woody encroachment ................................. 28
REFERENCES ......................................................................................................... 29
APPENDIX A: OBSERVATION OF A TWO-LAYER SOIL MOISTURE
INFLUENCE ON SURFACE ENERGY DYNAMICS AND PLANETARY
BOUNDARY LAYER CHARACTERISTICS IN A SEMIARID
SHRUBLAND ........................................................................................................... 42
1. Instruction .............................................................................................................. 45
8
TABLE OF CONTENTS
2. Study Area and Methods........................................................................................ 49
2.1 Study Area ....................................................................................................... 49
2.2 Data ................................................................................................................. 51
2.2.1 Radiation and Surface Energy ................................................................... 51
2.2.2 Soil Moisture ............................................................................................. 53
2.2.3 Atmospheric Sounding .............................................................................. 54
2.3 Soil Moisture Conceptual Framework ............................................................ 56
3. Results .................................................................................................................... 58
3.1 Soil Moisture Influence on Radiation and Surface Energy Components ....... 59
3.1.1 Wet and Dry Season .................................................................................. 59
3.1.2 Two-layer Conceptual Framework ............................................................ 60
3.2 Soil Moisture Influence on the Planetary Boundary Layer ............................. 62
4. Discussion .............................................................................................................. 63
References .................................................................................................................. 68
APPENDIX B: QUANTIFYING THE INFLUENCE OF DEEP SOIL
MOISTURE ON ECOSYSTEM ALBEDO: THE ROLE OF VEGETATION ........ 89
1. Instruction .............................................................................................................. 92
2. Study Area and Method ......................................................................................... 96
2.1 Study description ............................................................................................. 96
2.2 Field Campaign ............................................................................................... 97
2.2.1 Sampling design......................................................................................... 99
9
TABLE OF CONTENTS
2.2.2 Analysis ..................................................................................................... 100
2.3 Camera-Derived Greenness ............................................................................ 101
3. Results and discussions .......................................................................................... 102
3.1 Two-Layer Soil Moisture Influence on Greenness ......................................... 103
3.2 Two-Layer Soil Moisture Influence on Albedo .............................................. 104
3.3 Influence of Vegetation on Ecosystem Albedo via Access to Two-Layer
Soil Moisture ......................................................................................................... 106
3.4 Implications of Soil Moisture Drydown Differences between Bare and
Canopy Patches ..................................................................................................... 109
4. Conclusions ............................................................................................................ 111
References .................................................................................................................. 112
APPENDIX C: EMPIRICAL RELATIONSHIPS BETWEEN SOIL
MOISTURE AND THE PLANETARY BOUNDARY LAYER HEIGHT: A
TWO-LAYER BUCKET MODEL APPROACH ..................................................... 134
1. Instruction .............................................................................................................. 137
2. Methods.................................................................................................................. 140
2.1 Micrometeorological, Soil Moisture and Radiation Data ............................... 140
2.2 Atmospheric Sounding Data ......................................................................... 143
2.3 Development of Empirical Relationships ..................................................... 144
2.4. Two-Layer Bucket Model of Soil Moisture Dynamics ............................... 146
2.4.1 Model Parameters .................................................................................... 149
2.4.2 Model Parameters Input .......................................................................... 150
3. Results and discussion ........................................................................................... 153
10
TABLE OF CONTENTS
3.1 Soil moisture and albedo influence on the planetary boundary layer height .. 153
3.2 Empirical relationships for estimating planetary boundary layer height ........ 154
3.3 Influence of change in precipitation regime on land surface –PBL
dynamics ............................................................................................................... 157
3.4 Implications of modeling two-layer soil moisture dynamics........................ 160
References .................................................................................................................. 161
11
ABSTRACT
Uncertainty of predicted change in precipitation frequency and intensity motivates the
scientific community to better understand, quantify, and model the possible outcome of
dryland ecosystems. In pulse dependent ecosystems (i.e. monsoon driven) soil moisture is
tightly linked to atmospheric processes. Here, I analyze three overarching questions; Q1)
How does soil moisture presence or absence in a shallow or deep layer influence the
surface energy budget and planetary boundary layer characteristics?, Q2) What is the
role of vegetation on ecosystem albedo in the presence or absence of deep soil moisture?,
Q3) Can we develop empirical relationships between soil moisture and the planetary
boundary layer height to help evaluate the role of future precipitation changes in land
surface atmosphere interactions?. To address these questions I use a conceptual
framework based on the presence or absence of soil moisture in a shallow or deep layer. I
define these layers by using root profiles and establish soil moisture thresholds for each
layer using four years of observations from the Santa Rita Creosote Ameriflux site. Soil
moisture drydown curves were used to establish the shallow layer threshold in the
shallow layer, while NEE (Net Ecosystem Exchange of carbon dioxide) was used to
define the deep soil moisture threshold. Four cases were generated using these thresholds:
Case 1, dry shallow layer and dry deep layer; Case 2, wet shallow layer and dry deep
layer; Case 3, wet shallow layer and wet deep layer, and Case 4 dry shallow and wet deep
layer. Using this framework, I related data from the Ameriflux site SRC (Santa Rita
Creosote) from 2008 to 2012 and from atmospheric soundings from the nearby Tucson
Airport; conducted field campaigns during 2011 and 2012 to measure albedo from
12
individual bare and canopy patches that were then evaluated in a grid to estimate the
influence of deep moisture on albedo via vegetation cover change; and evaluated the
potential of using a two-layer bucket model and empirical relationships to evaluate the
link between deep soil moisture and the planetary boundary layer height under changing
precipitation regime. My results indicate that (1) the presence or absence of water in two
layers plays a role in surface energy dynamics, (2) soil moisture presence in the deep
layer is linked with decreased ecosystem albedo and planetary boundary layer height, (3)
deep moisture sustains vegetation greenness and decreases albedo, and (4) empirical
relationships are useful in modeling planetary boundary layer height from dryland
ecosystems. Based on these results we argue that deep soil moisture plays an important
role in land surface-atmosphere interactions.
13
CHAPTER 1: INTRODUCTION
1.1 Background
Almost forty percent of the Earth’s land surface is classified as dryland. Uncertainty in
how these dryland regions will respond to changes in precipitation and temperature is of
great concern for the scientific community and society [D'Odorico et al., 2013;
Rodriguez-Iturbe, 2000; Weltzin et al., 2003; Wiegand and Jeltsch, 2000]. In particular
because the ecosystem services provided by this regions influence the well-being of about
35% of the human population [Dooley, 2005; van Jaarsveld et al., 2005; Vihervaara et
al., 2010].
Drylands are unique in that they are water-limited and potential evapotranspiration
generally exceeds precipitation [Newman et al., 2006]. In addition to these environmental
conditions, drylands face management challenge [Miller, 2005]related to land use change
which are often as a consequence of human and natural disturbance, e.g.
“desertification”, salinization, soil erosion [D'Odorico et al., 2013; Miller, 2005].
Some identified knowledge gaps in our understanding of dryland ecosystems include 1)
ecohydrological impacts of invasive species [Nagler et al., 2009], 2) evaporation
partitioning [Newman et al., 2010], 3) scale and non-linear implications [Asbjornsen et
al., 2011], 4) shrub-grass interactions[Miller, 2005; Newman et al., 2006], 5) soil
moisture dynamics in canopy-intercanopy transitions[Bhark and Small, 2003], 6) water
availability for deep-rooted species as precipitation events become more intense
14
[Breshears et al., 2009; Raz-Yaseef et al., 2012], and 7) land surface - atmosphere models
with good representation of dryland ecohydrological triggers [Wang et al., 2012].
Arid and semiarid regions are characterized by “pulses” of precipitation [Huxman et al.,
2004; Lauenroth and Bradford, 2009] that have an effect on ecosystem functioning and
composition [Aguiar and Sala, 1999]. These “pulses” arrive as packets of moisture, as a
result of small or large precipitation events, that may differentially influence different
species within these arid and semiarid ecosystems. For instance in these regions, small
rainfall events (<5 mm) are especially important for certain functional types such as
grasses [Sala and Lauenroth, 1985] and soil crusts [Belnap et al., 2004].
In semiarid regions “pulses” of precipitation influence all of the components of the water
budget [Weltzin et al., 2003; Wilcox et al., 2006], and especially soil moisture [Entekhabi
et al., 1996]. Importantly, soil moisture links the ecological and climate system via
evapotranspiration and the surface energy budget [Chahine, 1992; Kustas et al., 1991;
Rodriguez-Iturbe, 2000]. Therefore, understanding mechanisms, interactions, thresholds,
and feedbacks between vegetation and the hydrologic cycle is of great importance
especially with anticipated changes in precipitation regimes [Rodriguez-Iturbe, 2000;
Seneviratne et al., 2010].
Soil moisture and vegetation are coupled in such a way that influences land surface
processes. This has been widely investigated especially in tropical and temperate forests
[Hutjes et al., 1998], e.g. by analyzing the effect of tropical deforestation [Berbert and
Costa, 2003; Gash and Shuttleworth, 1991] and temperate afforestation [Swann et al.,
15
2012] on albedo and the consequence of this on the surface energy balance and
furthermore on the climate system.
Processes such as evaporation depend on if soil moisture is within reach of atmospheric
demand [Orlandini, 1999; Wythers et al., 1999], while soil moisture beyond this depth is
transferred through transpiration as a consequence of photosynthesis of vegetation
[Cavanaugh et al., 2011]. Evapotranspiration partitioning is a subject of ongoing research
due to the complexity of surface energy balance processes and measurements
[Cavanaugh et al., 2011; Oren et al., 1998; Scott, 2010; Williams et al., 2004], and of
bellow ground processes such as hydraulic redistribution [Caldwell et al., 1998;
Giambelluca et al., 2009; Unsworth et al., 2004]. The partitioning of this is of great
importance, for instance, where there is a need to quantify the effects of soil moisture on
bare and vegetated patches individually to better predict the influence of vegetation
change (i.e. woody encroachment) [Rothfuss et al., 2010; Sakaguchi and Zeng, 2009;
Yepez et al., 2005] on the climate system.
Woody vegetation in semiarid regions is expected to change the ratio of transpiration to
total evapotranspiration because of changes in leaf area, volume of roots, and phenology
[Huxman et al., 2005; Rodriguez-Iturbe et al., 1999]. Changes in the land surface
characteristics as a result of shifts in vegetation impact more than one land surface
process. For instance, canopy architecture and greenness can influence albedo [Asner,
1998]. Moreover, belowground changes as a result of shifts in vegetation, such as root
16
distribution, also have the potential to influence land surface processes such as
transpiration [R Jackson et al., 2000].
The planetary boundary layer depends on mechanisms such as latent and sensible heat
partitioning, which can be influenced by transpiration if soil moisture (>20 cm) is
available [Basara and Crawford, 2002]. Latent and sensible heat from the land surface to
the atmosphere modify the lowest 100 to 3000 m, leading to the formation of this
planetary boundary layer [Stull, 1988]. Early studies on the planetary boundary layer
focused on the processes that affect air buoyancy and therefore the turbulence and height
of this layer, leading to linking the surface energy budget to the planetary boundary layer
[Bjerken, 1900; Ruzin, 1963]. More recently, research has focused on developing
diagnostic tools and relationships that couple land surface processes and the planetary
boundary layer [Chen et al., 2012; Kirtman et al., 2009; Santanello et al., 2013].
Albedo (the ratio of outgoing shortwave radiation to incoming shortwave radiation) is a
land surface property that is related to both vegetation and soil moisture [Charney, 1975;
Idso et al., 1975; Richardson et al., 2013]. For instance, as vegetation grows in the
presence of soil moisture, albedo decreases and as vegetation senesces, albedo increases,
making it a useful metric for analyzing phenological processes [Zhang et al., 2005]. As
another example, biophysical parameters such as albedo are used to study vegetation
processes using remote sensing in natural [Mendez-Barroso and Vivoni, 2010; Otterman,
1977] or agricultural [Mokhtari et al., 2013] systems.
17
Generally albedo decreases with a darker surface [Twomey et al., 1986]. For instance, the
presence of soil moisture darkens bare soil and changes albedo [Idso et al., 1975; R D
Jackson et al., 1976; Lobell and Asner, 2002] and the green-up of vegetation does as well
[Asner, 1998; Berbert and Costa, 2003; Song, 1999; Zhang et al., 2005]. However,
whether the soil or the vegetation has a stronger influence on albedo is not clear
[Domingo et al., 2000], and at what depth the presence or absence of moisture will
influence the surface and as a consequence impact albedo is also not well understood
[Small and Kurc, 2003]. For instance, albedo plays a key role in sensible and latent heat
partitioning, especially when moisture is present in a shallow layer [Baldocchi et al.,
2004; Domingo et al., 2000; Small and Kurc, 2003], but less is understood of the effect
when there is moisture in a deep layer [Bracho et al., 2008].
Modeling efforts at different scales have improved our understanding of land surfaceatmospheric interactions [Pitman, 2003; Rosolem et al., 2010; Seneviratne et al., 2010;
Shuttleworth, 1994]. However there is still a need to link these interactions at multiple
scales and to engage in understanding on how ecosystems will trigger and feedback to
future climate perturbations [Asbjornsen et al., 2011].
To contribute to these knowledge gaps, my doctoral research is focused in the study of
the presence or absence of soil moisture at different depths and its overall influence on
radiation and surface energy budget dynamics and planetary boundary layer
characteristics. This research will result in three peer-reviewed research papers. My first
paper assesses the influence of two-layer soil moisture on radiation and surface energy
18
dynamics and planetary boundary layer characteristics by coupling continuously
monitored ecosystem level measurements of soil moisture and surface energy fluxes
within a desert shrubland using the eddy covariance technique and atmospheric sounding
data. My second paper builds off of the findings from my first paper in the design of a
field campaign to address the differential influence of vegetation patches on ecosystem
albedo due to the influence of deep soil moisture on vegetation dynamics. Finally, in my
third paper, using my findings from my first and second papers I will incorporate deep
soil moisture dynamics into a simple land-surface modeling approach linking a bucket
model with empirical relationships to assess the value of including this information on
applications related to land-atmosphere interactions.
1.2 Research Objectives
The overall objective of this research is to contribute to watershed management and
ecohydrology in arid and semiarid ecosystems. Arid and semiarid ecosystems face many
challenges in both the near and distant future. The success of mitigation, conservation,
and managing strategies to address these challenges will depend on both scientific and
social effort .
The training that I have gained through this study include critical thinking, the use of
emerging tools such as eddy covariance, and experience in land surface – atmosphere
modeling, have prepared me for these challenges, enabling me to contribute in research
and mentoring as I return to Mexico.
19
With this in mind, my research aims to answer three broad questions:
1) How does soil moisture presence or absence in a shallow or deep layer influence the
surface energy budget and planetary boundary layer characteristics?
Understanding soil moisture control on land-atmosphere interactions will become
increasingly important as global changes continue to alter the water availability of our
ecosystems. For my analysis, I used four years of data from the Santa Rite Creosote
Ameriflux site. I categorized each day of our record into (1) wet or dry seasons and (2)
one of four Cases within a two-layer framework based on the presence or absence of
moisture in a shallow and a deep soil layer. Using these categorizations I quantified the
soil moisture control on the radiation and surface energy budgets and planetary boundary
layer characteristics using both average responses and linear regression.
2) What is the role of vegetation on ecosystem albedo in the presence or absence of deep
soil moisture?
Because vegetation plays a critical role in ecosystem processes, understanding what
controls vegetation change and when is it triggered is especially important in arid and
semiarid regions. To address this, I made use of the two-layer conceptual framework. I
hypothesized that when the vegetation had access to soil moisture available only in the
shallow layer (Case 2), albedo would be higher than when the vegetation had soil
moisture available in the shallow and deep layer (Case 3) or deep layer (Case 4). With
this we know that deep soil moisture is linked to surface albedo. I further hypothesized
20
that when the shallow layer is wet (Case 2 and 3), bare albedo will be lower than when
the shallow layer is dry (Case 1 and 4). To test this hypothesis I designed a field
campaign to measure albedo above canopy and bare patches for each Case. These Cases
were identified in real-time using data from a nearby meteorological station at a local
highschool ( http://sahuarita.cals.arizona.edu/). Field measurements were taken in two
locations within the footprint of an eddy covariance tower, at multiple bare and canopy
patches. Probability distribution functions (PDFs) were developed from these albedo data
collected from the field campaigns. I then populated a 100x100 grid of bare and canopy
cells with albedo values weighted by these PDFs, and increased the canopy cover from 0
to 100% to look at the influence of the deep moisture on albedo by its influence on the
vegetation. Long-term monitoring with pheno-cams proved an important tool for better
understanding the vegetation phenology that was linked to deep moisture.
3) Can we use empirical relationships between soil moisture and planetary boundary
layer height to evaluate future precipitation changes?
One major knowledge gap in dryland ecohydrology is a good representation of the major
triggers in land surface – atmosphere interactions and how they are expressed in
modeling contexts.
To address this need, I followed an empirical approach, identifying the relationships
between shallow and deep soil moisture in addition to albedo on the planetary boundary
layer height. I used polynomial relationships to estimate the planetary boundary layer
height as a function of combining shallow soil moisture, deep soil moisture and albedo.
21
Then, I modeled soil moisture dynamics using a two-layer bucket model approach and
used these modeled results, and estimate albedo and the planetary boundary layer height.
Finally precipitation was modeled stochastically to generate treatments based on future
precipitation scenarios such as a 20% increase or decrease in average annual rainfall and
evaluate the potential for this empirical approach for looking at climate change
feedbacks.
Ultimately my doctoral research contributes to our ecohydrological knowledge on
thresholds, mechanisms, and interactions between the land surface and the atmosphere
through (1) analyzing long term data sets, (2) developing and analyzing field campaigns,
and (3) modeling exercises. Furthermore, my doctoral research is at the boundary
between ecohydrology and hydrometeorology, both interdisciplinary sciences applied to
better understand watershed processes and add to management and conservation
strategies.
1.3 Dissertation Format
Following the introduction is the “Present Study” which provides a summary of the
analyses and conclusions of three papers prepared for peer-reviewed publication, the first
submitted to Water Resources Research and the third to be submitted to the Advances in
Water Resources. The three articles included as appendices are as follow:
22
APPENDIX A: OBSERVATION OF A TWO-LAYER SOIL MOISTURE
INFLUENCE ON SURFACE ENERGY DYNAMICS AND PLANETARY
BOUNDARY LAYER CHARACTERISTICS IN A SEMIARID SHRUBLAND
APPENDIX B: QUANTIFYING THE INFLUENCE OF DEEP SOIL
MOISTURE ON ECOSYSTEM ALBEDO: THE ROLE OF VEGETATION
APPENDIX C: EMPIRICAL RELATIONSHIPS BETWEEN SOIL MOISTURE,
ALBEDO, AND THE PLANETARY BOUNDARY LAYER HEIGHT: A TWOLAYER BUCKET MODEL APPROACH
Altogether these three papers discuss important findings to improve what we know of
ecohydrology in semiarid and arid environments, improve our understanding of how
“pulse” ecosystem respond to soil moisture dynamics, and how soil moisture at different
depths influences land surface – atmosphere interactions through influencing the surface
energy dynamics, vegetation characteristics, and the development of the planetary
boundary layer height.
23
CHAPTER 2: PRESENT STUDY
The methodology, results, and conclusions of this study are presented in three papers
appended to this dissertation. These papers follow a logical progression from (1) the
analysis of continuous long term data, (2) the testing of hypotheses developed from that
analysis through a field campaign, and (3) a modeling exercise based on an empirical
approach that emphasizes the use of readily available data that support the concepts
developed through the analysis of the long term and field campaign data. The following
is a summary of important findings and take home points from each of these appendices.
2.1 APPENDIX A: OBSERVATION OF A TWO-LAYER SOIL MOISTURE
INFLUENCE ON SURFACE ENERGY DYNAMICS AND PLANETARY
BOUNDARY LAYER CHARACTERISTICS IN A SEMIARID SHRUBLAND
I present an observational analysis examining soil moisture control on surface energy
dynamics and planetary boundary layer characteristics. Understanding soil moisture
control on land-atmosphere interactions will become increasingly important as global
changes continue to alter the water availability of our ecosystems. For my analysis, I used
four years of data from the Santa Rite Creosote Ameriflux site. I categorized each day of
our record into (1) wet or dry seasons and (2) one of four Cases within a two-layer
framework based on the presence or absence of moisture in a shallow and a deep soil
layer. Using these categorizations I quantified the soil moisture control on the radiation
and surface energy budgets and planetary boundary layer characteristics using both
average responses and linear regression. My results highlight the importance of deep soil
24
moisture in land surface – atmosphere interactions. For instance, albedo for Cases with
any moisture were significantly lower than for a completely dry soil profile, even if this
moisture was only below the surface. Also, while the planetary boundary layer height
PBLh was largest when the whole soil profile was dry and smallest when the whole
profile was wet, even when shallow moisture was absent but deep moisture was present
the PBLh was significantly lower than when the entire profile was dry. This deep
moisture importance is likely site-specific and modulated through vegetation. Therefore
understanding these relationships also provides important insights into feedbacks
between vegetation and the hydrologic cycle and their consequent influence on the
climate system.
2.2 APPENDIX B: QUANTIFYING THE INFLUENCE OF DEEP SOIL MOISTURE
ON ECOSYSTEM ALBEDO: THE ROLE OF VEGETATION
I present an observational analysis from a field campaign designed to examine the
influence of two-layer soil moisture control on the albedo from bare and canopy patches.
Because changes in precipitation dynamics will continue to alter the water availability of
our ecosystems, understanding the feedbacks between the vegetation and the hydrologic
cycle and their consequent influence on the climate system will become increasingly
important. For my analysis, I conducted a series of field campaigns over two years (2011
– 2012) within the tower footprint of the Santa Rite Creosote Ameriflux site. The timing
of our campaigns always fell into one of four Cases within a two-layer framework based
25
on the presence or absence of moisture in a shallow and a deep soil layer. Using these
Cases I quantified the soil moisture control on the albedo of bare and canopy patches
using average responses. My results highlight the importance of deep soil moisture in
land surface – atmosphere interactions through its influence on aboveground vegetation
characteristics. Despite a dry shallow layer in both Cases, ecosystem albedo for the Case
when soil moisture was present in the deep layer was significantly lower than for the
Case when soil moisture was absent in the deep layer. Greenness of the vegetation
differed between these two Cases, higher in the Case with soil moisture in the deep layer,
suggesting that deep moisture is expressing itself at the surface through an increase in
greenness of the vegetation. Understanding these relationships between vegetation and
deep soil moisture will provide important insights into feedbacks between the hydrologic
cycle and the climate system.
2.3 APPENDIX C: EMPIRICAL RELATIONSHIPS BETWEEN SOIL MOISTURE,
ALBEDO, AND THE PLANETARY BOUNDARY LAYER HEIGHT: A TWO-LAYER
BUCKET MODEL APPROACH
My study presents a modeling approach to study the soil moisture and albedo control on
planetary boundary layer height (PBLh). Interactions between the land surface and the
atmosphere regulate the coupled soil moisture-precipitation system. This is especially
important in semiarid regions, where water resources are limited, as precipitation
dynamics are changing. I used data from the Santa Rita Creosote Ameriflux site and
Tucson Airport atmospheric sounding to generate empirical relationships between soil
26
moisture, albedo and PBLh. Empirical relationships show that ~50% of PBLh can be
explained by soil moisture and albedo. Then, a two-layer bucket model was used to
model soil moisture, and was coupled to these empirical relationships to model PBLh. As
part of this modeling exercise, I explored soil moisture dynamics under three different
precipitation regimes: current, decreased, and increased, to look at the influence on soil
moisture on land surface-atmospheric processes. While my precipitation regimes are
simple, they represent future precipitation regimes that can influence the two layers in our
conceptual framework: an increase in annual precipitation, could have an impact on deep
soil moisture and atmospheric processes if the precipitation event is intense, while a
decrease in annual precipitation, has more of an impact on the soil moisture at the
shallow layer. I observed that the response of soil moisture, albedo, and the PBLh will
depend not only on annual change, but on the frequency and intensity of this change. I
argue that because albedo and soil moisture data are readily available at multiple
temporal and spatial scales, developing empirical relationships that can be used in land
surface – atmosphere applications are of great value.
2.4 Future Research Opportunities
Here, I outline suggestions for future research opportunities that could follow research
development from my dissertation.
2.4.1 Land surface- atmosphere interactions: dryland ecohydrology modeling
27
A knowledge gap previously identified is that land surface - atmosphere models need a
better representation of dryland ecohydrological processes [Wang et al., 2012].
Ecohydrological processes in arid and semiarid regions can be better understood using a
two-layer conceptual framework [Kurc and Small, in review; Sanchez-Mejia and Papuga,
in prep; in review; Sanchez-Mejia et al., in review]. The reasoning for this is that the
presence or absence of soil moisture will trigger biological, hydrological, surface energy
and planetary boundary layer processes [Basara and Crawford, 2002; Cavanaugh et al.,
2011; Kurc and Small, 2004; Sanchez-Mejia and Papuga, in prep]. For instance, soil
moisture in the shallow layer of bare areas controls soil evaporation and bare soil albedo
[Idso et al., 1975; Sanchez-Mejia et al., in review], while soil moisture in the deep layer
of a shrub controls transpiration [Cavanaugh et al., 2011], greenness[Kurc and Benton,
2010] and canopy albedo [Sanchez-Mejia et al., in review]. However, these controls may
be limited to warm deserts and therefore may be site-specific. For instance, in the
Mojave desert the behavior of the controls could be different, so further research in sitespecific local triggers will enrich the good representation of drylands in land surfaceatmospheric interactions models.
While previous research has suggested diagnostic parameters such as stability or
evaporative fraction as controls on land surface- atmospheric interactions [Santanello et
al., 2013], I show that other influences such as albedo and evaporative fraction indirectly
incorporate the role of vegetation as part of the land surface influencing the planetary
boundary layer height. However, the hierarchy of these parameters and the control they
have on planetary boundary layer height in semiarid regions should be further analyzed.
28
I plan to explore the impact of vegetation change on land surface- atmospheric
interactions using a Daisy World approach in combination with the two-layer bucket
model to get insight on how future precipitation regime and percent cover will influence
soil moisture, albedo, and therefore the planetary boundary layer height.
The “Daisy World” is a model developed by [Watson and Lovelock, 1983] to examine
feedbacks between the climate and terrestrial systems, in which competition for bare
space is based on the albedo of dark and bright daisies, and the ideal temperature at which
they can survive. Modifies the model so that competition is based on “wet” or “dry”
daisies and incorporates a transpiration. Because it is a 2-D model, meaning that bare and
canopy patches can be represented by independent cells, I will use this approach and
integrate it with the two-layer bucket model.
2.4.2 The shrubby “gray” –zone of woody encroachment
Better understanding of the ecohydrological impact of woody encroachment is priority
for more than one science discipline [Newman et al., 2006]. Woody encroachment
research has focused on species such as creosotebush [Kurc and Small, 2004], mesquite
[Archer, 1990], pinon-juniper and on grasses and their possible responses to warmer and
drier climates . However less worked has focused in the “gray” zone that cacti represent
[Ansley and Castellano, 2007; Unland et al., 1996].
Plant physiology has an effect on soil moisture and transpiration processes , and therefore
cacti which have a specific Crassulacean Acid Metabolisme (CAM) mechanism for
29
avoiding water stress. However, little is known about what controls their phenological
changes and the associated changes in albedo that accompany those phenological
changes.
In future work it would be beneficial to use this field and modeling approach using more
than one functional type to better represent ecosystem dynamics in the face of less
precipitation and warmer temperatures.
REFERENCES
Aguiar, M., and O. Sala (1999), Patch structure, dynamics and implications for the
functioning of arid ecosystems, Trends in Ecology & Evolution, 14(7), 273-277.
Ansley, R. J., and M. J. Castellano (2007), Prickly pear cactus responses to summer and
winter fires, Rangeland Ecology & Management, 60(3), 244-252.
Archer, S. (1990), Development and stability of grass woody mosaics in subtropical
savanna parkland, Texas, USA, Journal of Biogeography, 17(4-5), 453-462.
Asbjornsen, H., et al. (2011), Ecohydrological advances and applications in plant-water
relations research: a review, Journal of Plant Ecology, 4(1-2), 3-22.
Asner, G. (1998), Biophysical and biochemical sources of variability in canopy
reflectance, Remote Sensing of Environment, 64(3), 234-253.
30
Baldocchi, D., L. Xu, and N. Kiang (2004), How plant functional-type, weather, seasonal
drought, and soil physical properties alter water and energy fluxes of an oak-grass
savanna and an annual grassland, Agricultural and Forest Meteorology, 123(1-2),
13-39.
Basara, J., and K. Crawford (2002), Linear relationships between root-zone soil moisture
and atmospheric processes in the planetary boundary layer, Journal of Geophysical
Research-Atmospheres, 107(D15).
Belnap, J., S. Phillips, and M. Miller (2004), Response of desert biological soil crusts to
alterations in precipitation frequency, Oecologia, 141(2), 306-316.
Berbert, M., and M. Costa (2003), Climate change after tropical deforestation: Seasonal
variability of surface albedo and its effects on precipitation change, Journal of
Climate, 16(12), 2099-2104.
Bhark, E. W., and E. E. Small (2003), Association between plant canopies and the spatial
patterns of infiltration in shrubland and grassland of the Chihuahuan Desert, New
Mexico, Ecosystems, 6(2), 185-196.
Bjerken, V. (1900), Circulation in the atmosphere, edited, Physikalische Zeitschrift.
Bracho, R., T. L. Powell, S. Dore, J. H. Li, C. R. Hinkle, and B. G. Drake (2008),
Environmental and biological controls on water and energy exchange in Florida
31
scrub oak and pine flatwoods ecosystems, Journal of Geophysical ResearchBiogeosciences, 113(G2), 13.
Breshears, D., O. Myers, and F. Barnes (2009), Horizontal heterogeneity in the frequency
of plant-available water with woodland intercanopy-canopy vegetation patch type
rivals that occuring vertically by soil depth, Ecohydrology, 2(4), 503-519.
Caldwell, M. M., T. E. Dawson, and J. H. Richards (1998), Hydraulic lift: Consequences
of water efflux from the roots of plants, Oecologia, 113(2), 151-161.
Cavanaugh, M., S. Kurc, and R. Scott (2011), Evapotranspiration partitioning in semiarid
shrubland ecosystems: a two-site evaluation of soil moisture control on transpiration,
Ecohydrology, 4(5), 671-681.
Chahine, M. (1992), The Hydrological Cycle and its influence on Climate, Nature,
359(6394), 373-380.
Charney, J. G. (1975), DYNAMICS OF DESERTS AND DROUGHT IN SAHEL,
Quarterly Journal of the Royal Meteorological Society, 101(428), 193-202.
Chen, X., Z. Su, Y. Ma, and F. Sun (2012), Analysis of Land-Atmosphere Interactions
over the North Region of Mt. Qomolangma (Mt. Everest), Arctic Antarctic and
Alpine Research, 44(4), 412-422.
D'Odorico, P., A. Bhattachan, K. Davis, S. Ravi, and C. Runyan (2013), Global
desertification: Drivers and feedbacks, Advances in Water Resources, 51, 326-344.
32
Domingo, F., L. Villagarcia, A. Brenner, and J. Puigdefabregas (2000), Measuring and
modelling the radiation balance of a heterogeneous shrubland, Plant Cell and
Environment, 23(1), 27-38.
Dooley, E. (2005), Millennium ecosystem assessment, Environmental Health
Perspectives, 113(9), A591-A591.
Entekhabi, D., I. Rodriguez-Iturbe, and F. Castelli (1996), Mutual interaction of soil
moisture state and atmospheric processes, Journal of Hydrology, 184(1-2), 3-17.
Gash, J., and W. Shuttleworth (1991), Tropical deforestation- albedo and the surfaceenergy balance, Climatic Change, 19(1-2), 123-133.
Giambelluca, T. W., F. G. Scholz, S. J. Bucci, F. C. Meinzer, G. Goldstein, W. A.
Hoffmann, A. C. Franco, and M. P. Buchert (2009), Evapotranspiration and energy
balance of Brazilian savannas with contrasting tree density, Agricultural and Forest
Meteorology, 149(8), 1365-1376.
Hutjes, R., et al. (1998), Biospheric aspects of the hydrological cycle - Preface, Journal
of Hydrology, 212(1-4), 1-21.
Huxman, T., K. Snyder, D. Tissue, A. Leffler, K. Ogle, W. Pockman, D. Sandquist, D.
Potts, and S. Schwinning (2004), Precipitation pulses and carbon fluxes in semiarid
and arid ecosystems, OECOLOGIA, 141(2), 254-268.
33
Huxman, T., B. Wilcox, D. Breshears, R. Scott, K. Snyder, E. Small, K. Hultine, W.
Pockman, and R. Jackson (2005), Ecohydrological implications of woody plant
encroachment, Ecology, 86(2), 308-319.
Idso, S., R. Jackson, R. Reginato, B. Kimball, and F. Nakayama (1975), Dependence of
bare soil albedo on soil-water content, Journal of Applied Meteorology, 14(1), 109113.
Jackson, R., et al. (2000), Belowground consequences of vegetation change and their
treatment in models, Ecological Applications, 10(2), 470-483.
Jackson, R. D., S. B. Idso, and R. J. Reginato (1976), Calculation of evaporation rates
during transition from energy-limitig to soil-limiting phases using albedo data, Water
Resources Research, 12(1), 23-26.
Kirtman, B. P., D. M. Straus, D. Min, E. K. Schneider, and L. Siqueira (2009), Toward
linking weather and climate in the interactive ensemble NCAR climate model,
Geophysical Research Letters, 36.
Kurc, S., and E. Small (2004), Dynamics of evapotranspiration in semiarid grassland and
shrubland ecosystems during the summer monsoon season, central New Mexico,
Water Resources Research, 40(9).
34
Kurc, S., and L. Benton (2010), Digital image-derived greenness links deep soil moisture
to carbon uptake in a creosotebush-dominated shrubland, Journal of Arid
Environments, 74(5), 585-594.
Kurc, S., and E. Small (in review), Simple ecohydrological models: comparison to
observations of dryland water and carbon fluxes, edited, Adv. Water Resour.
Kustas, W., et al. (1991), An Interdisciplinary field-study of the energy ans water fluxes
in the atmosphere-biosphere system over semiarid rangelands -description and some
preliminary-results, Bulletin of the American Meteorological Society, 72(11), 16831705.
Lauenroth, W., and J. Bradford (2009), Ecohydrology of dry regions of the United States:
precipitation pulses and intraseasonal drought, Ecohydrology, 2(2), 173-181.
Lobell, D. B., and G. P. Asner (2002), Moisture effects on soil reflectance, Soil Science
Society of America Journal, 66(3), 722-727.
Mendez-Barroso, L., and E. Vivoni (2010), Observed shifts in land surface conditions
during the North American Monsoon: Implications for a vegetation-rainfall feedback
mechanism, Journal of Arid Environments, 74(5), 549-555.
Miller, M. (2005), The structure and functioning of dryland ecosystem - conceptual
models to informe long-term monitoring, edited, p. 73, U.S. Geological Survey
Scientific Investigations.
35
Mokhtari, M. H., R. Adnan, and I. Busu (2013), A new approach for developing
comprehensive agricultural drought index using satellite-derived biophysical
parameters and factor analysis method, Natural Hazards, 65(3), 1249-1274.
Nagler, P. L., K. Morino, K. Didan, J. Erker, J. Osterberg, K. R. Hultine, and E. P. Glenn
(2009), Wide-area estimates of saltcedar (Tamarix spp.) evapotranspiration on the
lower Colorado River measured by heat balance and remote sensing methods,
Ecohydrology, 2(1), 18-33.
Newman, B., D. Breshears, and M. Gard (2010), Evapotranspiration Partitioning in a
Semiarid Woodland: Ecohydrologic Heterogeneity and Connectivity of Vegetation
Patches, Vadose Zone Journal, 9(3), 561-572.
Newman, B., B. Wilcox, S. Archer, D. Breshears, C. Dahm, C. Duffy, N. McDowell, F.
Phillips, B. Scanlon, and E. Vivoni (2006), Ecohydrology of water-limited
environments: A scientific vision, Water Resources Research, 42(6).
Oren, R., N. Phillips, G. Katul, B. E. Ewers, and D. E. Pataki (1998), Scaling xylem sap
flux and soil water balance and calculating variance: a method for partitioning water
flux in forests, Annales Des Sciences Forestieres, 55(1-2), 191-216.
Orlandini, S. (1999), Two-Layer Model of Near-Surface Soil Drying for TimeContinuous Hydrologic Simulations, J. Hydrol. Eng., 4(2), 91-99.
36
Otterman, J. (1977), Monitoring surface albedo change with LANDSAT, Geophysical
Research Letters, 4(10), 441-444.
Pitman, A. (2003), The evolution of, and revolution in, land surface schemes designed for
climate models, International Journal of Climatology, 23(5), 479-510.
Raz-Yaseef, N., D. Yakir, G. Schiller, and S. Cohen (2012), Dynamics of
evapotranspiration partitioning in a semi-arid forest as affected by temporal rainfall
patterns, Agricultural and Forest Meteorology, 157, 77-85.
Richardson, A. D., T. F. Keenan, M. Migliavacca, Y. Ryu, O. Sonnentag, and M.
Toomey (2013), Climate change, phenology, and phenological control of vegetation
feedbacks to the climate system, Agricultural and Forest Meteorology, 169, 156-173.
Rodriguez-Iturbe, I. (2000), Ecohydrology: A hydrologic perspective of climate-soilvegetation dynamics, Water Resources Research, 36(1), 3-9.
Rodriguez-Iturbe, I., P. D'Odorico, A. Porporato, and L. Ridolfi (1999), On the spatial
and temporal links between vegetation, climate, and soil moisture, Water Resources
Research, 35(12), 3709-3722.
Rosolem, R., W. Shuttleworth, X. Zeng, S. Saleska, and T. Huxman (2010), Land surface
modeling inside the Biosphere 2 tropical rain forest biome, Journal of Geophysical
Research-Biogeosciences, 115.
37
Rothfuss, Y., P. Biron, I. Braud, L. Canale, J.-L. Durand, J.-P. Gaudet, P. Richard, M.
Vauclin, and T. Bariac (2010), Partitioning evapotranspiration fluxes into soil
evaporation and plant transpiration using water stable isotopes under controlled
conditions, Hydrological Processes, 24(22), 3177-3194.
Ruzin, M. I. (1963), Effect of the horizontal temperature gradient on the planetary
boundary layer of the atmosphere, edited, Bulletin of the Academy of Sciences of the
USSR, Geophysics Serie, 3, 306-308.
Sakaguchi, K., and X. Zeng (2009), Effects of soil wetness, plant litter, and under-canopy
atmospheric stability on ground evaporation in the Community Land Model
(CLM3.5), Journal of Geophysical Research-Atmospheres, 114.
Sala, O., and W. Lauenroth (1985), Root profiles and the ecological effect of light
rainshowers in arid and semiarid regions, American Midland Naturalist, 114(2), 406408.
Sanchez-Mejia, Z., and S. Papuga (in prep), Empirical relationships between soil
moisture, albedo, and the planetary boundary layer height: a two-layer bucket model
approach, edited, Journal of Hydrometeorology.
Sanchez-Mejia, Z., and S. Papuga (in review), Observation of a two-layer soil moisture
influence on surface energy dynamics and planetary boundary layer characteristics in
a semiarid shrubland, edited, Water Resources Research.
38
Sanchez-Mejia, Z., S. Papuga, and J. Swetish (in review), Quantifying the influence of
deep soil moisture on ecosystem albedo: the role of vegetation, edited, Water
Resources Research.
Santanello, J., C. Peters-Lidard, A. Kennedy, and S. Kumar (2013), Diagnosing the
Nature of Land-Atmosphere Coupling: A Case Study of Dry/Wet Extremes in the US
Southern Great Plains, Journal of Hydrometeorology, 14(1), 3-24.
Scott, R. (2010), Using watershed water balance to evaluate the accuracy of eddy
covariance evaporation measurements for three semiarid ecosystems, Agricultural
and Forest Meteorology, 150(2), 219-225.
Seneviratne, S., T. Corti, E. Davin, M. Hirschi, E. Jaeger, I. Lehner, B. Orlowsky, and A.
Teuling (2010), Investigating soil moisture-climate interactions in a changing
climate: A review, Earth-Science Reviews, 99(3-4), 125-161.
Shuttleworth, W. (1994), Large-scale experimental and modeling studies of hydrological
processes, Ambio, 23(1), 82-86.
Small, E., and S. Kurc (2003), Tight coupling between soil moisture and the surface
radiation budget in semiarid environments: Implications for land-atmosphere
interactions, Water Resources Research, 39(10).
Song, J. (1999), Phenological influences on the albedo of prairie grassland and crop
fields, Int. J. Biometeorol., 42(3), 153-157.
39
Stull, R. B. (1988), An Introduction to Boundary Layer Meteorology, edited, p. 666,
Kluwer Acad. Publ., Boston, London.
Swann, A., I. Fung, and J. Chiang (2012), Mid-latitude afforestation shifts general
circulation and tropical precipitation, Proceedings of the National Academy of
Sciences of the United States of America, 109(3), 712-716.
Twomey, S. A., C. F. Bohren, and J. L. Mergenthaler (1986), Reflectance and albedo
differences between wet and dry surfaces, Appl. Optics, 25(3), 431-437.
Unland, H., P. Houser, W. Shuttleworth, and Z. Yang (1996), Surface flux measurement
and modeling at a semi-arid Sonoran Desert site, Agricultural and Forest
Meteorology, 82(1-4), 119-153.
Unsworth, M. H., N. Phillips, T. Link, B. J. Bond, M. Falk, M. E. Harmon, T. M.
Hinckley, D. Marks, and K. T. P. U (2004), Components and controls of water flux
in an old-growth Douglas-fir-western hemlock ecosystem, Ecosystems, 7(5), 468481.
van Jaarsveld, A., R. Biggs, R. Scholes, E. Bohensky, B. Reyers, T. Lynam, C. Musvoto,
and C. Fabricius (2005), Measuring conditions and trends in ecosystem services at
multiple scales: the Southern African Millennium Ecosystem Assessment (SAfMA)
experience, Philosophical Transactions of the Royal Society B-Biological Sciences,
360(1454), 425-441.
40
Vihervaara, P., M. Ronka, and M. Walls (2010), Trends in Ecosystem Service Research:
Early Steps and Current Drivers, Ambio, 39(4), 314-324.
Wang, L., P. D'Odorico, J. P. Evans, D. J. Eldridge, M. F. McCabe, K. K. Caylor, and E.
G. King (2012), Dryland ecohydrology and climate change: critical issues and
technical advances, Hydrol. Earth Syst. Sci., 16(8), 2585-2603.
Watson, A. J., and J. E. Lovelock (1983), Biological homeostasis of the global
environment-the parable of Daisyworld, Tellus Series B-Chemical and Physical
Meteorology, 35(4), 284-289.
Weltzin, J., et al. (2003), Assessing the response of terrestrial ecosystems to potential
changes in precipitation, Bioscience, 53(10), 941-952.
Wiegand, T., and F. Jeltsch (2000), Long-term dynamics in arid and semiarid ecosystems
- synthesis of a workshop, Plant Ecology, 150(1-2), 3-6.
Wilcox, B., S. Dowhower, W. Teague, and T. Thurow (2006), Long-term water balance
in a semiarid shrubland, Rangeland Ecology & Management, 59(6), 600-606.
Williams, D. G., et al. (2004), Evapotranspiration components determined by stable
isotope, sap flow and eddy covariance techniques, Agricultural and Forest
Meteorology, 125(3-4), 241-258.
Wythers, K., W. Lauenroth, and J. Paruelo (1999), Bare-soil evaporation under semiarid
field conditions, Soil Science Society of America Journal, 63(5), 1341-1349.
41
Yepez, E. A., T. E. Huxman, D. D. Ignace, N. B. English, J. F. Weltzin, A. E.
Castellanos, and D. G. Williams (2005), Dynamics of transpiration and evaporation
following a moisture pulse in semiarid grassland: A chamber-based isotope method
for partitioning flux components, Agric. For. Meteorol., 132(3-4), 359-376.
Zhang, Y., E. Munkhtsetseg, T. Kadota, and T. Ohata (2005), An observational study of
ecohydrology of a sparse grassland at the edge of the Eurasian cryosphere in
Mongolia, Journal of Geophysical Research-Atmospheres, 110(D14).
42
APPENDIX A: OBSERVATIONS OF A TWO-LAYER SOIL MOISTURE
INFLUENCE ON SURFACE ENERGY DYNAMICS AND PLANETARY
BOUNDARY LAYER CHARACTERISTICS IN A SEMIARID SHRUBLAND
Zulia Mayari Sanchez-Mejiaa and Shirley A. Papugaa
Submitted to Water Resources Research
a
School of Natural Resources and the Environment, University of Arizona,
Tucson,AZ 85719
43
Abstract
We present an observational analysis examining soil moisture control on surface energy
dynamics and planetary boundary layer characteristics. Understanding soil moisture
control on land-atmosphere interactions will become increasingly important as climate
change continues to alter water availability. In this study, we analyzed four years of data
from the Santa Rita Creosote Ameriflux site. We categorized our data independently in
two ways (1) wet or dry seasons, and (2) one of four Cases within a two-layer soil
moisture framework based on the presence or absence of moisture in shallow (0-20 cm)
and deep (20 - 60 cm) soil layers. Using these categorizations, we quantified the soil
moisture control on surface energy dynamics and planetary boundary layer characteristics
using both average responses and linear regression. Our results highlight the importance
of deep soil moisture in land surface – atmosphere interactions. The presence of deep soil
moisture decreased albedo by about 10 % and significant differences were observed in
evaporative fraction even in the absence of shallow moisture. The planetary boundary
layer height PBLh was largest when the whole soil profile was dry, decreasing by about 1
km when the whole profile was wet. Even when shallow moisture was absent but deep
moisture was present the PBLh was significantly lower than when the entire profile was
dry. This deep moisture importance is likely site-specific and modulated through
vegetation. Therefore understanding these relationships also provides important insights
into feedbacks between vegetation and the hydrologic cycle and their consequent
influence on the climate system.
44
Keywords:
albedo; available energy; evaporative fraction; eddy covariance; creosotebush; Larrea
tridentata, Santa Rita Experimental Range
45
1. Introduction
The land surface and the atmosphere are tightly coupled through the exchange of energy
and water [Nicholson, 2000; Shukla and Mintz, 1982; Shuttleworth, 1991]. Soil moisture
plays an important role in this exchange [e.g. Seneviratne et al., 2010; Vereecken et al.,
2008] through the partitioning of available energy into sensible and latent heating
[Brubaker and Entekhabi, 1996; Colby, 1984]. Soil moisture control on the exchange of
energy and water is especially strong in dryland ecosystems [Small and Kurc, 2003;
Vivoni et al., 2008; Williams and Albertson, 2004]. Because over 40% of the Earth’s land
surface can be classified as arid to semiarid [e.g. Okin et al., 2009; Reynolds et al., 2007],
understanding soil moisture control on the interactions between the land surface and the
atmosphere will be critical for anticipating feedbacks associated with global change
[Betts, 2000; D'Odorico et al., 2013; Taylor et al., 2002].
Interest in the soil moisture influence on land surface-atmosphere interactions arises from
observations of feedbacks between soil moisture and precipitation, where high rainfall
may lead to increased soil moisture which, in turn, promotes increased rainfall, or vice
versa [Findell and Eltahir, 1997; Koster et al., 2003]. One possible mechanism for this
feedback is through direct contribution of surface moisture to precipitation through
increased evapotranspiration (ET) [Dekker et al., 2007; Dominguez et al., 2008; Eltahir
and Bras, 1996]. However, soil moisture may also influence on the surface energy budget
46
and its relationship to the development of the planetary boundary layer (PBL) [Betts et
al., 1996; Eltahir, 1998].
The surface energy budget can be summarized as Rn – G = LH + SH, where Rn is net
radiation, G is ground heat flux, LH is latent heat flux, and SH is sensible heat flux
[Pitman, 2003]. The available energy, Qa, is the amount of energy available for sensible
and latent heat exchange with the atmosphere, i.e. Rn – G. The evaporative fraction, EF,
is the fraction of Qa that is partitioned into latent heating, i.e. LH/Qa. The surface energy
budget is coupled with the radiation budget which can be summarized as Rn = SWn +
LWn, where SWn is net shortwave radiation and LWn is net longwave radiation. Net
shortwave radiation is influenced by the land surface through albedo, , i.e. the amount
of incoming shortwave radiation that is reflected. Net longwave radiation is influenced by
the temperature of the land surface.
The presence of soil moisture is expected to increase the EF [Betts and Ball, 1998;
Eltahir, 1998]. Higher EF is expected to lower the land surface temperature and therefore
outgoing longwave radiation [Brubaker and Entekhabi, 1996; Eltahir, 1998]. Further, an
increase in ET associated with increased soil moisture should increase water vapor in the
atmosphere, increasing the longwave radiation emitted back to the land surface [Eltahir,
1998; Miller et al., 2009]. Therefore, the combined effect of higher EF and higher ET
should lead to a higher LWn. Additionally, increased soil moisture should be associated
with increased SWn because wet soils tend to have a lower albedo than dry soils
47
[Cunnington and Rowntree, 1986; Small and Kurc, 2003; Twomey et al., 1986]. Overall
then, an increased Rn is expected with increased in surface soil moisture [Eltahir, 1998].
Finally, an increased Rn is expected to increase the energy transported into the PBL by
increasing Qa [Betts, 2000; Eltahir, 1998; Quinn et al., 1995].
These couplings between the land surface and the atmosphere have been shown to be
especially strong in semiarid ecosystems [e.g. Charney, 1975]. Previous research has
demonstrated the sensitivity of semiarid ecosystems to pulses of moisture [Austin et al.,
2004; Huxman et al., 2004; Loik et al., 2004], i.e. concentrations of moisture that come in
discrete “packets”. Pulses of moisture arrive either by frequent small storms that wet only
the shallow surface layer or large but infrequent storms that wet deeper soil layers [Kurc
and Small, 2007; Sala and Lauenroth, 1985; Yaseef et al., 2010]. The layering of this soil
moisture has important but poorly understood consequences for the partitioning the
vertical fluxes of energy and water in semiarid ecosystems.
In semiarid ecosystems, Qa increases as much as 80 W m-2 [Kurc and Small, 2004] with
increasing soil moisture at the surface , a magnitude larger than changes in Qa associated
with major land surface changes such as deforestation [Gash and Nobre, 1997] or shrub
encroachment [Kurc and Small, 2004]. Evaporation is also driven by moisture near the
surface [Cavanaugh et al., 2011; Kurc and Small, 2004]. However, transpiration tends to
be driven by large storms or by soil moisture at depths greater than 20[Cavanaugh et al.,
2011; Domingo et al., 1999; Scott et al., 2006; Zeppel et al., 2008] and beyond the reach
48
of atmospheric demand. Because transpiration and photosynthesis are inextricably linked,
regardless of vegetation type, net ecosystem uptake of carbon dioxide also tends to occur
only when moisture reaches depths greater than 20 cm [Kurc and Small, 2007; Kurc and
Benton, 2010]. Further, analysis of time-lapse digital photography has demonstrated that
green-up of creosotebush (Larrea tridentata), a widespread species in the Sonora,
Chihuahua, and Mojave deserts, is also driven by moisture that reaches depths greater
than 20 cm[Kurc and Benton, 2010]. These observations argue for the importance of the
consideration of deep soil moisture in the cycling of energy and water, and therefore land
surface -atmosphere interactions, in semiarid ecosystems.
Based on the contrasting influence of shallow and deep soil moisture on surface energy
dynamics in semiarid ecosystems, we have developed a paradigm for evaluating their
relative roles in land surface-atmosphere interactions. Our objective is to evaluate how
moisture in shallow (0-20 cm) and deep (20-60 cm) soil layers influences surface energy
dynamics and planetary boundary layer characteristics. Specifically we provide
quantitative estimates of (1) the magnitude of changes in the surface energy budget
associated with the presence or absence of soil moisture in each layer and (2) the
influence of soil moisture in each layer on planetary boundary layer characteristics. This
analysis is critical in assessing the importance of the inclusion of soil moisture in multiple
soil layers in models of land surface - atmosphere interactions. Additionally, our results
point to the consideration of size-of-storm in addition to timing and frequency when
49
evaluating climate change impacts in semiarid ecosystems and their feedbacks to the
atmosphere.
2. Study Area and Methods
2.1 Study Area
Our study site is located within the Santa Rita Experimental Range (SRER),
approximately 50 km south of Tucson, Arizona, USA. This site is co-located with the
Santa Rita Creosote (US-SRC) Ameriflux eddy covariance site (http://ameriflux.ornl.gov)
in the northern portion of the SRER (UTM: 12 R 515177, 3530284) at 950 m above sea
level (Figure 1). With adherence to Ameriflux protocol, 30 min averaged CO2, H2O, and
energy fluxes are calculated using 10 Hz measurements from an open path CO2/H2O
infrared gas analyzer (LI-7500, LI-COR Inc., Lincoln, NE, USA) and a 3-D sonic
anemometer (CSAT-3, Campbell Scientific Inc., Logan, UT, USA), both at 3.75 m. Data
are stored in a CR5000 data logger (Campbell Scientific Inc., Logan, UT, USA) and
downloaded every two weeks.
Site-specific mean annual precipitation is 294 mm (calculated from a 4-year record),
more than 50% of which occurs from July to September. Small rainfall events (< 5mm)
are most frequent; however, events > 20 mm make the largest contributions to the annual
rainfall. As typical for the region, this site is characterized by a bimodal precipitation
distribution. However, winter rains (December, January, February) account only for ~20
50
% of the annual precipitation while precipitation in the months of July, August, and
September account for about 60% of the annual precipitation. We define the monsoon
season as “the shortest continuous period of the year during which 50% of the annual
precipitation accumulates”. To determine the monsoon for our study site, we analyzed
long-term precipitation data from stations within 1 km, provided by the Santa Rita
Experimental Range Digital Database (http://ag.arizona.edu/SRER/data.html). According
to these data, the monsoon season occurs during the months of July, August, and
September, hereafter referred to as the “wet season.” The “dry season” for this site is
defined as the months of May and June (4% of total precipitation). Mean annual surface
temperature is about 20°C, with monthly mean temperatures ranging from about 10°C
during the winter to about 35°C during the summer.
The physical landscape of the flux site is gently sloping (slopes < 2%), and the soils are
sandy-loam with a 10% increase of clay and silt from 35 cm to 75 cm depth. The site has
23-24% canopy cover, of which 14-15% is creosote bush (Larrea tridentata) and 8% is
grasses, forbs, and cacti unpublished data. The root distribution in the soil profile differs
between bare and canopy patches; highest densities of roots are present at 10 and 35 cm
in bare patches, while canopy patches have their highest density at 25 cm (Figure 2).
These distributions were determined by excavating six 1-m soil pits, three bare and three
canopy, and extracting 10 cm x 10 cm soil samples every 5 cm. These soil samples were
brought back to the laboratory, weighed, and then placed in a drying oven for 24 hrs at 60
°C. After they were dried, they were reweighed. After this drying, roots were collected
51
through a tiered sieving process until no further roots could be identified with the naked
eye. The collected roots were then weighed and the root density was calculated as grams
of roots per kilograms of dry soil.
2.2 Data
In this study, we present data from 2008 to 2011. In this section, we introduce how the
data were collected and the instrumentation used. We also introduce how we calculate
site-averaged values by combining measurements from bare soil and plant canopies. All
surface and radiation variables were recorded every 30 minutes. Unless otherwise
indicated, half-hour data were aggregated to a 24-hr period starting at midnight local
time. Data gaps (DOY 81-133 and 166 – 190 in 2008, DOY 154-159 in 2009, DOY 261280 in 2010, DOY 30-34 in 2011) were caused by external disturbances to equipment or
power failures.
2.2.1 Radiation and Surface Energy
Incoming (SWin) and outgoing (SWout) shortwave radiation, as well as incoming (LWin)
and outgoing (LWout) long wave radiation are measured with a four-component net
radiometer (CNR1, Kipp & Zonen, Inc., Delft, Netherlands) installed at 2.75 m above the
surface and 10 m from the eddy covariance flux tower to avoid shading from other
sensors. Net radiation (Rn) is calculated as Rn = SWn-LWn where SWn is shortwave net
radiation and LWn is longwave net radiation. Shortwave net radiation is calculated as the
52
sum of SWin and SWout. Longwave net radiation is calculated as the sum of LWin and
LWout.
Half-hour sensible and latent heat fluxes (SH and LH, respectively) were obtained for
US-SRC from the Ameriflux website (http://ameriflux.ornl.gov). These fluxes have been
corrected for an apparent flux occurring from density fluctuations [Webb et al., 1980].
Additionally, we established and used a friction velocity (u*) threshold of 0.25 m s-1
[Blanken et al., 1998].
Soil heat flux is calculated using measurements from six heat plates (HFP01SC,
Hukseflux Thermal Sensors, Elektronicaweg, The Netherlands) of which three were
installed at a depth of 5 cm on bare ground between shrub canopies and three beneath
shrub canopy sites. Six soil temperature probes (TCAV-L50, Campbell Scientific Inc.,
Logan, UT, USA) are co-located with the heat flux plates. Ground heat flux (G) is
calculated from the soil heat flux and soil temperature using a combined calorimetric heat
flux approach [Kimball et al., 1976; Kurc and Small, 2004].
Surface albedo () is calculated as the ratio of outgoing to incoming shortwave radiation:
u
(1)
53
By removing ground heat flux from the net radiation, the available energy (Qa) that could
be transferred from the land surface to the atmosphere is calculated as:
a
-
(2)
The evaporative fraction (EF) is the ratio of LH to Qa, the fraction of available energy
that is used toward latent heating, and is calculated as [Shuttleworth, 2012]
(3)
Midday averages were calculated from 30-min data for , Qa, EF, and components of the
surface energy budget. Midday averages (10:00 am to 2:00 pm, UTC/GMT-7, Mountain
Standard Time, no daylight saving) are used because we assume this is the time when
available energy is at its maximum and incoming shortwave radiation is relatively stable.
2.2.2 Soil Moisture
Six soil moisture profiles were monitored since 2008 using factory-calibrated water
content reflectometers (CS616, Campbell Scientific Inc., Logan, UT, USA) at five
different depths (2.5, 12.5, 22.5, 37.5, and 52.5), in three sites for both bare and shrub
canopy locations. Average soil moisture at each depth was weighted based on bare
ground and shrub canopy percent cover (Eq.4) [Kurc and Small, 2004; Small and Kurc,
2003]:
54
-
(4)
B
where  is volumetric soil moisture (m3m-3) in the ecosystem, f is the fraction of canopy
cover for the site (14 % for the US-SRC), C is shrub canopy soil moisture, and B is bare
ground soil moisture.
Vertical moisture was aggregated into two different layers based on reach of atmospheric
demand, where the shallow layer (0-20 cm) is within reach and the deep (20-60 cm) is
beyond the reach of atmospheric demand. Weighted averages were based on the relative
contribution of depth to the shallow (Eq. 5) or deep (Eq.6) layers of the profile:
shallow
deep
0.
0. 5
.5
.5
0.5
0.
5
.5
.5
0.
0.
.5
5
5 .5
(5)
(6)
2.2.3 Atmospheric Sounding
Atmospheric sounding data were obtained from the Department of Atmospheric Science,
University of Wyoming (http://weather.uwyo.edu/upperair/sounding.html). The sounding
data correspond to the National Weather Service surface Tucson station (KTUS, WMO:
72274) located at the Tucson International Airport (UTM: 12 R 504112, 3555012). In
this study, we analyzed the PBL characteristics using sounding data at 00 UTC
(Coordinated Universal Time).
55
Planetary boundary layer height (PBLh) was determined by analyzing potential
temperature profiles (p) [Stull, 1988]. In a mixed layer, p remains constant with height,
therefore the height of the PBL can be determined by p/z (Figure 3), p is calculated
as follows:
(7)
where p (°K) is the potential temperature, T (°K) is temperature at each level , p0 (Pa) is
pressure at sea level, p (Pa) is pressure at each level, R is assumed to be  Rd= (287 J K1
kg-1) and cpd  1004 (J K-1kg-1 ).
An air parcel that travels from the surface adiabatically will reach a saturation point, i.e.
the lifting condensation level (LCL, m) in which the temperature of the air parcel and
dewpoint are equal. We can calculate the height of displacement from:
(8)
(9)
(10)
where T0 is the initial temperature, Tdew0 is the initial dewpoint temperature,
adiabatic lapse rate, and
dew
d
is the dry
is the dewpoint lapse rate. We assume a constant dry
adiabatic lapse rate with height
d=
9.8 °C/km. Tdew0 is calculated as:
(11)
56
where Tdew0 is in °C, and e is the actual vapor pressure (see [Wallace and Hobbs, 2006]),
i.e.
(12)
where w is the mixing ratio (g/kg), = Rd/Rv= 0.622, and p is (kPa) is the pressure at the
T level. Finally,
dew
is calculated as:

where
dew
(13)
is in °C/km, g is 9.8 m s-2 and lv is the latent heat of vaporization (2.5 x106 J
kg-1), derived from Clausius – lapeyron and Poisson’s equation [Tsonis, 2007; Wallace
and Hobbs, 2006].
In addition, data available from the soundings include convective available potential
energy (CAPE) and Precipitable Water for the whole sounding (PWAT). The CAPE is a
measure of a potentially unstable atmosphere. It represents the potential energy available
in an air parcel that can be transformed to kinetic energy in a buoyant updraft, i.e. CAPE
> 2500J/kg supplies enough energy for strong updrafts and therefore thunderstorms
[Renno and Ingersoll, 1996]. PWAT is an indicator of the moisture in the atmosphere.
PWAT values above > 30 mm generally suggest that thunderstorms are likely
[Kirkpatrick et al., 2011; Means, 2012].
2.3 Soil Moisture Conceptual Framework
Soil moisture drydown dynamics differ between the shallow and deep soil, therefore two
different approaches were used to define dry and wet periods in each layer (Figure 4a).
57
For the shallow layer, soil moisture drydown curves were used to determine a moisture
threshold. To do this, we first identified large storms, i.e. precipitation events >8 mm
[Sala and Lauenroth, 1982]. Soil moisture values for the 14 days following each of the
large precipitation events were used to develop an average drydown curve (Figure 4a).
Then, an exponential model (Eq. 9) was fit to this average drydown curve:
i-
e
-
(14)
where  is volumetric soil moisture (m3 m-3), t is the time in days from the rainfall event,
i is the soil moisture the first day after the rainfall event, f is the soil moisture the last
day of the drying curve, and  is the exponential time constant [Kurc and Small, 2004;
Lohmann and Wood, 2003; Scott et al., 1997]. We identified the shallow soil moisture
layer threshold as the time when 1/3 of the moisture remains (i.e. at time ). For our site
this threshold occurred on day 4, when shallow was 0.1229 (more than two decimal places
are needed for this analysis). Therefore, the shallow layer was considered dry at moisture
values less than 0.1229.
The deep soil moisture threshold was determined by using carbon flux data from the eddy
covariance tower, specifically using the net ecosystem exchange (NEE) of CO2. In this
approach, we assume that uptake of CO2 (negative NEE) is an indicator of plant activity
and that plant activity only occurs when deep moisture is available for the plant to use
[Kurc and Small, 2007]. For this threshold analysis, we selected a 10-day window in
which NEE shifted from negative (uptake, 5 continuous days) to positive (release, 5
58
continuous days) (Figure 4a). This transition signal indicated that low soil moisture levels
in the deep layer were reducing photosynthetic activity. We assumed the fifth day of the
positive NEE (release of CO2) represented the time when soil moisture was unable to
continue to support plant activity (Figure 4a). For our four year period (2008-2011) we
extracted all 10 day windows that represented this transition and noted the soil moisture
value on the 5th day of positive NEE for each. To be conservative, we used the maximum
value of the time series generated from soil moisture values on the fifth day. Therefore,
the deep layer was considered dry at deep < 0.1013.
These threshold values (shallow=0.1229 and deep=0.1013) were used to design the
conceptual framework (Figure 3b) composed of four Cases: 1) dry shallow soil (shallow <
0.1229) and dry deep soil (deep < 0.1013); 2) wet shallow soil (shallow > 0.1229) and dry
deep soil (deep < 0.1013); 3) wet shallow soil (shallow > 0.1229) and wet deep soil (deep >
0.1013); 4) dry shallow soil (shallow < 0.1229) and wet deep soil (deep > 0.1013) (Figure
4b).
3. Results
Here we demonstrate how soil moisture present or absent in shallow or deep soil layers
influences land surface processes. We do this by analyzing 1) the surface energy budget
components and 2) analyzing the planetary boundary layer both in the context of our both
in the context of wet and dry seasons and in the context of our two-layer conceptual
framework. Overall, from our 4 year record, the dry season consisted of 244 days, while
59
the wet season consisted of 368 days. Using our two-layer soil moisture framework
categorization, 62% (795 days) of our 4-year record relates to Case 1, 1% (18 days) to
Case 2, 14% (178 days) to Case 3, and 22% (286 days) to Case 4.
3.1 Soil Moisture Influence on Radiation and Surface Energy Components
3.1.1 Wet and Dry Seasons
When calculated using midday averages, differences in all radiation and surface energy
components were statistically significant between the wet and dry seasons, with the
exception of Rn and Qa (Table 1). The similar average Rn between seasons is a result of
higher average SWn during the dry season and higher LWn during the wet season (Table
1). The average ground heat flux G is positive during the dry season while negative
during the wet season (Table 1). However, G comprises less than 4% of Rn in either
season which leads to a negligible difference between wet and dry seasons (Table 1;
Figure 5b).
While overall, average SH was lower during the wet season than the dry season (Table 1),
SH dominated the partitioning between SH and LH in both seasons. However, average
LH was significantly higher during the wet season than the dry season by about 81 Wm-2
(Table 1). Because the difference in average midday available energy Qa was negligible
between the seasons, this lower SH and higher LH led to a significantly higher average
midday EF during the wet season than the dry season (Table 1; Figure 5c).
60
Average midday surface albedo was significantly higher during the dry season (0.192)
than during the wet season (0.173) (Table 1; Figures 5a). This difference of 0.02
corresponded to about a 10% decrease in albedo under wet conditions.
3.1.2 Two-Layer Conceptual Framework
When calculated using midday averages, differences in radiation and surface energy
components were always statistically significant between conceptual framework Cases
when presence of soil moisture in the shallow layer differed, i.e. Cases 1 and 4 were
always significantly different than Cases 2 and 3 (Table 1).
Average Rn was lowest for Case 2 when moisture was present in the shallow layer but
absent in the deep layer (Table 1). This was a result of significantly lower average SWn
for Case 2 than for the other Cases (Table 1). The average ground heat flux G was
positive for Cases 1 and 4 when moisture was absent in the shallow layer but negative for
Cases 2 and 3 when moisture was present in the shallow layer (Table 1). Similarly,
average SH was significantly higher for Cases 1 and 4 when moisture was absent in the
shallow layer than for Cases 2 and 3 when moisture was present in the shallow layer
(Table 1 As for wet and dry seasons, regardless of Case, SH dominated the partitioning
between SH and LH, with the exception of Case 2 where LH is slightly higher than SH
(Table 1). Average LH was significantly higher for Cases 2 and 3 when moisture was
present in the shallow layer than for Cases 1 and 4 (Table 1).
61
Average midday available energy Qa was highest for Case 4 (499 W m-2) when moisture
was present in the deep layer but absent in the shallow layer (Table 1; Figure 5e). This is
a consequence of Rn being largest for Case 4 (Table 1). Because Case 4 is likely under
“drying” conditions, sparse cloud cover probably leads to the high SWn (Table 1).
Average midday available energy Qa was lowest for Case 2 (290 W m-2) when moisture
was present in the shallow layer but absent in the deep layer (Table 1; Figure 5e). Again,
in Case 2 Rn was lowest (Table 1). Because Case 2 is likely under short duration
“wetting” conditions from small storm before the surface has time to dry, clouds likely
lead to the low SWn (Table 1). As a consequence, differences in Qa were significant even
when similar moisture conditions were present in the shallow layer (Table 1). Despite the
differences in Qa between the Cases, Qa did not appear to be strongly influenced by
increasing soil moisture in either the shallow layer (R2 = 0.03) or the deep layer (R2 =
0.01) (Figure 6e,f).
EF was highest for Cases 2 (0.475) and 3 (0.419) when moisture was present in the
shallow layer (Table 1; Figure 5f). While EF was lowest when moisture was absent in the
shallow layer, EF was significantly higher for Case 4 than for Case 1 (Table 1, Figure
5f), presumably because the moisture in the deep layer can be used for transpiration. EF
tended to increase with increasing soil moisture in either layer, but this influence was
stronger for the shallow (R2 = 0.66) than for the deep soil layer (R2 = 0.32) (Figure 6c,d).
62
Difference in albedo from a complete dry Case (4) to wet Case (3) is about 0.01 (Table 1;
Figure 5d ); again this corresponded to about a 9% decrease in albedo under wet
conditions. Albedo for Cases with any moisture at all (Cases 2, 3, 4) were significantly
lower than for a completely dry soil profile (Case 1), even if this moisture was only
below the surface (Case 4) (Table 1; Figure 5d ). Further, albedo tended to decrease with
increasing soil moisture in either layer, but this influence was stronger for the shallow (R2
= 0.38) than for the deep soil layer (R2 = 0.15) (Figure 6a,b).
3.2 Soil Moisture Influence on the Planetary Boundary Layer
The PBLh was significantly lower when soil moisture was present in shallow layer (Cases
2 and 3) (Table 2). The PBLh extended the most when the whole soil profile was dry
(Case 1) and extended the least when the whole profile was wet (Case 3) (Table 2).
However, even when moisture was absent in the shallow layer but was present in the
deep layer (Case 4), the height of the PBL was significantly lower than when the entire
profile was dry (Case 1) (Table 2). Further, the PBLh tended to decrease with decreasing
albedo (R2=0.21) under the presence of moisture, but this influence was stronger for the
shallow (R2 = 0.32) than for the deep soil layer (R2 = 0.10) (Figure 7).
The LCL was highest when whole soil profile was dry (Case 1) and was lowest when the
shallow layer was wet (Cases 2 and 3) (Table 2). Moisture in the deep layer tended to
decrease the LCL whether or not moisture was present in the shallow layer (Cases4)
63
(Table 2). In general, the CAPE was lowest when the shallow layer was dry (Cases 1 and
4) (Table 2). The CAPE was highest when the whole profile was wet (Case 3), and is
statistically different from other cases which were not statistically different from one
another. PWAT was lowest when the entire soil profile was dry (Case 1) and highest
when the entire soil profile was wet (Case 3) (Table 2). Only in Case 3 was PWAT
greater than 30 mm (Table 2), suggesting that thunderstorms were likely only under Case
3 conditions.
4. Discussion
Our data demonstrate the importance of the presence of soil moisture on the surface
energy dynamics in a semiarid shrubland and therefore its importance in modeling local
land surface – atmosphere interactions. Consistent with previous theory [Betts and Ball,
1998; Eltahir, 1998], the presence of soil moisture resulted in an increase of EF (Table 1,
Figure 5c,f). However, our data also suggest that regardless of the presence of moisture in
the shallow layer, the presence of moisture in the deep layer also increases the EF (Figure
5f). This is most likely a consequence of the shrubs being able to access moisture in the
deep layer for transpiration [Cavanaugh et al., 2011; Zeppel et al., 2008], even when
moisture from the shallow layer is unavailable for latent heating. Also consistent with
previous theory [Eltahir, 1998; Miller et al., 2009], higher LWn was associated with the
presence of soil moisture, as much as 60 W m-2 between wet and dry seasons (Table 1).
In fact, LWn was higher when moisture was present in the deep layer but absent from the
64
shallow layer (Case 4) than when the entire profile was dry (Case 1), though this
difference was not significant.
Consistent with previous studies [Cunnington and Rowntree, 1986; Small and Kurc,
2003; Twomey et al., 1986], albedo tended to be lower under wet conditions than dry
conditions (Table 1; Figure 5d). In fact, albedo was significantly lower when moisture
was present in the deep layer but absent from the shallow layer (Case 4) than when the
entire profile was dry (Case 1). This suggests that moisture deep in the profile has an
influence on the characteristics of the land surface. We suspect that because the shrubs
have access to deep moisture through their roots, that the moisture in the deep layer is
altering the vegetation at the surface, which is altering the albedo. Supporting this notion,
a recent study showed that in a semiarid shrubland, deep moisture, beyond the reach of
atmospheric demand, is responsible for the greening of vegetation in these shrubland
ecosystems[Kurc and Benton, 2010]. Because greening of vegetation influences albedo
[Asner, 1998; Berbert and Costa, 2003; Song, 1999], the idea that deep soil moisture
influences ecosystem albedo is reasonable.
Generally speaking, despite lower albedos under wet conditions than dry conditions SWn
was higher under dry than wet conditions (Table 1), which was not expected [Small and
Kurc, 2003]. This is likely a result of the use of a two-layer soil moisture conceptual
framework to classify processes through time rather than looking for cloudy versus clear
sky days. The use of cloudy days also has implications for the values of Rn. Contrary to
65
expectations [Eltahir, 1998] , our data shows that Rn is not significantly different between
wet and dry conditions, even between an entirely dry (Case 1) and an entirely wet (Case
3) soil profile (Table 1). These negligible differences in Rn associated with soil moisture,
are reflected in similar negligible differences in Qa (Table 1). Therefore at our site, we
did not observe the expected increase in Qa associated with moisture driven increase in Rn
to increase the energy transported into the PBL [Betts, 2000; Eltahir, 1998; Quinn et al.,
1995]. Because changes in PBL characteristics are observed under different moisture
conditions at our site (Table 2), other soil moisture driven processes must be influencing
the PBL. For instance, increasing albedo [presumably by drying the soil] is associated
with increasing PBLh (Figure 7c).
Because the two-layer conceptual framework is based thresholds influenced by wetting
from precipitation pulses [Sala and Lauenroth, 1985], our data also demonstrate the
importance of storm size and the consequent temporal dynamics of the “layering” of soil
moisture associated with storm size on the surface energy components and planetary
boundary layer characteristics. To summarize, for each day within our study period, the
shrubland falls into a Case (1 to 4) (Figure 8e,j) based on the presence or absence of
moisture in a shallow and a deep layer (Figure 3b). Generally before precipitation events
both soil layers are dry (Case 1). After small precipitation events, the shallow layer is
wetted, but the deep layer remains dry (Case 2; Figure 8a,f), after a larger precipitation
event the deep layer is wetted in addition to the shallow layer (Case 3; Figure 8 e,j).
However, Case 3 is more persistent following larger storms (Figure 8a,f). Following both
66
large and small storms, the surface energy budget is modified by a decrease in albedo
(Figure 8d,i), an increase in LH (Figure 8c,h), and a decrease in SH (Figure 8c,h). These
changes correspond to a shrinking of the PBL after both small and large precipitation
events (Figure 8b,g). Importantly, the persistence of soil moisture in the deep layer
through Case 3 and Case 4 (Figure 8e,j) is closely linked to the height of the PBL (Figure
8b,g) and the components of the surface energy budget (Figure 8c,h). This suggests that
larger storms may have a greater influence on land-surface atmosphere interactions than
small storms, and that this influence is linked to the presence of deep soil moisture.
The value of continuous measurements of atmospheric and hydrological properties to
investigate land-atmosphere interactions has been previously highlighted [e.g. Baldocchi
et al., 2001; Basara and Crawford, 2002] link the importance of deep soil moisture to
surface energy dynamics by showing linear relationships between variability in soil
moisture at 20 – 60 cm depths and variability of multiple atmospheric properties. In their
study, the strength of the relationships between soil moisture and SH, LH, and EF were
always largest at depths between 20 and 60 cm depth rather than at the surface. The
results from our study support highlighting the importance of deep soil moisture in the
consideration of land surface-atmosphere interactions.
Insights from our research reflect the importance of considering site-specific nature of the
role of soil moisture in land surface atmosphere interactions, both in a shallow and deep
layer. Because the mechanism by which deep soil moisture influences surface energy
67
dynamics and planetary boundary characteristics is likely through transpiration [Yaseef et
al., 2010], understanding site-specific controls on feedbacks between vegetation and the
hydrologic cycle is critically important for land surface – atmosphere research [Chahine,
1992; Dekker et al., 2007; Scheffer et al., 2005]. For instance, root density with depth is
likely important in the determination of the depth at which soil moisture most strongly
influences surface energy dynamics. At our shrubland site, roots are concentrated at
depths > 20 cm (Figure 2), and therefore the strength of the relationship between
moisture in the deep soil layer with radiation and surface energy components is
reasonable. Furthermore, site-specific soil characteristics will also play a role in the
layering of soil moisture and in the development of root density with depth [Gregory et
al., 1987]. For instance, in desert ecosystems of the southwestern United States, a caliche
layer , “bedrock”, or a argilic horizon may inhibit root growth and soil moisture
movement beyond depths of 40 to 60 cm [Hennessy et al., 1983].
Additional sources of uncertainty must be considered in the interpretation of the surface
energy dynamics. Firstly, the temporal distribution of the Cases is subject to variability in
solar zenith angle throughout the year. Differences in solar zenith angle is likely to have a
confounding effect on the albedo values for each Case [Wang et al., 2005] as Cases
tended to be associated with particular times of year, e.g. Case 3 always occurred during
the wet season. Additionally, our two-layer framework does not discern between cloudy
or clear sky days, which may add uncertainty to our findings. Clouds influence the
incoming shortwave radiation and other surface energy components. While variations in
68
albedo are not necessarily driven by cloud cover at local scales [Small and Kurc, 2003],
at regional scales this may be an important source of uncertainty because cloud cover is
not necessarily homogeneous [van Leeuwen and Roujean, 2002].
Acknowledgments.
This research was supported in part by The University of Arizona College of Agriculture
and Life Sciences (CALS), The Arizona University System Technology and Research
Initiative Fund (TRIF), The University of Arizona Office of the Vice President for
Research (VPR), SAHRA (Sustainability of semi-Arid Hydrology and Riparian areas)
under the STC Program of the National Science Foundation (NSF), NSF CAREER
Award # EAR-1255013 and a doctoral studies fellowship provided through Mexican
National Council for Science and Technology (CONACYT).
References.
Asner, G. P. (1998), Biophysical and biochemical sources of variability in canopy
reflectance, Remote Sens. Environ., 64(3), 234-253.
Austin, A. T., L. Yahdjian, J. M. Stark, J. Belnap, A. Porporato, U. Norton, D. A.
Ravetta, and S. M. Schaeffer (2004), Water pulses and biogeochemical cycles in arid
and semiarid ecosystems, Oecologia, 141(2), 221-235.
Baldocchi, D., et al. (2001), FLUXNET: A new tool to study the temporal and spatial
variability of ecosystem-scale carbon dioxide, water vapor, and energy flux densities,
Bull. Amer. Meteorol. Soc., 82(11), 2415-2434.
69
Basara, J., and K. Crawford (2002), Linear relationships between root-zone soil moisture
and atmospheric processes in the planetary boundary layer, J. Geophys. Res.-Atmos.,
107(D15).
Berbert, M. L. C., and M. H. Costa (2003), Climate change after tropical deforestation:
Seasonal variability of surface albedo and its effects on precipitation change, J.
Clim., 16(12), 2099-2104.
Betts, A. K. (2000), Idealized model for equilibrium boundary layer over land, J.
Hydrometeorol., 1(6), 507-523.
Betts, A. K., and J. H. Ball (1998), FIFE surface climate and site-average dataset 198789, J. Atmos. Sci., 55(7), 1091-1108.
Betts, A. K., J. H. Ball, A. C. M. Beljaars, M. J. Miller, and P. A. Viterbo (1996), The
land surface-atmosphere interaction: A review based on observational and global
modeling perspectives, J. Geophys. Res.-Atmos., 101(D3), 7209-7225.
Blanken, P. D., T. A. Black, H. H. Neumann, G. Den Hartog, P. C. Yang, Z. Nesic, R.
Staebler, W. Chen, and M. D. Novak (1998), Turbulent flux measurements above
and below the overstory of a boreal aspen forest, Bound.-Layer Meteor., 89(1), 109140.
Brubaker, K. L., and D. Entekhabi (1996), Analysis of feedback mechanisms in landatmosphere interaction, Water Resour. Res., 32(5), 1343-1357.
Cavanaugh, M., S. Kurc, and R. Scott (2011), Evapotranspiration partitioning in semiarid
shrubland ecosystems: a two-site evaluation of soil moisture control on transpiration,
Ecohydrology, 4(5), 671-681.
70
Chahine, M. T. (1992), The hydrological cycle and its influence on climate, Nature,
359(6394), 373-380.
Charney, J. G. (1975), Dynamics of deserts and drought in the Sahel, Q. J. R. Meteorol.
Soc., 101(428), 193-202.
Colby, F. P. (1984), Convective inhibition as a predictor of convection during AVESESAME-II., Mon. Weather Rev., 112(11), 2239-2252.
Cunnington, W. M., and P. R. Rowntree (1986), Simulations of the Saharan atmosphere Dependence on moisture and albedo, Q. J. R. Meteorol. Soc., 112(474), 971-999.
D'Odorico, P., A. Bhattachan, K. F. Davis, S. Ravi, and C. W. Runyan (2013), Global
desertification: Drivers and feedbacks, Adv. Water Resour., 51, 326-344.
Dekker, S. C., M. Rietkerk, and M. F. P. Bierkens (2007), Coupling microscale
vegetation-soil water and macroscale vegetation-precipitation feedbacks in semiarid
ecosystems, Glob. Change Biol., 13(3), 671-678.
Domingo, F., L. Villagarcia, A. J. Brenner, and J. Puigdefabregas (1999),
Evapotranspiration model for semi-arid shrub-lands tested against data from SE
Spain, Agricultural and Forest Meteorology, 95(2), 67-84.
Dominguez, F., P. Kumar, and E. R. Vivoni (2008), Precipitation Recycling Variability
and Ecoclimatological Stability - A Study Using NARR Data. Part II: North
American Monsoon Region, J. Clim., 21(20), 5187-5203.
Eltahir, E. A. B. (1998), A soil moisture rainfall feedback mechanism 1. Theory and
observations, Water Resour. Res., 34(4), 765-776.
71
Eltahir, E. A. B., and R. L. Bras (1996), Precipitation recycling, Rev. Geophys., 34(3),
367-378.
Findell, K. L., and E. A. B. Eltahir (1997), An analysis of the soil moisture-rainfall
feedback, based on direct observations from Illinois, Water Resour. Res., 33(4), 725735.
Gash, J. H. C., and C. A. Nobre (1997), Climatic effects of Amazonian deforestation:
Some results from ABRACOS, Bull. Amer. Meteorol. Soc., 78(5), 823-830.
Gregory, P. J., J. V. Lake, and D. A. Rose (1987), Root development and structure,
Cambridge University Press.
Hennessy, J. T., R. P. Gibbens, J. M. Tromble, and M. Cardenas (1983), Water properties
of caliche, Journal of Range Management, 36(6), 723-726.
Huxman, T. E., K. A. Snyder, D. Tissue, A. J. Leffler, K. Ogle, W. T. Pockman, D. R.
Sandquist, D. L. Potts, and S. Schwinning (2004), Precipitation pulses and carbon
fluxes in semiarid and arid ecosystems, Oecologia, 141(2), 254-268.
Kimball, B. A., R. D. Jackson, R. J. Reginato, F. S. Nakayama, and S. B. Idso (1976),
Comparison of Field-measured and Calculated Soil-heat Fluxes, Soil Sci. Soc. Am. J.,
40(1), 18-25.
Kirkpatrick, C., E. W. McCaul, and C. Cohen (2011), Sensitivities of Simulated
Convective Storms to Environmental CAPE, Mon. Weather Rev., 139(11), 35143532.
72
Koster, R. D., M. J. Suarez, R. W. Higgins, and H. M. Van den Dool (2003),
Observational evidence that soil moisture variations affect precipitation, Geophys.
Res. Lett., 30(5).
Kurc, S., and E. Small (2004), Dynamics of evapotranspiration in semiarid grassland and
shrubland ecosystems during the summer monsoon season, central New Mexico,
Water Resources Research, 40(9).
Kurc, S., and E. Small (2007), Soil moisture variations and ecosystem-scale fluxes of
water and carbon in semiarid grassland and shrubland, Water Resources Research,
43(6).
Kurc, S., and L. Benton (2010), Digital image-derived greenness links deep soil moisture
to carbon uptake in a creosotebush-dominated shrubland, Journal of Arid
Environments, 74(5), 585-594.
Lohmann, D., and E. F. Wood (2003), Timescales of land surface evapotranspiration
response in the PILPS phase 2(c), Glob. Planet. Change, 38(1-2), 81-91.
Loik, M. E., D. D. Breshears, W. K. Lauenroth, and J. Belnap (2004), A multi-scale
perspective of water pulses in dryland ecosystems: climatology and ecohydrology of
the western USA, Oecologia, 141(2), 269-281.
Means, J. D. (2012), GPS Precipitable Water as a Diagnostic of the North American
Monsoon in California and Nevada, J. Clim., 26(4), 1432-1444.
Miller, R. L., A. Slingo, J. C. Barnard, and E. Kassianov (2009), Seasonal contrast in the
surface energy balance of the Sahel, J. Geophys. Res.-Atmos., 114.
73
Nicholson, S. E. (2000), Land surface processes and Sahel climate, Rev. Geophys., 38(1),
117-139.
Okin, G. S., A. J. Parsons, J. Wainwright, J. E. Herrick, B. T. Bestelmeyer, D. C. Peters,
and E. L. Fredrickson (2009), Do Changes in Connectivity Explain Desertification?,
Bioscience, 59(3), 237-244.
Pitman, A. J. (2003), The evolution of, and revolution in, land surface schemes designed
for climate models, Int. J. Climatol., 23(5), 479-510.
Quinn, P., K. Beven, and A. Culf (1995), The introduction of macroscale hydrological
complexity into land surface-atmosphere transfer models and the effect on planetary
boundary layer development, J. Hydrol., 166(3-4), 421-444.
Renno, N. O., and A. P. Ingersoll (1996), Natural convection as a heat engine: A theory
for CAPE, J. Atmos. Sci., 53(4), 572-585.
Reynolds, J. F., et al. (2007), Global desertification: Building a science for dryland
development, Science, 316(5826), 847-851.
Sala, O., and W. Lauenroth (1982), Small rainfall events: An ecological role in semiarid
regions, Oecologia, 53(3), 301-304.
Sala, O., and W. Lauenroth (1985), Root profiles and the ecological effect of light
rainshowers in arid and semiarid regions, American Midland Naturalist, 114(2), 406408.
Scheffer, M., M. Holmgren, V. Brovkin, and M. Claussen (2005), Synergy between
small- and large-scale feedbacks of vegetation on the water cycle, Glob. Change
Biol., 11(7), 1003-1012.
74
Scott, R., D. Entekhabi, R. Koster, and M. Suarez (1997), Timescales of land surface
evapotranspiration response, Journal of Climate, 10(4), 559-566.
Scott, R., T. Huxman, W. Cable, and W. Emmerich (2006), Partitioning of
evapotranspiration and its relation to carbon dioxide exchange in a Chihuahuan
Desert shrubland, Hydrological Processes, 20(15), 3227-3243.
Seneviratne, S. I., T. Corti, E. L. Davin, M. Hirschi, E. B. Jaeger, I. Lehner, B. Orlowsky,
and A. J. Teuling (2010), Investigating soil moisture-climate interactions in a
changing climate: A review, Earth-Sci. Rev., 99(3-4), 125-161.
Shukla, J., and Y. Mintz (1982), Influence of land-surface evapo-transpiration on the
Earth's climate Science, 215(4539), 1498-1501.
Shuttleworth, W. (1991), The Modellion Concept, Reviews of Geophysics, 29(4), 585606.
Shuttleworth, W. (2012), Terrestrial Hydrometeorology, edited, p. 441, Wiley-Blackwell,
UK.
Small, E. E., and S. A. Kurc (2003), Tight coupling between soil moisture and the surface
radiation budget in semiarid environments: Implications for land-atmosphere
interactions, Water Resour. Res., 39(10), 14.
Song, J. (1999), Phenological influences on the albedo of prairie grassland and crop
fields, Int. J. Biometeorol., 42(3), 153-157.
Stull, R. B. (1988), An Introduction to Boundary Layer Meteorology, edited, p. 666,
Kluwer Acad. Publ., Boston, London.
75
Taylor, C. M., E. F. Lambin, N. Stephenne, R. J. Harding, and R. L. H. Essery (2002),
The influence of land use change on climate in the Sahel, J. Clim., 15(24), 36153629.
Tsonis, A. A. (2007), An Introduction to Atmospheric Thermodynamics, edited,
Cambridge University Press, United Kingdom.
Twomey, S. A., C. F. Bohren, and J. L. Mergenthaler (1986), Reflectance and albedo
differences between wet and dry surfaces, Appl. Optics, 25(3), 431-437.
van Leeuwen, W. J. D., and J. L. Roujean (2002), Land surface albedo from the
synergistic use of polar (EPS) and geo-stationary (MSG) observing systems - An
assessment of physical uncertainties, Remote Sens. Environ., 81(2-3), 273-289.
Vereecken, H., J. A. Huisman, H. Bogena, J. Vanderborght, J. A. Vrugt, and J. W.
Hopmans (2008), On the value of soil moisture measurements in vadose zone
hydrology: A review, Water Resour. Res., 44.
Vivoni, E. R., H. A. Moreno, G. Mascaro, J. C. Rodriguez, C. J. Watts, J. GaratuzaPayan, and R. L. Scott (2008), Observed relation between evapotranspiration and soil
moisture in the North American monsoon region, Geophys. Res. Lett., 35(22).
Wallace, J. M., and P. V. Hobbs (2006), Atmospheric Science: An Introductory Survey,
edited, Academic Press, London.
Wang, Z., M. Barlage, X. B. Zeng, R. E. Dickinson, and C. B. Schaaf (2005), The solar
zenith angle dependence of desert albedo, Geophys. Res. Lett., 32(5).
76
Webb, E. K., G. I. Pearman, and R. Leuning (1980), Correction of flux measurements for
density effects due to heat and water vapour transfer, Q. J. R. Meteorol. Soc.,
106(447), 85-100.
Williams, C. A., and J. D. Albertson (2004), Soil moisture controls on canopy-scale water
and carbon fluxes in an African savanna, Water Resour. Res., 40(9), 14.
Yaseef, N. R., D. Yakir, E. Rotenberg, G. Schiller, and S. Cohen (2010), Ecohydrology
of a semi-arid forest: partitioning among water balance components and its
implications for predicted precipitation changes, Ecohydrology, 3(2), 143-154.
Zeppel, M., C. M. O. Macinnis-Ng, C. R. Ford, and D. Eamus (2008), The response of
sap flow to pulses of rain in a temperate Australian woodland, Plant and Soil, 305(12), 121-130.
77
Table 1. Midday average values of season and case analysis. Here we present net
radiation (Rn; Wm-2), net shortwave radiation (SW; Wm-2), net longwave radiation (SW;
Wm-2), ground heat flux (G; Wm-2), sensible heat flux (SH; Wm-2), latent heat flux (LH;
Wm-2), available energy (Qa, Wm-2), albedo, and evaporative fraction (EF).
n
Dry
244
Wet
368
Case 1
795
Case 2
18
Case 3
178
Case 4
286
Rn
510 A
515 A
444 b
296 c
410 b
490 a
SWn
722 A
662 B
613 a
369 c
521 b
648 a
LWn
-212 B
-147 A
-169 b
-73a
-111a
-158 b
G
18 A
-3 B
23 a
-6 c
-9 c
9b
SH
286 A
213 B
232 a
113 b
139 b
220 a
LH
23 B
105 A
32 c
116 a
105 a
72 b
Qa
526 A
508 A
467 b
290 d
400 c
499 a
Albedo
0.193 A
0.173 B
0.182 a
0.167 c
0.163 c
0.174 b
EF
0.090 B
0.325 A
0.132 c
0.475 a
0.419 a
0.249 b
p-Value <0.01, A,B Season t-test, a,b,c,d Case t-test, n=under each
category
78
Table 2. The planetary boundary layer characteristics analyzed using a two-layer soil
moisture framework. We present planetary boundary layer height (PBLh; m), lifting
condensation level (LCL; m), convective available potential energy (CAPE; J kg-1),
Precipitable Water (PWAT; mm).
n
PBLh
LCL
CAPE
Case 1
Case 2
Case 3
Case 4
32
11
32
32
mean
SE
mean
SE
mean
SE
mean
SE
2983a
79
1969c
992
1855c
787
2592b
764
399
c
136
b
154
a
4859
b
0
PWAT 6c
112
c
2797
b
2599
a
3503
b
0
104
46
239
55
61b
25
0.48
30a
4
33a
2
20b
2
p-Value <0.01,t-test a,b,c,d, n= under each case, SE standard error mean
79
List of Figures.
Figure 1. Location of study site. The Santa Rita Experimental Range (SRER) is shown in
solid gray and the white circle indicates the location of Santa Rita Creosote Ameriflux
tower. The distribution range of L. tridentata is shown for Arizona in light gray semicircles, and the dominated areas in the SRER are shown in solid black.
Figure 2. Root density (g/kg, root/soil) under bare (fill) and canopy (open). Each line
represents the average from three independent profiles, error bar shows standard
deviation.
Figure 3. The planetary boundary height is establish following the P/h [Stull, 1988],
here we present a typical profile from Case 1 to demonstrate the cutoff.
Figure 4.Our two-layer soil moisture conceptual framework. Soil moisture is influenced
by its vertical distribution whether is at reach of atmospheric demand (0-20 cm) or not
(20-60 cm). We use a drydown curve to determine the soil moisture threshold in the
shallow layer (a), and a relationship between carbon uptake and soil moisture to
determine the soil moisture threshold in the deep layer (a). Using these thresholds, all
days within the study period are categorized into Cases (b), where Case 1 represents the
dry state, Case 2 represents small precipitation events, Case 3 represents large
80
precipitation events, and Case 4 represents drying of the surface after a large precipitation
event.
Figure 5. Average diurnal cycles of albedo, available energy (Qa), and evaporative
fraction (EF), by season (a,b,c): wet (black line) and dry season (gray line) and by Case
(d,e,f): Case 1 (thick gray line), Case 2 (thin gray line), Case 3 (thick black line) and Case
4 (thin black line).
Figure 6. Linear regressions between: albedo and (a) shallow soil moisture (shallow) and
(b) deep soil moisture (deep); evaporative fraction (EF) and (c) shallow soil moisture and
(d) deep soil moisture; and available energy (Qa) and (e) shallow soil moisture and (f)
deep soil moisture.
Figure 7. Linear regressions between the planetary boundary layer height and (a) shallow
soil moisture (shallow), (b) deep soil moisture (deep), and (c) albedo.
Figure 8. Summary showing two examples of (a,f) precipitation events and how they
influence (b,g) planetary boundary layer height (PBLh), (c,h) surface energy components
sensible heat (SH) and latent heat (LH) heat, (d, i) albedo, and Case (e, j) over time as the
soil dries.
81
Figure 1. Location of study site. The Santa Rita Experimental Range (SRER) is shown in
solid gray and the white circle indicates the location of Santa Rita Creosote Ameriflux
tower. The distribution range of L. tridentata is shown for Arizona in light gray semicircles, and the dominated areas in the SRER are shown in solid black.
82
Figure 2. Root density (g/kg, root/soil) under bare (fill) and canopy (open). Each line
represents the average from three independent profiles, error bar shows standard
deviation.
83
Figure 3. The planetary boundary height is establish following the P/h [Stull, 1988],
here we present a typical profile from Case 1 to demonstrate the cutoff.
84
Figure 4.Our two-layer soil moisture conceptual framework. Soil moisture is influenced
by its vertical distribution whether is at reach of atmospheric demand (0-20 cm) or not
(20-60 cm). We use a drydown curve to determine the soil moisture threshold in the
shallow layer (a), and a relationship between carbon uptake and soil moisture to
determine the soil moisture threshold in the deep layer (a). Using these thresholds, all
days within the study period are categorized into Cases (b), where Case 1 represents the
dry state, Case 2 represents small precipitation events, Case 3 represents large
precipitation events, and Case 4 represents drying of the surface after a large precipitation
event.
85
Figure 5. Average diurnal cycles of albedo, available energy (Qa), and evaporative
fraction (EF), by season (a,b,c): wet (black line) and dry season (gray line) and by Case
(d,e,f): Case 1 (thick gray line), Case 2 (thin gray line), Case 3 (thick black line) and Case
4 (thin black line).
86
Figure 6. Linear regressions between: albedo and (a) shallow soil moisture (shallow) and
(b) deep soil moisture (deep); evaporative fraction (EF) and (c) shallow soil moisture and
(d) deep soil moisture; and available energy (Qa) and (e) shallow soil moisture and (f)
deep soil moisture.
87
Figure 7. Linear regressions between the planetary boundary layer height and (a) shallow
soil moisture (shallow), (b) deep soil moisture (deep), and (c) albedo.
88
Figure 8. Summary showing two examples of (a,f) precipitation events and how they
influence (b,g) planetary boundary layer height (PBLh), (c,h) surface energy components
sensible heat (SH) and latent heat (LH) heat, (d, i) albedo, and Case (e, j) over time as the
soil dries.
89
APPENDIX B: QUANTIFYING THE INFLUENCE OF DEEP SOIL MOISTURE ON
ECOSYSTEM ALBEDO: THE ROLE OF VEGETATION
Zulia Mayari Sanchez-Mejiaa, Shirley Anne Papugaa, and Jessica Blaine Swetisha
Submitted to Water Resources Research
a
School of Natural Resources and the Environment, University of Arizona,
Tucson, AZ 85719
90
Abstract.
As changes in precipitation dynamics continue to alter the water availability in the
ecosystem, understanding the feedbacks between the vegetation and the hydrologic cycle
and their influence on the climate system is critically important. We designed a field
campaign to examine the influence of two-layer soil moisture control on bare and canopy
albedo dynamics in a semiarid shrubland ecosystem. We conducted this campaign during
2011 and 2012 within the tower footprint of the Santa Rite Creosote Ameriflux site.
Albedo field measurements fell into one of four Cases within a two-layer soil moisture
framework based on the presence or absence of moisture in a shallow and a deep soil
layer. The average response of albedo from canopy and bare patches to soil was
quantified using these Cases. Using these averages, we explore the influence of
vegetation and soil moisture on ecosystem albedo in a gridded framework. Our results
highlight the importance of deep soil moisture in land surface – atmosphere interactions
through its influence on aboveground vegetation characteristics. For instance we show
how green-up of the vegetation is triggered by deep soil moisture, and link deep soil
moisture to a decrease in canopy albedo. Understanding relationships between vegetation
and deep soil moisture will provide important insights into feedbacks between the
hydrologic cycle and the climate system.
91
Keywords:
semiarid shrubland; greenness; eddy covariance; creosotebush; Larrea tridentata, Santa
Rita Experimental Range
92
1. Introduction
Arid and semiarid environments are susceptible to major land surface alterations such as
desertification [e.g. Carrion et al., 2010; Otterman, 1974] and shrub encroachment [e.g.
Asner et al., 2003; Grover and Musick, 1990; Van Auken, 2000] that result from climate
change and anthropogenic pressure. These changes are important as the land surface
strongly influences the dynamics of energy and water fluxes within these environments
[Gu et al., 2007]. Previous studies have shown that feedbacks between the land surface
and the atmosphere may, for instance, increase drought persistence [Charney, 1975] and
affect the planetary boundary layer development through wet or dry periods [Nicholson,
2000].
These feedbacks can be quite dynamic in space and time in arid and semiarid
environments due to the strong linkages between soil moisture and vegetation [Philippon
et al., 2005; Rodriguez-Iturbe, 2000; Rodriguez-Iturbe et al., 1999]. For instance,
moisture changes due to changes in precipitation intensity and frequency in arid and
semiarid regions are likely to affect vegetation cover [Nicholson, 2000; Thomey et al.,
2011]. Furthermore, vegetation cover, type, and phenology can exert a strong influence
on the radiation budget [e.g. Villegas et al., 2010a]. Therefore variations on land-surface
atmospheric interactions are to be expected with changes in precipitation dynamics
[Overpeck and Udall, 2010].
93
In arid and semiarid regions, soil conditions have a large effect on surface energy fluxes
and atmospheric conditions [e.g. Ek and Cuenca, 1994; Nicholson, 2000], particularly by
changing the albedo of the surface. Albedo generally decreases with increasing soil
moisture content [Idso et al., 1975; Small and Kurc, 2003; K Wang et al., 2005].
Therefore, soil moisture is a key component in controlling the variation of surface albedo
[Entekhabi et al., 1996; Fuller and Ottke, 2002; G Wang et al., 2007].
Land cover also plays a significant role in surface energy dynamics of arid and semiarid
regions through modifying albedo. Generally speaking, surfaces with scarce vegetation
cover tend to have higher albedo than more vegetated landscapes [e.g. Charney, 1975;
Nicholson et al., 1998]. The influence of land cover on albedo can vary temporally in
semiarid and arid regions which are so strongly linked to moisture inputs [Huxman et al.,
2004]. For instance, surface albedo has been shown to be lower during the wet season
due to an increase in leaf area index and greenness of the vegetation in semiarid
landscapes [Colwell, 1974; G Wang et al., 2007]. Remote sensing observations in
semiarid ecosystems have also shown a decrease in albedo during the monsoon season
that is associated with an increase in vegetation greenness [Mendez-Barroso et al., 2009;
Wu et al., 1995].
The dynamics of precipitation in semiarid regions exert a strong control over soil
moisture [Huxman et al., 2004] and vegetation dynamics [Hastings et al., 2005], and
therefore should also influence albedo. For instance, in these areas, small frequent
94
precipitation events moisten the surface layer of the soil[Sala and Lauenroth, 1982] .
Because this moisture darkens the color of the soil surface [Y Wang et al., 2011], land
surface albedo has been shown to decrease after these small storms [Duchon and Hamm,
2006; Small and Kurc, 2003]. On the other hand, large infrequent precipitation events in
these semiarid regions are capable of wetting the soil deeper below the surface [Huxman
et al., 2004; Kurc and Small, 2007; Schwinning and Sala, 2004]. This deep moisture
stimulates the vegetation [Grover and Musick, 1990; Noy-Meir, 1973], and leads to a
green up of the shrub canopy[Kurc and Benton, 2010]. Greening and thickening of
canopy foliage have been shown to decrease albedo [Xuezhen et al., 2013; Zhang and
Walsh, 2006]. Therefore, large precipitation events and therefore deep soil moisture is
likely to influence ecosystem albedo in arid and semiarid regions.
Furthermore, different plant functional types will allocate roots at different depths in the
soil to access water [Asbjornsen et al., 2011; R B Jackson et al., 2000; Loik et al., 2004].
This conceptual thinking has been established to study ecosystems in what is known as
the two-layer hypothesis [Mahrt and Pan, 1984; Walter, 1972; Wiegand et al., 2006].
Because certain plants in arid and semiarid ecosystems can access longer lasting water
from deeper soil layers [Raz-Yaseef et al., 2012], there is growing interest in knowing the
effects of this deep moisture in surface and atmospheric processes [Basara and
Crawford, 2002; Siqueira et al., 2009], especially on albedo [Sanchez-Mejia and Papuga,
in review].
95
The aim of this research is to evaluate how the presence or absence of moisture in two
different soil layers influences ecosystem albedo through modifying aboveground
characteristics of the vegetation. We demonstrate that changes in vegetation due to deep
soil moisture availability will result in changes in albedo, thereby linking deep soil
moisture and albedo. We do this by modifying a simple two-layer soil moisture
conceptual framework of four Cases: (1) a dry shallow layer (0 – 20 cm) with a dry deep
layer (20 – 60 cm), (2) a wet shallow layer with a dry deep layer, (3) a wet shallow layer
with a wet deep layer and (4) a dry shallow layer with a wet deep layer [Sanchez-Mejia
and Papuga, in review]. For the purposes of this study we evaluate each Case for a bare
patch and for a shrub (creosotebush, Larrea tridentada) patch (Figure 1a). We
hypothesize that the absence of deep soil moisture
ases
and
will result in a “dry”
and light-colored canopy, while the presence of soil moisture in the deep layer (Cases 3
and 4 will result in a “wet” and dark-colored (greener) canopy (Figure 1a). We further
hypothesize that a “dry” land surface, either through dry surface soil
ase bare, ase 4
bare) or through dry vegetation (Case 1 canopy, Case 2 canopy) should yield high albedo
values Figure b . Alternatively, a “wet” land surface, either through wet surface soil
(Case 2 bare, Case 3, bare) or through darker (greener) vegetation (Case 3 canopy, Case 4
canopy) should yield low albedo values. Using this framework, we demonstrate the role
of deep soil moisture on land surface-atmosphere interactions through its influence on
vegetation.
96
2. Study Area and Methods
2.1 Site description
The study site is located within a 250 m footprint of an eddy covariance (EC) tower
located in a creosotebush dominated ecosystem on the Santa Rita Experimental Range
(SRER), approximately 50 km south of Tucson, Arizona [Kurc and Benton, 2010;
Sanchez-Mejia and Papuga, in review]. In addition to typical water and carbon fluxes,
incident and reflected solar radiation are measured at 2.75 m above the surface and at 10
m from the EC tower with a four-component net radiometer (CNR1, Kipp & Zonen, Inc.,
Delft, Netherlands). Six soil moisture profiles are also monitored at the EC tower with
water content reflectometers (CS616, Campbell Scientific Inc., Logan, UT, USA) at
seven different depths (2.5, 12.5, 22.5, 37.5, 52.5, 67.5 and 82.5 cm) in 3 bare and 3
shrub canopy sites. Additionally, within the footprint of this EC tower are three timelapse digital cameras (pheno-cams) used to monitor plant phenology .
Long-term annual precipitation average for the closest rain gage is 260 mm (Santa Rita
Experimental Range Digital Database (http://ag.arizona.edu/SRER/data.html). According
to these data, most of the rain occurs from July, August, and September (~ 60%), while
winter rains (December, January, February) account for ~ 20%. Mean annual surface
temperature is ~20°C, with monthly mean temperatures ranging from ~10°C during the
winter to ~35°C during the summer.
97
Vegetation cover at the site is about 24%, from which 14% is creosotebush (Larrea
tridentata); while herbaceous cover and cacti account for nearly 10% [Kurc and Benton,
2010]. The topography is relatively flat (slopes < 2%), and soils are sandy loams, with a
10% increase of clay and silt, from 35 cm to 75 cm depth. Root profile shows vertical and
horizontal heterogeneity similar to what has been described in a conceptual model for
shallow-extracting woody plants [Breshears and Barnes, 1999]; root density under bare
patches is highest at 10 and at 35 cm depth, while under the canopy it is highest at 25 cm
depth [Sanchez-Mejia and Papuga, in review].
2.2 Field Campaign
Field campaigns were conducted in 2011 and 2012 to assess the influence of deep soil
moisture on albedo of bare and vegetated patches. The soil profile was viewed in two
layers where the surface (0 – 20 cm) and deep (20 – 60 cm) layers can differ in moisture
content as per Figure 1 and Sanchez-Mejia and Papuga [in review]. Over a two year
period, albedo measurements were made at eight bare and eight vegetated patches on
three different occasions in each of the Cases.
To sample in Case 1 (dry/dry), measurements were obtained during periods with no
precipitation events for at least 2 months. To sample in Case 2 (wet/dry), measurements
were taken just after a small precipitation events (<8 mm) in July, 2011 and July, 2012,
respectively so that the shallow layer will be wet but the deep layer would be dry. To
sample Case 3 (wet/wet), measurements were taken just after a large precipitation events
98
(>8 mm) in August and September, 2011 and August and September 2012 so that both
the shallow and deep layers would be wet. And finally, to sample Case 4 (dry/wet),
measurements were taken several days (~ 4) after the large precipitation events used in
Case 3. Real-time data from the nearby Sahuarita High School meteorological station
(sahuarita.cals.arizona.edu) located 13 km from the EC tower site was used to track
precipitations events. Soil moisture conditions used for the four Cases are summarized in
Figure 2.
We verified that our sampling days matched the soil moisture condition for each Case by
using the soil moisture profile data. Vertical moisture distribution was aggregated at the
surface (0-20 cm) and deep (20-60 cm) using weighted averages [Sanchez-Mejia and
Papuga, in review]. The fractions are based on the contribution of each layer to shallow
or deep moisture as below:
shallow
deep
0.
0. 5
.5
.5
0.5
0.
5
.5
.5
0.
0.
.5
5
5 .5
(1)
(2)
We then used the thresholds determined in Sanchez-Mejia and Papuga [in review] to
determine in hindsight the actual Case status for each sampling period. Generally
speaking, Case 1 was sampled during dry season of the spring before the onset of the
monsoon, Case 2 was sampled in the late spring when small rainstorms dominated the
99
precipitation, Case 3 was sampled in the summer when large rainstorms were present as a
result of the established monsoon, and Case 4 was sampled just after Case 3 but before
additional rainfall occurred (Figure 2).
2.2.1 Sampling Design
Measurements were made at two sites within the EC tower footprint, approximately 120
m apart. The first sampling site was approximately 100 m northwest of the tower and the
second 20 m south of the tower. We made measurements at four shrub canopy and four
bare soil patches at each sampling site, for a total of 8 shrub canopy patches and 8 bare
ground patches. At each patch, we measured reflected shortwave radiation using an
Eppley PSP Precision Spectral Pyranometer (The Eppley Laboratory, Inc., Newport,
Rhode Island) attached to a tripod and connected to a CR10x datalogger (Campbell
Scientific, Inc., Logan, Utah) for a two minute time period between 12:00 to 2:00 pm,
when incoming shortwave radiation is maximized. The tripod was oriented so that the
sensor was 0.3 m above the canopy of each shrub and 0.3 m above the ground level of
each bare patch. The height of the sensor ensured that the majority of the source area for
the measurement was directly from the feature of interest.
In addition to shortwave radiation measurements, we also made qualitative measurements
at each patch. Downward looking photos above canopy (Figure 3) and bare patches were
taken to follow leaf color and abundance and bare ground color. We also recorded leaf,
100
branch, and soil color, and presence or absence of blooms, flowers, seed pods, and litter
(Table 1).
2.2.2 Analysis
Albedo () was calculated as = SWout/SWin using the incoming shortwave radiation
(SWin [W m-2]) from the EC tower and the reflected shortwave radiation (SWout [W m-2])
from the field campaign measurements. Because the EC data is averaged every 30
minutes, we used the 2-minute measurement to determine a half-hour SWout and then
calculated a half hourly albedo. The eight bare measurements were pooled together and
then averaged as were the eight canopy measurements.
Field campaign albedo measurements from bare and canopy were used to generate
probability distribution functions (PDFs) describing the frequency of albedo values
around the mean for each Case (Figure 4), using a normal distribution:
(3)
where  is the mean or expected distribution,  is the standard deviation, and  2 is the
variance.
We then used these PDFs in a simple model in which each cell of a 100 x 100 grid was
populated with either a bare (0) or canopy (1) patch [e.g. Baldocchi et al., 2005]. Albedos
101
for each patch were assigned based on the PDFs. In this virtual landscape, vegetation
cover can be modified and bare and canopy albedo values are coupled to the landscape.
For our study, we increased cover from 0 to 100% and then calculated ecosystem albedo
as a function of vegetation cover for each soil moisture Case, i.e.:
(4)
where, e is ecosystem albedo, f is the percent canopy, b is bare albedo and c is canopy
albedo.
With this model we can investigate the influence of vegetation cover on albedo in the
presence or absence of moisture in a shallow or deep layer. In this exercise, we assumed
that a grid cell was either completely covered by a shrub (canopy) or completely
uncovered (bare). While this assumption is a simplification of canopy architecture and
vegetation composition, it is a starting point that can eventually be strengthened with a
dynamic vegetation model [e.g. Bonan et al., 2003; Krinner et al., 2005].
2.3 Camera-Derived Greenness
The three pheno-cams were originally placed to account for different sizes and aggregate
structure of the creosotebush community within the footprint of the EC tower. These
digital cameras (Moultrie Game Spy I-60) are mounted on a pole at 1m (lens-to-ground
distance), oriented with a field-of-view parallel to the ground surface, facing northward to
maximize sunlight and to minimize shading[Kurc and Benton, 2010].
102
Regardless of cloud cover, the daily solar noon image from each camera was analyzed for
greenness by utilizing a constant rectangular region of interest (ROI) [Richardson et al.,
2007], approximately 400 by 2800 pixels, selected to maximize creosotebush density by
integrating several individuals within each image [Kurc and Benton, 2010]. Each ROI is
then analyzed for greenness Ig using the following relationship Richardson et al. [2007]:
(5)
where Green is the average green intensity of the ROI, Red is the average red intensity,
and Blue is the average blue intensity. We then calculate a daily Ig study site by averaging
the Ig from each of the three pheno-cams . Daily Ig data is limited to 2011 due to theft of
these cameras from the field site[Kurc and Benton, 2010].
3. Results and Discussion
Here we demonstrate how soil moisture present or absent in shallow or deep soil layers
influences albedo and suggest how this is modulated by the vegetation. We do this by
analyzing 1) vegetation greenness and 2) bare and canopy patch scale albedo in the
context of our two-layer conceptual framework (Figure 1).
103
3.1 Two-Layer Soil Moisture Influence on Greenness
As hypothesized, we found the canopies of the shrubs to be the greenest (i.e. highest Ig)
when soil moisture was present in the deep layer (Cases 3 and 4) (Figure 5). Despite the
fact that creosotebush is an evergreen shrub, change in greenness is visible via the
greenness index because of changes in LAI (leaf area index) [Migliavacca et al., 2011]
associated with moisture in the soil beyond the reach of atmospheric demand that triggers
vegetation processes[Cavanaugh et al., 2011; Kurc and Small, 2007; Kurc and Benton,
2010] . While Ig in Case 4 (mean= 0.71) was not significantly different than Case 3 (mean
= 0.67), there was higher variability in Case 3 than in Case 4 that could be associated
with meteorological conditions such as cloudiness [Migliavacca et al., 2011] (Figure 5).
Because Case 4 tended to follow Case 3 (e.g. Figure 2), how moisture in the deep layer
sustains canopy characteristics at the surface [e.g. Eastham et al., 1984; Ezcurra et al.,
1992; Monson and Smith, 1982; Neufeld et al., 1988; Ong et al., 1985] must be
considered.
While moisture is available in the shallow layer in Case 2, this moisture leaves the
shallow layer within about two days[R Jackson et al., 1976; Kurc and Small, 2004],
presumably before the shrubs are able to use it[Kurc and Small, 2007] . Therefore it is not
surprising that Ig was lowest for Cases when moisture was absent from the deep layer
(Cases 1 and 2) (Figure 5). Further, Ig was lower by 0.3 when moisture was present in the
shallow layer (Case 2) than when it was not (Case 1) (Figure 5). This finding confirms
our hypothesis that the canopy is perceived as “drier” with absence of deep soil moisture,
104
and is consistent with a response mechanism to drought were evergreen desert shrubs
shed leaves [Hamerlynck and Huxman, 2009]. An alternative mechanism for the lower Ig
in Case 1 than in 2 could be a result of canopy-litter interactions [van Leeuwen and
Huete, 1996] that result from the physical removal of leaves during precipitation events
that led to the moisture in Case 2. Overall, these results support the differences between
Cases when moisture was present in the deep layer (Cases 3 and 4) and Cases when
moisture was absent in the deep layer (Cases 1 and 2) with respect to the canopy portion
of our two-layer conceptual framework (Figure 1).
3.2 Two-Layer Soil Moisture Influence on Albedo
Ecosystem albedo was calculated as an average from measurements of incoming and
outgoing shortwave radiation at the EC tower for days aligning with the field campaign.
Ecosystem albedo (Eq. 2) was highest ( = 0.205) when the entire soil profile was dry
(Case 1), and in general albedo was higher for Cases when soil moisture was absent from
the shallow layer (Cases 1 and Case 4) than when the shallow layer was wet (Cases 2 and
3) (Figure 6). Lower land surface albedo is expected when the surface soil is wet
[Fritschen, 1967; Small and Kurc, 2003; Twomey et al., 1986], so this is not surprising.
Land surface albedo is influenced by the presence of vegetation [Idso, 1972], and
therefore, we suspect that ecosystem albedo for Case 4 ( = 0.192) was lower than for
Case 1 due because the shrubs were accessing deep moisture available in Case 4 and
changing their reflectivity through greening [Kurc and Benton, 2010] .
105
We hypothesized that bare albedo values for Cases when the shallow layer was dry
(Cases 1 and 4) would be similar, and that values for Cases when the shallow layer was
wet (Cases 2 and 3) would be similar (Figure 1). However, while albedo values were
highest for Cases 1 ( = 0.249) and 4 ( = 0.220) (Figure 6b), the albedo values were
statistically different. This suggests that surface characteristics in Case 4 differ from
those in Case 1. We suspect this is through the presence of litter in Case 4 that
accumulates after rainfall events which is not as prevalent in Case 1 (Table 1).
When evaluating albedo for canopy patches, we expected that the albedo values for Cases
when the deep layer was wet (Cases 3 and 4) would be comparably lower than the albedo
values for Cases when the deep layer was dry (Cases 1 and 2) (Figure 1), indicative of the
influence of deep soil moisture on shrub canopy greenness. In fact, when evaluated for
canopy patches, albedo values were lowest and statistically similar for Case 3 ( = 0.166)
and 4 ( = 0.165) despite their differences in shallow moisture (Figure 6c). Canopy
albedo for Cases 3 and 4 was significantly different from Case 1 (= 0.193) and 2 (=
0. 8
Figure 6c . In addition, ases
and
have a “sparser” canopy than ase
and 4
(Figure 3), resulting in a complex canopy-bare soil interaction. Because of this complex
interaction, we suspect that albedo from Case 2 appears significantly lower than in Case
, mostly because of the influence of the wet soil “under” the canopy.
As expected, albedo is always highest in the absence of moisture in the shallow and deep
layer (Case 1) (Figure 6a, b, and c). Under these conditions, both the soils and the
106
vegetation tend to be “lighter” in color Figure , Figure 5, Table
and therefore more
reflective. In Case 1, albedo is higher in the bare patches (=0.249) than in the canopy
patches (=0.193) (Figure 6b, c). Ecosystem albedo calculated from tower data (=.205)
falls between these values (Figure 6a), reflecting the influence of both bare and canopy
albedo on the ecosystem.
Differences between albedo values from bare and canopy patches are largest for Cases
when the shallow layer is dry (Cases 1 and 4) (Figure 6b, c), suggesting the presence of
vegetation influences the ecosystem albedo most under conditions when the surface of
the soil is dry, despite the characteristics of the canopy (Figure 3). Albedo values from
bare and canopy patches are most similar under Case 2 (Figure 6b, c), when the shallow
layer is wet but the deep layer is dry. Under this condition, the vegetation is “dry”, light
color, with scarce leaves (Figure 3)[Kurc and Benton, 2010]. This suggests that under
these conditions the leaf area index (LAI) of the vegetation is low enough so that the wet
soil conditions pass through the canopy [van Leeuwen and Huete, 1996] and influence the
shrub albedo signal [Asner, 1998].
3.3 Influence of Vegetation on Ecosystem Albedo via Access to Two-Layer Soil
Moisture
In semiarid ecosystems, scarce vegetation results in higher albedo due to 1) the amount of
bare ground exposed [Charney, 1975; Otterman, 1974] and 2) how quickly this bare
ground loses moisture [R Jackson et al., 1976; Wythers et al., 1999]. To investigate this,
107
we used PDFs of patch scale albedo Figure 4 within the context of a simple “grid
model”, in which each cell of a 00 x 00 grid was populated with either a bare (0) or
canopy (1) patch [e.g. Baldocchi et al., 2005]. We assigned an albedo value for each
patch based on the PDFs. For our study, we increased cover from 0 to 100% for each
Case (Figure 7a), and calculated an ecosystem albedo (Equation 4) as an average of the
patches within the grid. We remind the reader that in this exercise, we assumed that a
grid cell was either completely covered by a shrub (canopy) or completely uncovered
(bare), i.e. a sparse canopy patch would not influence the grid differently than a full
canopy patch.
We expected that at 0% cover, ecosystem albedo values for Case 1 and 4 would be
similar as would ecosystem albedo values for Case 2 and 3, representing the conditions of
the bare soil patches (Figure 7a). When moisture was absent in the shallow layer but
present in the deep layer (Case 4), we expected that as cover increased ecosystem albedo
would converge toward the ecosystem albedo for Case 3 because the deep layer moisture
is accessible by the vegetation under these conditions creating “wet” and therefore “dark”
canopies (Figures 3 and 5). Therefore, at 00% cover the surface should appear “dark” in
color, regardless of the color of the soil below the canopy (Figure 7a). Further, when
moisture was present in the shallow layer but absent in the deep layer (Case 2), we
expected that as cover increased ecosystem albedo would converge toward the ecosystem
albedo for Case 1 because moisture is not accessible by the vegetation under these
conditions, creating “dry” and therefore “light” canopies Figures
and 5 . Therefore, at
108
100% cover the surface should appear “light” in color regardless of the color of the soil
beneath it (Figure 7a). Finally, we expect that the convergence of Case 4 to Case 3 should
be stronger than the convergence of Case 2 to Case 1 because canopies present under the
conditions of Cases 3 and 4 should have higher LAI because they are under less water
stressed conditions. Therefore, the canopies in Cases 3 and 4 should have less influence
from the soil beneath the canopy than Cases 1 and 2.
The biggest change in ecosystem albedo with increasing cover (~ 8% decrease) was for
Case 4 (a dry shallow layer and a wet deep layer) (Figure 7b). This result suggests that at
this site, the albedo of “wet” vegetation with high Ig (Figure 5) is very different than the
albedo of “dry” surface soil Figure b . ase
also underwent a large decrease in
ecosystem albedo with increasing cover (~5%), becoming more similar to Case 2 (Figure
7b), as expected. The smallest change in ecosystem albedo with increasing cover (~ 2%)
occurred when the shallow layer was wet but the deep layer was dry (Case 2). This
suggests that, at least at this site, the albedo of “dry” vegetation Figure
with low Ig
(Figure 5) is more similar to the albedo of wet surface soil than to the albedo of dry
surface soil.
As expected, the albedo values for Cases 3 and 4 converged with increasing cover
(Figure 7b). In fact, the ecosystem albedo for Case 4 was nearly identical to the
ecosystem albedo for Case 3 after cover was increased to 100%, supporting the
hypothesis that the canopies sufficiently cover the soil surface under these conditions
109
such that the ecosystem albedo is not influenced by it. While the albedo values of Cases 1
and 2 also began to converge when cover was increased from zero to 100%, ecosystem
albedo for Cases 1 and 2 were still quite different (Figure 7b). This difference was nearly
as large as the difference between Cases 2 and 3 at 100% cover. This supports the
hypothesis that the canopies do not sufficiently cover the soil surface under these
conditions and therefore the ecosystem albedo is still influenced by the soil surface.
Finally, according to our toy grid model, albedo values for Cases 2 and 3 should look
identical at around 30% cover (Figure 7b). Additionally, albedo values for Cases 2 and 4
should look identical at around 75% cover (Figure 7b).
3.4 Implications of Soil Moisture Drydown Differences between Bare and Canopy
Patches
Based on previous studies, we expected that the length of time that a bare and canopy
patch remained in each Case would be different and that these differences would have
implications for land surface – atmosphere modeling applications that identify bare and
canopy patches within a two-layer soil moisture framework. In particular, we expected
evaporation to be higher in bare patches [Villegas et al., 2010b] so that bare patches
would remain in Cases 2 and 3 for less time than canopy patches.
Similar to methods using in studies aimed at evaluating the characteristic timescales of
the evaporation response in land surface models [Lohmann and Wood, 2003; Scott et al.,
110
1997], we modeled the decrease in soil moisture in the surface layer and in the deep layer
as an exponential relationship through time [Hunt et al., 2002; Kurc and Small, 2004]:
(6)
where, is volumetric soil moisture [m3m-3], t is the number of days since the rainfall
event, 1 is the soil moisture observed on the first day following the rainfall event, f is
the soil moisture on the last day of the drydown, and  is a best fit exponential time
constant. We only used “large storms” > 8 mm in our analysis .
Interestingly, we found that the shallow layers of canopy patches actually dried more
quickly (Figure 8b;  = 3.15 days) than the shallow layers of bare patches (Figure 8a; =
4.6 days), by more than one full day. This suggests that contrary to our initial
expectation, bare patches would remain in Cases 2 and 3 longer than canopy patches
would. Infiltration is expected to be higher under canopy patches than under bare patches
[Bhark and Small, 2003] due in part to increased interception and stemflow [e.g.
Reynolds et al., 1999], and the movement of moisture belowground through root
macropores [e.g. Mitchell et al., 1995]. Evapotranspiration from these canopy shallow
111
layers may be large relative to bare shallow layers because of larger pore spaces in the
soil beneath the canopy or compaction of soil in the bare patches.
4. Conclusions
Our findings highlight that only considering shifts in vegetation [e.g. Beltran-Przekurat
et al., 2008] or changes in surface soil moisture [e.g. Small, 2001] may limit our
understanding of land surface-atmosphere interactions. In particular, we suggest that
understanding how deep soil moisture is expressed at the surface through the vegetation
may have implications for surface energy dynamics and planetary boundary layer
characteristics [Sanchez-Mejia and Papuga, in review]. We expect that this is especially
important in semiarid ecosystems, which are largely confined to the water-limited, rather
than energy-limited conditions that influence evapotranspiration dynamics [R Jackson et
al., 1976]in addition to other land surface-atmosphere exchanges. In these semiarid
ecosystems, moisture only moves beyond a shallow surface layer after large rainfall
events or series of events [Kurc and Small, 2007]. While precipitation is likely to
continue to decrease in semiarid areas [e.g. Woodhouse et al., 2010], whether this
reduced precipitation arrives as smaller rainfall events or more infrequent large rainfall
events will have important consequences for the vegetation [Heisler-White et al., 2008;
Huxman et al., 2004] and therefore the partitioning of energy in these ecosystems. These
changes will result in further feedbacks with the climate system [e.g. Notaro et al., 2006;
Penuelas et al., 2009].
112
Acknowledgments.
This research was supported in part by The University of Arizona College of Agriculture
and Life Sciences (CALS), The Arizona University System Technology and Research
Initiative Fund (TRIF), The University of Arizona Office of the Vice President for
Research (VPR), SAHRA (Sustainability of semi-Arid Hydrology and Riparian areas)
under the STC Program of the National Science Foundation (NSF), NSF CAREER
Award # EAR-1255013 and a doctoral studies fellowship provided through Mexican
National Council for Science and Technology (CONACYT). Patch scale shortwave
radiation measurements were made possible through field equipment provided by Dr.
Russell Scott. Additional field assistance for the campaign measurements was provided
by Lori Lovell, Evan Kipnis, Krystine Nelson, and Daniel Bunting.
References.
Asbjornsen, H., et al. (2011), Ecohydrological advances and applications in plant-water
relations research: a review, Journal of Plant Ecology, 4(1-2), 3-22.
Asner, G. P. (1998), Biophysical and biochemical sources of variability in canopy
reflectance, Remote Sens. Environ., 64(3), 234-253.
Asner, G. P., S. Archer, R. F. Hughes, R. J. Ansley, and C. A. Wessman (2003), Net
changes in regional woody vegetation cover and carbon storage in Texas Drylands,
1937-1999, Glob. Change Biol., 9(3), 316-335.
113
Baldocchi, D. D., T. Krebs, and M. Y. Leclerc (2005), "Wet/dry Daisyworld": a
conceptual tool for quantifying the spatial scaling of heterogeneous landscapes and
its impact on the subgrid variability of energy fluxes, Tellus Ser. B-Chem. Phys.
Meteorol., 57(3), 175-188.
Basara, J. B., and K. C. Crawford (2002), Linear relationships between root-zone soil
moisture and atmospheric processes in the planetary boundary layer, J. Geophys.
Res.-Atmos., 107(D15).
Beltran-Przekurat, A., R. A. Pielke, D. P. C. Peters, K. A. Snyder, and A. Rango (2008),
Modeling the effects of historical vegetation change on near-surface atmosphere in
the northern Chihuahuan Desert, J. Arid. Environ., 72(10), 1897-1910.
Bhark, E. W., and E. E. Small (2003), Association between plant canopies and the spatial
patterns of infiltration in shrubland and grassland of the Chihuahuan Desert, New
Mexico, Ecosystems, 6(2), 185-196.
Bonan, G. B., S. Levis, S. Sitch, M. Vertenstein, and K. W. Oleson (2003), A dynamic
global vegetation model for use with climate models: concepts and description of
simulated vegetation dynamics, Glob. Change Biol., 9(11), 1543-1566.
Breshears, D. D., and F. J. Barnes (1999), Interrelationships between plant functional
types and soil moisture heterogeneity for semiarid landscapes within the
grassland/forest continuum: a unified conceptual model, Landsc. Ecol., 14(5), 465478.
114
Carrion, J. S., S. Fernandez, G. Jimenez-Moreno, S. Fauquette, G. Gil-Romera, P.
Gonzalez-Samperiz, and C. Finlayson (2010), The historical origins of aridity and
vegetation degradation in southeastern Spain, J. Arid. Environ., 74(7), 731-736.
Cavanaugh, M., S. Kurc, and R. Scott (2011), Evapotranspiration partitioning in semiarid
shrubland ecosystems: a two-site evaluation of soil moisture control on transpiration,
Ecohydrology, 4(5), 671-681.
Charney, J. G. (1975), Dynamics of deserts and drought in the Sahel, Q. J. R. Meteorol.
Soc., 101(428), 193-202.
Colwell, J. (1974), Vegetation canopy reflectance, Remote Sens. Environ., 3(3), 175-183.
Duchon, C. E., and K. G. Hamm (2006), Broadband albedo observations in the southern
Great Plains, Journal of Applied Meteorology and Climatology, 45(1), 210-235.
Eastham, J., D. M. Oosterhuis, and S. Walker (1984), Leaf water and turgor potential
threshold values for leaf growth of wheat, Agron. J., 76(5), 841-847.
Ek, M., and R. H. Cuenca (1994), Variation in soil parameters - implications for
modeling surface fluxes and atmospheric boundary-layer development, Bound.Layer Meteor., 70(4), 369-383.
Entekhabi, D., I. RodriguezIturbe, and F. Castelli (1996), Mutual interaction of soil
moisture state and atmospheric processes, J. Hydrol., 184(1-2), 3-17.
Ezcurra, E., S. Arizaga, P. L. Valverde, C. Mourelle, and A. Floresmartinez (1992),
Foliole movement and canopy architecture of Larrea-tridentata (DC) Cov in
Mexican deserts, Oecologia, 92(1), 83-89.
115
Fritschen, L. J. (1967), Net and solar radiation relations over irrigated field crops, Agr
Meteorol, 4((1)), 55-62.
Fuller, D. O., and C. Ottke (2002), Land cover, rainfall and land-surface albedo in West
Africa, Clim. Change, 54(1-2), 181-204.
Grover, H. D., and H. B. Musick (1990), Shrubland encroachment in southern New
Mexico, U.S.A.: An analysis of desertification processes in the American southwest
Clim. Change, 17(2-3), 305-330.
Gu, L. H., et al. (2007), Influences of biomass heat and biochemical energy storages on
the land surface fluxes and radiative temperature, J. Geophys. Res.-Atmos., 112(D2),
11.
Hamerlynck, E. P., and T. E. Huxman (2009), Ecophysiology of two Sonoran Desert
evergreen shrubs during extreme drought, Journal of Arid Environments, 73(4-5),
582-585.
Hastings, S. J., W. C. Oechel, and A. Muhlia-Melo (2005), Diurnal, seasonal and annual
variation in the net ecosystem CO2 exchange of a desert shrub community
(Sarcocaulescent) in Baja California, Mexico, Glob. Change Biol., 11(6), 927-939.
Heisler-White, J. L., A. K. Knapp, and E. F. Kelly (2008), Increasing precipitation event
size increases aboveground net primary productivity in a semi-arid grassland,
Oecologia, 158(1), 129-140.
Hunt, J. E., F. M. Kelliher, T. M. McSeveny, and J. N. Byers (2002), Evaporation and
carbon dioxide exchange between the atmosphere and a tussock grassland during a
summer drought, Agric. For. Meteorol., 111(1), 65-82.
116
Huxman, T. E., K. A. Snyder, D. Tissue, A. J. Leffler, K. Ogle, W. T. Pockman, D. R.
Sandquist, D. L. Potts, and S. Schwinning (2004), Precipitation pulses and carbon
fluxes in semiarid and arid ecosystems, Oecologia, 141(2), 254-268.
Idso, S. B. (1972), A note on some recently proposed mechanisms of genesis of deserts,
Quarterly Journal of the Royal Meteorological Society, 103(436), 369-370.
Idso, S. B., R. D. Jackson, R. J. Reginato, B. A. Kimball, and F. S. Nakayama (1975),
The Dependence of Bare Soil Albedo on Soil Water Content, J. Appl. Meteorol.,
14(1), 109-113.
Jackson, R., S. Idso, and R. Reginato (1976), Calculation of evaporation rates during the
transition from energy-limiting to soil-limiting phases using albedo data, Water
Resources Research, 12(1), 23-26.
Jackson, R. B., J. S. Sperry, and T. E. Dawson (2000), Root water uptake and transport:
using physiological processes in global predictions, Trends Plant Sci., 5(11), 482488.
Krinner, G., N. Viovy, N. de Noblet-Ducoudre, J. Ogee, J. Polcher, P. Friedlingstein, P.
Ciais, S. Sitch, and I. C. Prentice (2005), A dynamic global vegetation model for
studies of the coupled atmosphere-biosphere system, Glob. Biogeochem. Cycle,
19(1).
Kurc, S., and E. Small (2004), Dynamics of evapotranspiration in semiarid grassland and
shrubland ecosystems during the summer monsoon season, central New Mexico,
Water Resources Research, 40(9).
117
Kurc, S., and E. Small (2007), Soil moisture variations and ecosystem-scale fluxes of
water and carbon in semiarid grassland and shrubland, Water Resources Research,
43(6).
Kurc, S., and L. Benton (2010), Digital image-derived greenness links deep soil moisture
to carbon uptake in a creosotebush-dominated shrubland, Journal of Arid
Environments, 74(5), 585-594.
Lohmann, D., and E. F. Wood (2003), Timescales of land surface evapotranspiration
response in the PILPS phase 2(c), Glob. Planet. Change, 38(1-2), 81-91.
Loik, M. E., D. D. Breshears, W. K. Lauenroth, and J. Belnap (2004), A multi-scale
perspective of water pulses in dryland ecosystems: climatology and ecohydrology of
the western USA, Oecologia, 141(2), 269-281.
Mahrt, L., and H. Pan (1984), A 2-layer model of soil hydrology, Bound.-Layer Meteor.,
29(1), 1-20.
Mendez-Barroso, L. A., E. R. Vivoni, C. J. Watts, and J. C. Rodriguez (2009), Seasonal
and interannual relations between precipitation, surface soil moisture and vegetation
dynamics in the North American monsoon region, J. Hydrol., 377(1-2), 59-70.
Migliavacca, M., et al. (2011), Using digital repeat photography and eddy covariance
data to model grassland phenology and photosynthetic CO2 uptake, Agricultural and
Forest Meteorology, 151(10), 1325-1337.
Mitchell, A. R., T. R. Ellsworth, and B. D. Meek (1995), Effect of root systems on
preferential slow in swelling soil, Commun. Soil Sci. Plant Anal., 26(15-16), 26552666.
118
Monson, R. K., and S. D. Smith (1982), Seasonal water potential components of Sonoran
Desert plants, Ecology, 63(1), 113-123.
Neufeld, H. S., F. C. Meinzer, C. S. Wisdom, M. R. Sharifi, P. W. Rundel, M. S. Neufeld,
Y. Goldring, and G. L. Cunningham (1988), Canopy architecture of Larrea
tridentata (DC) Cov, a desert shrub - Foliage orientation and direct beam radiation
interception, Oecologia, 75(1), 54-60.
Nicholson, S. E. (2000), Land surface processes and Sahel climate, Rev. Geophys., 38(1),
117-139.
Nicholson, S. E., C. J. Tucker, and M. B. Ba (1998), Desertification, drought, and surface
vegetation: An example from the West African Sahel, Bull. Amer. Meteorol. Soc.,
79(5), 815-829.
Notaro, M., Z. Liu, and J. W. Williams (2006), Observed vegetation-climate feedbacks in
the United States, J. Clim., 19(5), 763-786.
Noy-Meir, I. (1973), Desert Ecosystems: Environment and Producers, Annu. Rev. Ecol.
Syst., 4, 25-51.
Ong, C. K., C. R. Black, L. P. Simmonds, and R. A. Saffell (1985), Influence of
saturation deficit on leaf production and expansion in stands of groundnut (arachis
hypogaea l) grown without irrigation, Ann. Bot., 56(4), 523-536.
Otterman, J. (1974), Baring high-albedo soils by overgrazing - hypothesized
desertification mechanism, Science, 186(4163), 531-533.
Overpeck, J., and B. Udall (2010), Dry Times Ahead, Science, 328(5986), 1642-1643.
119
Penuelas, J., T. Rutishauser, and I. Filella (2009), Phenology Feedbacks on Climate
Change, Science, 324(5929), 887-888.
Philippon, N., E. Mougin, L. Jarlan, and P. L. Frison (2005), Analysis of the linkages
between rainfall and land surface conditions in the West African monsoon through
CMAP, ERS-WSC, and NOAA-AVHRR data, J. Geophys. Res.-Atmos., 110(D24).
Raz-Yaseef, N., D. Yakir, G. Schiller, and S. Cohen (2012), Dynamics of
evapotranspiration partitioning in a semi-arid forest as affected by temporal rainfall
patterns, Agricultural and Forest Meteorology, 157, 77-85.
Reynolds, J. F., R. A. Virginia, P. R. Kemp, A. G. de Soyza, and D. C. Tremmel (1999),
Impact of drought on desert shrubs: Effects of seasonality and degree of resource
island development, Ecological Monographs, 69(1), 69-106.
Richardson, A. D., J. P. Jenkins, B. H. Braswell, D. Y. Hollinger, S. V. Ollinger, and M.
L. Smith (2007), Use of digital webcam images to track spring green-up in a
deciduous broadleaf forest, Oecologia, 152(2), 323-334.
Rodriguez-Iturbe, I. (2000), Ecohydrology: A hydrologic perspective of climate-soilvegetation dynamics, Water Resour. Res., 36(1), 3-9.
Rodriguez-Iturbe, I., P. D'Odorico, A. Porporato, and L. Ridolfi (1999), On the spatial
and temporal links between vegetation, climate, and soil moisture, Water Resour.
Res., 35(12), 3709-3722.
Sala, O., and W. Lauenroth (1982), Small rainfall events: An ecological role in semiarid
regions, Oecologia, 53(3), 301-304.
120
Sanchez-Mejia, Z. M., and S. A. Papuga (in review), Observations of a two-layer soil
moisture influence on surface energy dynamics and planetary boundary layer
characteristics in a semiarid shrubland, Water Resour. Res.
Schwinning, S., and O. Sala (2004), Hierarchy of responses to resource pulses in and and
semi-arid ecosystems, Oecologia, 141(2), 211-220.
Scott, R., D. Entekhabi, R. Koster, and M. Suarez (1997), Timescales of land surface
evapotranspiration response, J. Clim., 10(4), 559-566.
Siqueira, M., G. Katul, and A. Porporato (2009), Soil Moisture Feedbacks on Convection
Triggers: The Role of Soil-Plant Hydrodynamics, J. Hydrometeorol., 10(1), 96-112.
Small, E. E. (2001), The influence of soil moisture anomalies on variability of the North
American monsoon system, Geophys. Res. Lett., 28(1), 139-142.
Small, E. E., and S. A. Kurc (2003), Tight coupling between soil moisture and the surface
radiation budget in semiarid environments: Implications for land-atmosphere
interactions, Water Resour. Res., 39(10), 14.
Thomey, M. L., S. L. Collins, R. Vargas, J. E. Johnson, R. F. Brown, D. O. Natvig, and
M. T. Friggens (2011), Effect of precipitation variability on net primary production
and soil respiration in a Chihuahuan Desert grassland, Glob. Change Biol., 17(4),
1505-1515.
Twomey, S. A., C. F. Bohren, and J. L. Mergenthaler (1986), Reflectance and albedo
differences between wet and dry surfaces, Appl. Optics, 25(3), 431-437.
Van Auken, O. W. (2000), Shrub invasions of North American semiarid grasslands,
Annu. Rev. Ecol. Syst., 31, 197-215.
121
van Leeuwen, W. J. D., and A. R. Huete (1996), Effects of standing litter on the
biophysical interpretation of plant canopies with spectral indices, Remote Sens.
Environ., 55(2), 123-138.
Villegas, J. C., D. D. Breshears, C. B. Zou, and P. D. Royer (2010a), Seasonally Pulsed
Heterogeneity in Microclimate: Phenology and Cover Effects along Deciduous
Grassland-Forest Continuum, Vadose Zone J., 9(3), 537-547.
Villegas, J. C., D. D. Breshears, C. B. Zou, and D. J. Law (2010b), Ecohydrological
controls of soil evaporation in deciduous drylands: How the hierarchical effects of
litter, patch and vegetation mosaic cover interact with phenology and season, J. Arid.
Environ., 74(5), 595-602.
Walter, H. (1972), Ecology of Tropical and Subtropical Vegetation, Oliver and Boyd.
Wang, G., Y. Kim, and D. Wang (2007), Quantifying the strength of soil moistureprecipitation coupling and its sensitivity to changes in surface water budget, Journal
of Hydrometeorology, 8(3), 551-570.
Wang, K., P. Wang, J. Liu, M. Sparrow, S. Haginoya, and X. Zhou (2005), Variation of
surface albedo and soil thermal parameters with soil moisture content at a semidesert site on the western Tibetan Plateau, Boundary-Layer Meteorology, 116(1),
117-129.
Wang, Y., M. Shao, Y. Zhu, and Z. Liu (2011), Impacts of land use and plant
characteristics on dried soil layers in different climatic regions on the Loess Plateau
of China, Agricultural and Forest Meteorology, 151(4), 437-448.
122
Wiegand, K., D. Saitz, and D. Ward (2006), A patch-dynamics approach to savanna
dynamics and woody plant encroachment - Insights from an arid savanna, Perspect.
Plant Ecol. Evol. Syst., 7(4), 229-242.
Woodhouse, C. A., D. M. Meko, G. M. MacDonald, D. W. Stahle, and E. R. Cooke
(2010), A 1,200-year perspective of 21st century drought in southwestern North
America, Proc. Natl. Acad. Sci. U. S. A., 107(50), 21283-21288.
Wu, A. H., Z. Q. Li, and J. Cihlar (1995), Effects of land-cover type and greenness on
advanced very high-resolution radiometer bidirectional reflectances - analysis and
removal, J. Geophys. Res.-Atmos., 100(D5), 9179-9192.
Wythers, K. R., W. K. Lauenroth, and J. M. Paruelo (1999), Bare-soil evaporation under
semiarid field conditions, Soil Sci. Soc. Am. J., 63(5), 1341-1349.
Xuezhen, Z., T. Qiuhong, Z. Jingyun, and G. Quansheng (2013), Warming/cooling
effects of cropland greenness changes during 1982-2006 in the North China Plain,
Environmental Research Letters, 8(2), 024038 (024039 pp.)-024038 (024039 pp.).
Zhang, J., and J. E. Walsh (2006), Thermodynamic and hydrological impacts of
increasing greenness in northern high latitudes, Journal of Hydrometeorology, 7(5),
1147-1163.
123
Table 1. Qualitative data generated from the field campaign. Bloom, flower, seed pod,
and litter characteristics are presented as a fraction of the number of patches with the
given characteristic, where n= 8 for both canopy and bare patches.
Canopy Patches
Leaf color
Branch color
Blooms
Flowers
Seed pods
Bare Patches
Soil color
Litter
Case 1
Case 2
Case 3
Case 4
yellow-brown
light gray
0/8
0/8
2/8
light green
dark gray
2/8
0/8
3/8
dark green
dark gray
8/8
8/8
4/8
dark green
medium gray
3/8
6/8
7/8
light brown
2/8
brown
5/8
brown
5/8
light brown
7/8
124
Figure Captions.
Figure 1. a) The four Cases of our two-layer soil moisture conceptual framework broken
into bare and canopy patches. Gray coloring represents “wet” soil and “dark” green
canopies; white represents “dry” soil and “light” green canopies. b The expected relative
magnitude of albedo associated with bare and canopy patches for each Case.
Figure 2. Time series (2011-2012) of a) daily precipitation, b) shallow (black line) and
deep (gray line) soil moisture, and c) day on which the field campaign was conducted for
each Case (Case 1 open circles, Case 2 filled circles, Case 3 filled squares, Case 4 open
squares). Soil moisture was aggregated at the shallow (0-20 cm) and deep (20-60 cm)
depths using weighted averages as described in [Sanchez-Mejia and Papuga, in review]
Figure 3. Examples of images taken during the field campaigns at each shrub where
albedo was measured for a) Case 1, b) Case 2, c) Case 3. Images for Case 4 are missing
and therefore not included here.
Figure 4. Probability distribution functions (PDFs) for albedo derived from field
campaign measurements over bare and canopy and canopy patches for each Case: a) Case
1, bare; b) Case 2, bare; c) Case 3, bare; d) Case 4, bare; e) Case 1, canopy; f) Case 2,
canopy; g) Case 3, canopy; and h) Case 4, canopy. Albedo class intervals are defined as,
125
1 (<0.12), 2(0.12-0.14), 3(0.14-0.16), 4(0.16-0.18),5(0.18-0.20), 6(0.20-0.22), 7(0.220.24), 8(0.24-0.26), 9(>0.26).
Figure 5. Mean and standard deviation of greenness (Ig) computed from daily 2011
images that were categorized using two-layer soil moisture conceptual framework. The
different letters represent statistical significance p-Value <0.01.
Figure 6. Mean and standard deviation of a) ecosystem albedo, b) bare albedo, and c)
canopy albedo for each Case. The different letters represent statistical significance pValue <0.01.
Figure 7. a) Changes in the two-layer soil moisture framework with increasing percent
canopy cover from 0 to 100% for each Case and b) Changes in albedo resulting from
increasing percent canopy cover from 0 to 100% for each Case.
Figure 8. Drydown curves for shallow soil moisture in a) bare and b) canopy patches.
The markers indicate measurements and the line is an exponential fit,  in the time
constant which represents the day at which soil moisture returns to within one third of its
pre-storm value.
126
Figure 1.
Figure 1. a) The four Cases of our two-layer soil moisture conceptual framework broken
into bare and canopy patches. Gray coloring represents “wet” soil and “dark” green
canopies; white represents “dry” soil and “light” green canopies. b The expected relative
magnitude of albedo associated with bare and canopy patches for each Case.
127
Figure 2.
Figure 2. Time series (2011-2012) of a) daily precipitation, b) shallow (black line) and
deep (gray line) soil moisture, and c) day on which the field campaign was conducted for
each Case (Case 1 open circles, Case 2 filled circles, Case 3 filled squares, Case 4 open
squares). Soil moisture was aggregated at the shallow (0-20 cm) and deep (20-60 cm)
depths using weighted averages as described in [Sanchez-Mejia and Papuga, in review]
128
Figure 3.
Figure 3. Examples of images taken during the field campaigns at each shrub where
albedo was measured for a) Case 1, b) Case 2, c) Case 3. Images for Case 4 are missing
and therefore not included here.
129
Figure 4.
Figure 4. Probability distribution functions (PDFs) for albedo derived from field
campaign measurements over bare and canopy and canopy patches for each Case: a) Case
1, bare; b) Case 2, bare; c) Case 3, bare; d) Case 4, bare; e) Case 1, canopy; f) Case 2,
canopy; g) Case 3, canopy; and h) Case 4, canopy. Albedo class intervals are defined as,
1 (<0.12), 2(0.12-0.14), 3(0.14-0.16), 4(0.16-0.18),5(0.18-0.20), 6(0.20-0.22), 7(0.220.24), 8(0.24-0.26), 9(>0.26).
130
Figure 5.
Figure 5. Mean and standard deviation of greenness (Ig) computed from daily 2011
images that were categorized using two-layer soil moisture conceptual framework. The
different letters represent statistical significance p-Value <0.01.
131
Figure 6.
Figure 6. Mean en standard deviation of a) ecosystem albedo, b) bare albedo, and c)
canopy albedo for each Case. The different letters represent statistical significance pValue <0.01.
132
Figure 7.
Figure 7. a) Changes in the two-layer soil moisture framework with increasing percent
canopy cover from 0 to 100% for each Case and b) Changes in albedo resulting from
increasing percent canopy cover from 0 to 100% for each Case.
133
Figure 8.
Figure 8. Drydown curves for shallow soil moisture in a) bare and b) canopy patches.
The markers indicate measurements and the line is an exponential fit,  in the time
constant which represents the day at which soil moisture returns to within one third of its
pre-storm value.
134
APPENDIX C: EMPIRICAL RELATIONSHIPS BETWEEN SOIL MOISTURE,
ALBEDO, AND THE PLANETARY BOUNDARY LAYER HEIGHT: A TWOLAYER BUCKET MODEL APPROACH
Zulia Mayari Sanchez-Mejiaa and Shirley A. Papugaa
For Submission Advances in Water Resources, Summer 2013
a
School of Natural Resources and the Environment, University of Arizona,
Tucson, AZ, 85719
135
Abstract.
In semiarid regions, where water resources are limited and precipitation dynamics are
changing, understanding land surface-atmosphere interactions that regulate the coupled
soil moisture-precipitation system is key for resource management and planning. We
present a modeling approach to study soil moisture and albedo controls on planetary
boundary layer height (PBLh). We used data from the Santa Rita Creosote Ameriflux site
and Tucson Airport atmospheric sounding to generate empirical relationships between
soil moisture, albedo and PBLh. Empirical relationships show that ~50% of the variation
in PBLh can be explained by soil moisture and albedo. Then, a two-layer bucket model
was used to model soil moisture, and was coupled to these empirical relationships to
model PBLh. We explored soil moisture dynamics under three different mean annual
precipitation regimes: CURRENT, INCREASE, and DECREASE, to look at the
influence on soil moisture on land surface-atmospheric processes. While our precipitation
regimes are simple, they represent future precipitation regimes that can influence the two
soil layers in our conceptual framework: an increase in annual precipitation, could have
an impact on deep soil moisture and atmospheric processes if the precipitation event is
intense, while a decrease in annual precipitation, has more of an impact on the soil
moisture at the shallow layer. We observed that the response of soil moisture, albedo, and
the PBLh will depend not only on changes in annual precipitation, but also on the
frequency and intensity of this change. We argue that because albedo and soil moisture
data are readily available at multiple temporal and spatial scales, developing empirical
136
relationships that can be used in land surface – atmosphere applications are of great
value.
Keywords
Precipitation regime, creosotebush, Larrea tridentata, eddy covariance, atmospheric
sounding, Santa Rita Experimental Range
137
1. Introduction
In his well cited scenario, [Charney, 1975] proposed a positive feedback between the
sparse desert and the dry atmosphere. Since then, for semiarid to arid regions worldwide,
coupled atmosphere-vegetation models have demonstrated how small scale land surfaceatmosphere interactions can influence climate variability at the larger scales [Claussen,
1997; Henderson-Sellers, 1993; Zeng et al., 1999]. Vegetation has also been shown to
have a strong influence on large scale hydrological processes [e.g. Ghan et al., 1997].
Ultimately, the climate system is largely dependent on the hydrologic cycle that this
vegetation influences [Chahine, 1992]. In fact, studies have suggested that small scale
positive feedbacks between vegetation and the hydrologic cycle may have the ability to
elicit non-linear responses with important large scale consequences [Dekker et al., 2007;
Rietkerk et al., 2002; Scheffer et al., 2005]. Capturing the synergies between hydrologic
processes at different space and time scales is necessary for appropriate modeling the
influence of vegetation and the hydrologic cycle on the climate system [Chahine, 1992;
Lyon et al., 2008]. These synergies are often expressed through soil moisture dynamics
[Chen and Dudhia, 2001]. Soil moisture is an important link in the hydrological cycle as
it couples precipitation and ecosystems [D'Odorico et al., 2010; Rodriguez-Iturbe, 2000;
Santanello et al., 2011; Seneviratne et al., 2010].
[Manabe, 1969] pioneered research on better understanding the effects of soil moisture
dynamics on the climate system by using a “bucket model” approach – where single root-
138
zone layer is essentially a bucket of moisture. The simplicity of this bucket model, and
others like it, make it an attractive method for looking at soil moisture control on
ecosystem processes. For instance, recent studies have focused on using the bucket model
approach for understanding ecosystem scale evapotranspiration dynamics [Guswa et al.,
2002; Porporato et al., 2004], and link this to transpiration and carbon assimilation[Daly
et al., 2004; Kurc and Small, in review] . Additionally, it has been used in combination
with “slab” models to look at the sensitivity of the planetary boundary layer development
in relation to soil moisture heterogeneity [Quinn et al., 1995]. At the larger scale, this
approach continues to be used in General Circulation Models (GCMs) [Pitman, 2003].
This simple bucket model has been shown to adequately represent soil moisture dynamics
and their influence on water cycling in ecosystems where precipitation events are large
and sufficiently frequent enough to wet the entire root zone [Salvucci, 2001]. However, in
arid to semiarid ecosystems, water and carbon cycles have been shown to be controlled
by soil moisture availability two discrete layers [Breshears et al., 1997; Kurc and Small,
in review]. Small precipitation events, which only wet the top surface layer [Sala and
Lauenroth, 1982], trigger evaporation-dominated evapotranspiration [Kurc and Small,
2004; Scott et al., 2006]. Large precipitation events provide soil moisture to the deep
layer [Kurc and Small, 2007; Raz-Yaseef et al., 2012] that is used toward transpiration
[Cavanaugh et al., 2011] and carbon uptake [Kurc and Benton, 2010; Scott et al., 2006].
Further, both shallow and deep soil moisture have been shown to have an effect on the
surface energy budget components, albedo, and planetary boundary characteristics
139
[Basara and Crawford, 2002; Sanchez-Mejia and Papuga, in review; Sanchez-Mejia et
al., in review; Small and Kurc, 2003]. Therefore, modifying the bucket model to include
two-layers will be important for representing soil moisture dynamics of arid to semiarid
areas in land surface – atmosphere applications.
The characteristics of the planetary boundary layer (PBL) reflect the effects of the surface
energy dynamics of the land surface, which play a large role in the development of the
climate system. Therefore, knowledge of the PBL height, and how it changes, is
important for understanding the influence of the land surface on the climate system.
Previous efforts to obtain PBL characteristics have tended to rely on in situ measurements
[e.g. Betts and Ball, 1994] or complex models [e.g. Diak and Whipple, 1993], which are
not ideal for applications that need daily estimates of PBL characteristics [Santanello et
al., 2005] . However, because of the strong relationships between the PBL height and
land surface characteristics such as soil moisture [Basara and Crawford, 2002] and
albedo [Sanchez-Mejia and Papuga, in review] that are generally available daily through
micrometeorological datasets or through remote sensing products, empirical relationships
that relate PBL height to land surface characteristics show promise [Santanello et al.,
2005] for improving our understanding of PBL dynamics and its feedback to the climate
system.
Given the sensitivity of the PBL to both shallow and deep soil moisture [Sanchez-Mejia
and Papuga, in review], the objectives of this study are to 1) to develop empirical
methods of estimating daily PBL height based on readily attainable data that has been
140
shown to influence the PBL, such as albedo and surface soil moisture; and 2) link the
two-layer bucket model to the empirical relationships to predict changes in PBL
dynamics under different rainfall regimes that represent different climate change
scenarios.
2. Methods
2.1 Micrometeorological, Soil Moisture and Radiation Data
Field observations used in this study were from the Santa Rita Creosote Ameriflux site
(US-SRC; http://ameriflux.ornl.gov) in a creosotebush (Larrea tridentata)-dominated
shrubland located near the northern-most boundary of the Santa Rita Experimental Range
(SRER) in southern Arizona. Annual precipitation at the site is bimodal and averages ~
350 mm yr-1 (http://ag.arizona.edu/SRER/data.html), ~ 55% occurring in summer (July,
August, September) and 15% occurring in winter (December, January, February). The
mean annual surface temperature is ~20°C, with monthly mean temperatures ranging
from ~10 °C during the winter to ~35°C during the summer. Soil is sandy-loam with a
10% increase of clay and silt from 35 cm to 75 cm depth. Highest densities of roots are
present at 10 and 35 cm in bare ground patches, while creosote bush canopy patches have
their highest density at 25 cm [Sanchez-Mejia and Papuga, in review].
Observations included four years of measurements (2008 – 2012) from an eddy
covariance tower with typical micrometeorological instrumentation [Moncrieff et al.,
141
2000; Shuttleworth, 1993]. Precipitation was recorded using a tipping-bucket rain gauge.
Volumetric water content was measured using Campbell Scientific factory-calibrated
water content reflectometers at depths of 2.5, 12.5, 22.5, 37.5 and 52.5 cm replicated in
six profiles, three under shrub canopies and three in bare soil patches. Incoming (SWin)
and outgoing (SWout) shortwave radiation, as well as incoming (LWin) and outgoing
(LWout) long wave radiation were measured with a four-component net radiometer
(CNR1, Kipp & Zonen, Inc., Delft, Netherlands).
Average soil moisture at each depth was calculated using a weighting based on bare
ground and shrub canopy percent cover [Kurc and Small, 2004; Small and Kurc, 2003],
i.e.:
-
(1)
B
where  is volumetric soil moisture (m3m-3), f is the fraction of shrub canopy cover (14 %
for the US-SRC), C is shrub canopy soil moisture, and B is bare soil moisture. Average
soil moisture values were then aggregated for two layers: shallow (0-20 cm) and deep
(20-60 cm). Aggregations were calculated as weighted averages based on the contribution
of each depth [Sanchez-Mejia and Papuga, in review]:
shallow
0.
.5
0.5
.5
0.
.5
(2)
142
deep
0. 5
.5
0.
5
.5
0.
5
5 .5
(3)
Soil moisture was then categorized into four cases based on the presence or absence of
moisture in the upper “shallow” layer and the lower “deep” layer [Sanchez-Mejia and
Papuga, in review]: Case 1 corresponds to a dry shallow and dry deep layer, Case 2
corresponds to a wet shallow and dry deep layer, Case 3 corresponds to a wet shallow
and wet deep layer, and Case 4 corresponds to a dry shallow and wet deep layer. The
presence or absence of moisture in each layer was determined based on soil moisture
thresholds derived based on soil moisture observations at the site [Sanchez-Mejia and
Papuga, in review]. Soil moisture thresholds for our site were identified for the
ecosystem based on percent shrub cover using 2009 – 2011 data (Equation 1), and also
for bare and shrub canopy patches using 2009 – 2012 data and are presented in Table 1.
For each day, these thresholds are then used to categorize soil moisture into one of the
four Cases.
Half-hourly surface albedo () was calculated as the ratio of outgoing to incoming
shortwave radiation, i.e.:
out
in
(4)
For albedo, midday averages, as opposed to daily averages, were calculated from 30-min
data. We used midday averages (10:00 am to 2:00 pm, UTC/GMT-7, Mountain Standard
143
Time, no daylight savings) because we assume that this is the time when available energy
is at its maximum and incoming shortwave radiation is relatively stable.
2.2 Atmospheric Sounding Data
Atmospheric sounding data were obtained from the Department of Atmospheric Science,
University of Wyoming (http://weather.uwyo.edu/upperair/sounding.html). The sounding
data correspond to the National Weather Service surface Tucson station (KTUS, WMO:
72274) located at the Tucson International Airport (UTM: 12 R 504112, 3555012). In
this study, we analyzed the PBL characteristics using sounding data at 00 UTC
(Coordinated Universal Time).
Planetary boundary layer height (PBLh) was determined by analyzing potential
temperature (P) profiles[Stull, 1988]. In a mixed layer, P remains constant with height,
therefore the height of the PBL can be determined by P/z, where P is calculated as
follows:
(5)
where P (°K) is the potential temperature, T (°K) is temperature at each level , p0 (Pa) is
pressure at sea level, p (Pa) is pressure at each level, R is assumed to be  Rd= (287 J K1
kg-1) and cpd  1004 (J K-1kg-1 ).
144
Atmospheric stability (ML; oK m-1) in the mixed layer relates to the change of P with
height; the thermal PBL development is influenced by this parameter[Santanello et al.,
2005]. We calculated stability as:
(6)
where P is potential temperature (°K), h is height (m), t refers to the top of PBL and s to
the surface just above the inversion zone[Santanello et al., 2005].
2.3 Development of Empirical Relationships
Data for the development of empirical relationships were selected based on soil moisture
dynamics defined by the Cases (described in Section 2.1). From 2008 to 2011, we
identified the days corresponding to each Case and selected representative days for each
Case based on surface and sounding data available (Case1=30, Case2=7, Case3=32, Case
4=32) for a total of 101 days. Case 2 occurred infrequently, and therefore comparisons
must be made cautiously in this analysis. For these 101 days we obtained atmospheric
soundings (Section 2.2), shallow and deep soil moisture (Section 2.1), and albedo
(Section 2.1). We used these data to develop empirical relationships between the land
surface and the PBLh using a polynomial approach.
The PBLh has been shown to be controlled by land surface parameters such as soil
moisture, atmospheric stability, evaporative fraction, and 2-m potential temperature
145
[Basara and Crawford, 2002; Santanello et al., 2005; Santanello et al., 2011; Santanello
et al., 2013]. Additionally, albedo has shown to have an important role as well in land
surface – atmosphere interactions [Jackson et al., 1976; Sanchez-Mejia and Papuga, in
review]. [Santanello et al., 2005]developed an empirical relationship between the PBLh
and soil moisture and stability for their site in the southern Great Plains using a 2nd order
polynomial:




(7)
where p0…p5 are the polynomial coefficients. Using atmospheric stability (Eq. 6) and
shallow soil moisture (Eq. 2), we adapted this empirical relationship for our site using all
of days of data from 2008 – 2011 (Figure 1) and also by categorizing the data by Case as
described above (Figure 2). We report our site-specific coefficients in Table 2.
Because previous studies have highlighted the sensitivity of PBLh to albedo and soil
moisture, including deep soil moisture, [Basara and Crawford, 2002; Sanchez-Mejia and
Papuga, in review] and albedo is a more readily available parameter than PBLh stability,
we also develop an empirical relationship between the PBLh and shallow soil moisture
(Eq. 2) and albedo (Eq. 4) using the 2nd order polynomial approach (Figure 3), i.e.:
(8)
146
We also develop an empirical relationship between the PBLh and shallow soil moisture
(Eq. 2) and deep soil moisture (Eq. 3) using the 2nd order polynomial approach (Figure 3),
i.e.:
(9)
These relationships are developed using all of 101 days of Case data from 2008 – 2011;
we report the polynomial coefficients in Table 2. Lastly, we develop a 2nd order
polynomial empirical relationship using both shallow (Eq. 2) and deep soil moisture (Eq.
3) with albedo (Eq. 4):
(10)
and report the polynomial coefficients in Table 2.
2.4 Two-Layer Bucket Model of Soil Moisture Dynamics
To model shallow and deep soil moisture dynamics, we modified the standard root zone
bucket model [Daly et al., 2004; Guswa et al., 2002; Laio et al., 2001; Rodriguez-Iturbe
et al., 1999] into a two-layer bucket model (Figure 4) [Kurc and Small, in review]. In this
two-layer bucket model, the root zone is divided into an upper shallow layer where
evaporation drives soil moisture, and a lower deep layer where transpiration controls soil
moisture [Kurc and Small, in review]. Leakage can occur from the upper to the lower
147
layer, but no redistribution is allowed from the lower to the upper layer [Kurc and Small,
in review]. The following equations describe this two-layer bucket model:
(11)
(12)
where, n is porosity, ZE is the depth of the upper layer, ZT is the depth of the lower layer,
sE is the relative soil moisture content in the upper layer, sT is the relative soil moisture
content in the lower layer I(sE,t) is the rate of infiltration, E(sE) is the rate of evaporation,
T(sT) is the rate of transpiration, L(sE) is the rate of leakage from the upper to the lower
layer, and L(sT) is the rate of leakage from the lower layer out of the system. Parameters
used for our specific study site are listed in Table 3.
As for evapotranspiration in the standard bucket model, losses from the system due to
evaporation E(sE) and transpiration T(sT) are described as piecewise functions of sE and
sT, where sE and sT are constant above a certain s* and decrease linearly to the
hygroscopic point sh in the case of E [Guswa et al., 2002]and to the wilting point sw in the
case of T [Guswa et al., 2002; Laio et al., 2001], i.e.:
148
(13)
(14)
where Emax and Tmax are the maximum daily rates under wet conditions [Guswa et al.,
2002]. Parameters used for our specific study site are listed in Table 3.
Previous studies in similar ecosystems have shown that leakage below the root zone (60 –
80 cm) is rare [Liu et al., 1995], or suggest that the lower layer never gets enough wet to
leak [Kurc and Small, in review]. However, because we consider increased precipitation
in our analysis, we include a leakage term from the lower layer in our model. Movement
out of the layers is only considered to be downward in our model. Two components
comprise this movement: (1) behavior of hydraulic conductivity and (2) a simple
empirically derived linear relationship regulating water movement out of the layer [Kurc
and Small, in review], i.e. 1% of the soil moisture in the upper layer is moved into the
lower layer and 0.1% of the soil moisture in the lower later is moved out of the bottom of
the system at any time step. We ignore hydraulic redistribution [e.g. Ryel et al., 2004;
Scott et al., 2008] between the layers mostly due to lack of information of this
mechanism for our study site. As such, leakage is modeled as a simple linear relationship
[Daly et al., 2004; Kurc and Small, in review]:
149
(15)
(16)
where, Ksat is saturated hydraulic conductivity in cm d-1 and b is the exponent of the
retention curve
where
is the soil water potential at saturation. Parameters
used for our specific study site are listed in Table 3.
2.4.1 Model Parameters
The depth of the upper layer ZE is set as 0 – 20 cm to minimize the influence of
atmospheric demand on the lower layer (Table 3). A root density profile has been
developed for this shrubland site and is reported in [Sanchez-Mejia and Papuga, in
review]. As was found for a similar shrubland[Kurc and Small, 2007] , while roots were
found to depths of 1 m in both the bare and canopy areas of our shrubland, the highest
root densities were found above 60 cm. Therefore, a maximum depth of 60 cm, i.e. ZT =
20 – 60 cm, was used in our modeled scenarios.
Soil texture at our study site classified as a sandy loam [Kurc and Benton, 2010], and
therefore we assigned a porosity (n) of 0.43 (Table 3). [Laio et al., 2001] suggest values
for the parameters sfc,sh, sw, and s* are based on the sandy loam soil texture using soil
water potential at different pressures using soil moisture retention curves . However, the
150
model is very sensitive to these parameters[Kurc and Small, in review] , and therefore
while we adopt the suggested value of 0.46 [Table 3; Laio et al., 2001], we recognize that
the wilting and hygroscopic point at our site are lower than what they report, i.e. at our
shrubland sh of 0.075 and sw of 0.085 are more appropriate (Table 3). These parameters
depend on both vegetation and soil properties, and wilting point values have been shown
to be defined as low as 0.03 [Guswa et al., 2002].
Partitioning evapotranspiration into its components of transpiration and evaporation is a
challenging enterprise, and has not been fully conducted at our shrubland site. Because
we do not have site-specific partitioning measurements, we defined the model parameters
of maximum evaporation from the upper layer and maximum transpiration from the deep
layer based on research conducted similar semiarid ecosystems [Kurc and Small, 2004;
Moran et al., 2009; Reynolds et al., 2000; Villegas et al., 2010]. Using these studies, we
define an individual maximum value for evaporation and transpiration at an ecosystem,
bare, and canopy level (Table 3).
2.4.2 Model Precipitation Input
The water from precipitation enters the upper layer of the model as:
(17)
where i corresponds to the time step and PPT is precipitation.
151
We calibrated the two-layer model using three years (2009 – 2011) of daily precipitation
from the micrometeorological tower as input and compared observed soil moisture in the
shallow and deep layers to modeled soil moisture (Figure 5). Linear regression between
observations and model output suggest that shallow moisture dynamics (R2= 0.60) are
better captured than deep moisture dynamics (R2= 0.12). In general, the modeled shallow
soil moisture follows the peaks and drydowns of the observed data. While deep soil
moisture follows the trends of the observed data, the magnitude does not reflect the
observations. We note that the model initialization results in a lag at the beginning of the
modeling time series which is resolved as however as moisture accumulates. In addition
we note that modeling the higher and lower values of deep soil moisture is complex,
despite the fact that the model follows the trend of the observed deep soil moisture. This
is a consequence for the development of Cases and the use of our Case thresholds (Table
1). When categorizing our model data by Case, it is clear that deep soil moisture is
misrepresented. For instance, Case 1 days decrease from 608 (observed) to 587 (model),
while Case 3 days increase from144 (observed) to 216 (model), similarly an increase is
observed in Case 4 (Table 4). When partitioning the ecosystem into bare and canopy to
analyze model soil moisture versus observed, the model overestimates the deep soil
moisture under bare (Table 4, Case 4) and underestimates deep soil moisture under
canopy (Table 4, Case 2 and 4).
To develop soil moisture scenarios for our Cases based on shift in the precipitation
regime, we model precipitation input as a stochastic process [Daly et al., 2004;
152
Rodriguez-Iturbe et al., 1999]. At a scale where soil-moisture recycling can be neglected,
the rainfall input can be generated as an external force independent from the soil moisture
state [Laio et al., 2001]. The response of vegetation and therefore land surface- PBL
interactions in a water-limited ecosystem is tightly coupled to the stochastic nature of soil
moisture given by precipitation [Katul et al., 2007]. Because of this, rainfall follows a
series of “events” in time with frequency ) and depth () following a typical Poisson
process [Bonser et al., 1985], and therefore the frequency in time is found using a typical
Poisson distribution:


(18)
where, Px(x) is the probability of observing x events in a time period and  is equal to the
average rate of occurrence of the event.
The depth () of the precipitation event can be calculated by using an exponential
function based on the average rainfall amount [Laio et al., 2001; Porporato et al., 2002].
This methodology to simulate daily precipitation regime is recognized to be a common
practice [Laio et al., 2001]. Frequency at the site was calculated from daily precipitation
data from 2008 to 2012, where the amount of days between storms was averaged,
resulting in 10 days average between storms. However, during a dry period a maximum
of 80 days without storms can be observed, while during the monsoon period a single day
between storms can be observed. Based on possible future changes in precipitation
153
[NAST, 2000; Trenberth et al., 2003; Weltzin et al., 2003], we assume the average 10 day
precipitation frequency and alter the average annual intensity from 8.3 mm with a 20%
increase and a 20% decrease.
3. Results and Discussion
3.1 Soil moisture and albedo influence on the planetary boundary layer height
Our analysis takes advantage of a two-layer soil moisture conceptual framework. Using
this framework, we categorize our time series data into four soil moisture cases based on
the presence or absence of moisture in the upper “shallow” layer and the lower “deep”
layer [Sanchez-Mejia and Papuga, in review]. Using representative days from each of
these categories (Figure 6a), we can see that the PBLh is sensitive to the soil moisture
derived Cases. Overall, the PBLh is largest when the shallow layer is dry (Cases 1 and 4)
and smallest when the shallow layer is wet (Cases 2 and 3). As has been demonstrated in
recent studies [Sanchez-Mejia and Papuga, in review], moisture presence in the deep
layer does tend to shrink the PBLh, for Case 3 the PBLh is generally lower than for Case 2
and for Case 4 the PBLh is generally lower than for Case 1 (Figure 6a). Obviously,
shallow soil moisture is highest for Case 2 and Case 3– but it is noticeably higher under
Case 3 (Figure 6b), which likely is influencing the increased shrinkage of the PBLh under
that moisture condition. Albedo is lowest in Case 2 and Case 3 (Figure 6c), which
supports the concept that surface soil moisture is influences the albedo [Small and Kurc,
154
2003] which is influencing the PBLh [Jackson et al., 1976; Sanchez-Mejia and Papuga, in
review].
We note that the scale at which soil moisture and albedo is measured in comparison to an
atmospheric sounding is a source of uncertainty in our analysis. For instance, spatial
variability in soil moisture associated with terrain can introduce complexity during moist
periods at multiple scales [Western et al., 1999], having an influence on surface fluxes
and ultimately the PBL [Giorgi and Avissar, 1997].
3.2 Empirical relationships for estimating planetary boundary layer height
[Santanello et al., 2005] argued for the need to develop empirical relationships that
focused on observable properties of the PBL and the land surface rather than model of
complex processes. Further, they emphasize using daily variability in these properties,
using the PBL as an integrator of surface conditions at the regional scale. A statistical
analysis found that the PBL was most sensitive to atmospheric stability ML, and that soil
moisture  also played a significant role . Therefore, they developed an empirical
relationship to estimate PBLh from daily ML and  (Equation 7). Because the empirical
relationship only captured about 60% of the variance between the variables, the predicted
range of PBLh was smaller than the actual range, which did have consequences when
using those PBLh estimates in other applications . We do note that the atmospheric
sounding from which PBLh is estimated has a larger source area than the , therefore a
difference between the actual and predicted PBLh is not necessarily unexpected.
155
Based on their findings, we calculated the ML for our site (Equation 6) and defined our
coefficients (Table 2) for their empirical relationship for PBLh using ML and shallow
(Equation 7, Figure 1). Using this approach, the variables explained about 50% of the
variance in PBLh; the estimated PBLh values fall between 298 m to about 4600 m (Figure
1), whereas the actual PBLh values have a range from about 300 m to about 4600 m with
a majority of those between 1500 – 2500m (Figure 6a). When we look at these by Case
(Figure 2), we see that the estimations for PBLh improve for the Cases when shallow soil
moisture is present, i.e. Case 3 (R2 = 0.51) and Case 2 (R2 = 0.97). Additionally, the
differences in ranges between the actual PBLh values (Figure 6a) and the estimated PBLh
values (Figure 1) almost disappear for these Cases.
Despite reasonable predictions of PBLh using ML and , we calculated ML using
atmospheric sounding data and therefore we presume that this parameter is not any easier
to access than the PBLh itself. Strong relationships between PBLh and albedo and PBLh
and soil moisture in both the shallow and deep layer have been previously identified
[Figure 6; Basara and Crawford, 2002; Sanchez-Mejia and Papuga, in review]. There is
value in having identified these relationships in part because observations of albedo and
soil moisture are readily available at multiple temporal and spatial scales [Asbjornsen et
al., 2011; Wang et al., 2012] – ranging from micrometeorological datasets [Baldocchi et
al., 2001; Mueller et al., 2011] to remotely sensed data [Mills, 1990; Nemani et al.,
156
1993]. Furthermore, soil moisture has been predicted from albedo [Jackson et al., 1976]
and vice versa [Sanchez-Mejia and Papuga, in review].
To explore the ability to predict PBLh using these readily available parameters, we
adopted the polynomial approach of [Santanello et al., 2005]. We developed an empirical
relationship to estimate PBLh from daily  and shallow (Equation 8), from daily shallow and
deep (Equation 9), and from daily , shallow and deep (Equation 10). Our analysis shows
that  and shallow explained about 40% of the variance in PBLh (Figure 7a). From the
surface contours, relationships between the variables and the PBLh are highlighted, e.g. at
high  and shallow, the PBLh develops more (Figure 7a). This is expected because under
these conditions, energy is partitioned towards turbulence (sensible heat) [Otterman et
al., 1993]. Estimating PBLh with shallow and deep did not show improvement (R2 = 0.40;
Figure 7b), while using , shallow, and deep provided the best estimates (R2 = 0.42; Figure
7c). The addition of deep moisture does not necessarily improve the predictability of the
PBLh, in our analysis. We suspect this is in part due to the lack of the ability of the model
to adequately capture deep moisture dynamics. That being said, deep moisture data are
not as readily available as shallow soil moisture, e.g. remotely sensed soil moisture
generally only captures the near surface signal [e.g. Calvet and Noilhan, 2000].
Therefore, based on the availability of the variables and their influence on the PBLh, we
suggest that the empirical relationship predicting PBLh from  and shallow (Equation 8) is
a reasonable approach.
157
It should be noted again that the consideration of scale will be important in the derivation
and use of these empirical relationships. This is because atmospheric soundings represent
a scale on the order of 10 km2, while flux towers generally represent a scales on the order
of 100 m2 [Asbjornsen et al., 2011]. How much these empirical models represent the land
surface that is linked to the PBL, will depend on surface heterogeneity, and the response
of the PBL to integrated fluxes over a larger spatial scale.
3.3 Influence of change in precipitation regime on land surface - PBL dynamics
With a combination approach using the modified two-layer bucket model of soil moisture
dynamics and the empirical relationships developed in this study, we estimate the PBLh
for three different modeled precipitation regimes (Equation 18), two based on possible
future changes in precipitation [NAST, 2000; Trenberth et al., 2003; Weltzin et al., 2003]:
(1) the current 10 day precipitation frequency with an average annual intensity of 8.3
mm, hereafter referred to as CURRENT; (2) the current 10 day precipitation frequency
with a 20% increase in average annual intensity, hereafter referred to as INCREASE; and
3) the current 10 day precipitation frequency with a 20% decrease in average annual
intensity, hereafter referred to as DECREASE. Using these precipitation regimes as
inputs into the two-layer bucket model, we obtain a daily soil moisture value for both the
shallow upper (Equation 11) and the deep lower layer (Equation 12) and calculate this for
both bare and canopy patches. Using the percent cover, we calculate ecosystem soil
moisture as a weighted average (Equation 1). Using modeled ecosystem shallow soil
moisture, we estimate albedo using a linear relationship [Sanchez-Mejia and Papuga, in
158
review]. Then, we use our empirical relationships (Equations 8, 9 and 10) to estimate the
PBLh.
Model results from the CURRENT precipitation regime (Figure 8) at the study site yield
total precipitation of 293 mm, which is consistent with our site characteristics. Shallow
ecosystem soil moisture ranges from 6.7% to 17.5% and generally following the
dynamics of the precipitation, increasing after every storm. Deep ecosystem soil moisture
ranges from 8.5% to 10.8%, only increasing after large storms. Albedo ranges from 15 %
to 19%, with a tendency toward an overall decrease following large storms. The PBLh is
variable depending on the empirical relationship: for f(,sh) PBLh ranges from 1567 m to
2919 m, for f(sh,d) PBLh ranges from 1353 m to 3130 m, and for (sh,d,) PBLh ranges
from 1725 m to 3202 m.
Model results from the INCREASE precipitation regime (Figure 9) at the study site yield
total precipitation of 433 mm. Shallow ecosystem soil moisture ranges from 6.9 % to
27.9 %. Deep ecosystem soil moisture ranges from 8.5 % to 11.6 %, only increasing after
large storms. Albedo ranges from 11.5% to 18.4%, which is just slightly lower than the
CURRENT scenario. The PBLh is variable depending on the empirical relationship: for
f(sh,) PBLh ranges from 1001 m to 2892 m, for f(sh,d) PBLh ranges from 1237 m to
3193 m, and for (sh,d,) PBLh ranges from 1622 m to 4792 m.
159
The DECREASE precipitation regime (Figure 10) results in annual average rainfall of
244 mm. Shallow soil moisture ranges from 6.6% to 19.2%. Deep soil moisture ranges
from 8.5% to 11%. Albedo ranges from 14.4 to 18.5 %. The PBLh is variable depending
on the empirical relationship: for f(sh,) PBLh ranges from 1419 m to 2939 m, for
f(sh,d) PBLh ranges from 1301 m to 2080 m, and for (s,d,) PBLh ranges from 1697 m
to 3237 m.
When categorizing by cases using the two-layer conceptual framework, the CURRENT
and INCREASE scenario present more days of Case 4 (CURRENT= 47%, INCREASE =
57%), while the scenario DECREASE has more days of Case 1 (45%).
We explored two simple possibilities for an increase and decrease by 20% of
precipitation inputting the system. We are conservative in this analysis especially, since
regional modeling is not completely consistent [Weltzin et al., 2003]. The decrease in
average precipitation could affect soil moisture-vegetation-atmosphere interactions
depending on the frequency and intensity of the events [Knapp et al., 2008]. In any case,
this framework allows for the exploration of changes in precipitation dynamics and the
influence these changes might have on the PBLh, an indicator of larger scale land surface
– atmosphere interactions.
160
3.4 Implications of modeling two-layer soil moisture dynamics
We explored an ecohydrological application using a two-layer bucket model (Figure 4).
Our results show that a two-layer approach is feasible and that using this approach, we
can simulate both shallow and deep layer soil moisture dynamics for use in multiple
applications. Notably however, shallow soil moisture is better represented than deep soil
moisture (Figure 5), leaving room for further improvement, perhaps through the inclusion
of hydraulic redistribution [Guswa, 2012; Nadezhdina et al., 2010; Ryel et al., 2004].
We also look at the two-layer soil moisture dynamics of bare and canopy patches
individually (Figures 8, 9, and 10). To do this, we take into account that roots do not
distribute homogeneously beneath the surface [Breshears et al., 2009; Loik et al., 2004],
and allocate more roots under the canopy than under the bare patches [Sanchez-Mejia and
Papuga, in review]. Because of the smaller fraction of roots under the bare patches, we
assume a smaller maximum transpiration (Tmax) from the bare patches than the canopy
patches for our site. Because these root distributions do generalize to all vegetation and
soil types, parameterizations should be based off of specific site characterizations. With
this in mind, the partitioning of the soil into a shallow and deep layer and the surface into
canopy and bare patches, we can improve our understanding the dynamics of evaporation
and transpiration partitioning by applying this two-layer model, especially if we know
where is water coming from [Cavanaugh et al., 2011; Kurc and Small, 2007; Yepez et al.,
2007].
161
Acknowledgments.
This research was supported in part by University of Arizona College of Agriculture and
Life Sciences (CALS), The Arizona University System Technology and Research
Initiative Fund (TRIF), The University of Arizona Office of the Vice President for
Research (VPR), SAHRA (Sustainability of Semi-Arid Hydrology and Riparian area)
under the STC Program of the National Science Foundation (NSF), NSF CAREER
Award # EAR-1255013, the UA Springfield Scholarship for Graduate Research in
Rangelands, the UA William A. Calder III PhD Scholarship for Mexican graduate
students focused in conservation, and the UA Kel M. Fox Watershed Scholarship.
References.
Asbjornsen, H., et al. (2011), Ecohydrological advances and applications in plant-water
relations research: a review, Journal of Plant Ecology, 4(1-2), 3-22.
Baldocchi, D., et al. (2001), FLUXNET: A new tool to study the temporal and spatial
variability of ecosystem-scale carbon dioxide, water vapor, and energy flux densities,
Bull. Amer. Meteorol. Soc., 82(11), 2415-2434.
Basara, J., and K. Crawford (2002), Linear relationships between root-zone soil moisture
and atmospheric processes in the planetary boundary layer, Journal of Geophysical
Research-Atmospheres, 107(D15).
Betts, A. K., and J. H. Ball (1994), Budget analysis of FIFE-1987 sonde data, J. Geophys.
Res.-Atmos., 99(D2), 3655-3666.
162
Bonser, J. D., T. E. Unny, and K. Singhal (1985), A marked Poisson-process model of
summer rainfall in southern Ontario, Canadian Journal of Civil Engineering, 12(4),
886-898.
Breshears, D., O. Myers, and F. Barnes (2009), Horizontal heterogeneity in the frequency
of plant-available water with woodland intercanopy-canopy vegetation patch type
rivals that occuring vertically by soil depth, Ecohydrology, 2(4), 503-519.
Breshears, D., O. Myers, S. Johnson, C. Meyer, and S. Martens (1997), Differential use
of spatially heterogeneous soil moisture by two semiarid woody species: Pinus edulis
and Juniperus monosperma, Journal of Ecology, 85(3), 289-299.
Calvet, J. C., and J. Noilhan (2000), From near-surface to root-zone soil moisture using
year-round data, J. Hydrometeorol., 1(5), 393-411.
Cavanaugh, M. L., S. A. Kurc, and R. L. Scott (2011), Evapotranspiration partitioning in
semiarid shrubland ecosystems: a two-site evaluation of soil moisture control on
transpiration, Ecohydrology, 4(5), 671-681.
Chahine, M. T. (1992), The hydrological cycle and its influence on climate, Nature,
359(6394), 373-380.
Charney, J. G. (1975), Dynamics of deserts and drought in the Sahel, Q. J. R. Meteorol.
Soc., 101(428), 193-202.
Chen, F., and J. Dudhia (2001), Coupling an advanced land surface-hydrology model
with the Penn State-NCAR MM5 modeling system. Part I: Model implementation
and sensitivity, Mon. Weather Rev., 129(4), 569-585.
163
Claussen, M. (1997), Modeling bio-geophysical feedback in the African and Indian
monsoon region, Clim. Dyn., 13(4), 247-257.
Cosby, B. J., G. M. Hornberger, R. B. Clapp, and T. R. Ginn (1984), A Statistical
Exploration of the Relationships of Soil Moisture Characteristics to the Physical
Properties of Soils, Water Resour. Res., 20(6), 682-690.
D'Odorico, P., F. Laio, A. Porporato, L. Ridolfi, A. Rinaldo, and I. R. Iturbe (2010),
Ecohydrology of Terrestrial Ecosystems, Bioscience, 60(11), 898-907.
Daly, E., A. Porporato, and I. Rodriguez-Iturbe (2004), Coupled dynamics of
photosynthesis, transpiration, and soil water balance. Part II: Stochastic analysis and
ecohydrological significance, Journal of Hydrometeorology, 5(3), 559-566.
Dekker, S. C., M. Rietkerk, and M. F. P. Bierkens (2007), Coupling microscale
vegetation-soil water and macroscale vegetation-precipitation feedbacks in semiarid
ecosystems, Glob. Change Biol., 13(3), 671-678.
Diak, G. R., and M. S. Whipple (1993), Improvements to models and methods for
evaluating the land-surface energy-balance and effective roughness using radiosonde
reports and satellite-measured skin temperature data, Agric. For. Meteorol., 63(3-4),
189-218.
Ghan, S. J., J. C. Liljegren, W. J. Shaw, J. H. Hubbe, and J. C. Doran (1997), Influence of
subgrid variability on surface hydrology, J. Clim., 10(12), 3157-3166.
Giorgi, F., and R. Avissar (1997), Representation of heterogeneity effects in earth system
modeling: Experience from land surface modeling, Reviews of Geophysics, 35(4),
413-437.
164
Guswa, A. J. (2012), Canopy vs. Roots: Production and Destruction of Variability in Soil
Moisture and Hydrologic Fluxes, Vadose Zone Journal, 11(3), 13.
Guswa, A. J., M. A. Celia, and I. Rodriguez-Iturbe (2002), Models of soil moisture
dynamics in ecohydrology: A comparative study, Water Resour. Res., 38(9), 15.
Henderson-Sellers, A. (1993), Continental Vegetation as a Dynamic Component of a
Global Climate Model - A Preliminary Assessment, Clim. Change, 23(4), 337-377.
Jackson, R., S. Idso, and R. Reginato (1976), Calculation of evaporation rates during the
transition from energy-limiting to soil-limiting phases using albedo data, Water
Resources Research, 12(1), 23-26.
Katul, G., A. Porporato, and R. Oren (2007), Stochastic dynamics of plant-water
interactions, in Annu. Rev. Ecol. Evol. Syst., edited, pp. 767-791.
Knapp, A. K., et al. (2008), Consequences of More Extreme Precipitation Regimes for
Terrestrial Ecosystems, Bioscience, 58(9), 811-821.
Kurc, S., and E. Small (2004), Dynamics of evapotranspiration in semiarid grassland and
shrubland ecosystems during the summer monsoon season, central New Mexico,
Water Resources Research, 40(9).
Kurc, S., and E. Small (2007), Soil moisture variations and ecosystem-scale fluxes of
water and carbon in semiarid grassland and shrubland, Water Resources Research,
43(6).
Kurc, S., and L. Benton (2010), Digital image-derived greenness links deep soil moisture
to carbon uptake in a creosotebush-dominated shrubland, Journal of Arid
Environments, 74(5), 585-594.
165
Kurc, S., and E. Small (in review), Simple ecohydrological models: comparison to
observations of dryland water and carbon fluxes, edited, Adv. Water Resour.
Laio, F., A. Porporato, L. Ridolfi, and I. Rodriguez-Iturbe (2001), Plants in watercontrolled ecosystems: active role in hydrologic processes and response to water
stress - II. Probabilistic soil moisture dynamics, Advances in Water Resources, 24(7),
707-723.
Liu, B. L., F. Phillips, S. Hoines, A. R. Campbell, and P. Sharma (1995), Water
movement in desert soil traced by hydrogen and oxygen isotopes, chloride, and
chlorine-36, southern Arizona, J. Hydrol., 168(1-4), 91-110.
Loik, M. E., D. D. Breshears, W. K. Lauenroth, and J. Belnap (2004), A multi-scale
perspective of water pulses in dryland ecosystems: climatology and ecohydrology of
the western USA, Oecologia, 141(2), 269-281.
Lyon, S. W., et al. (2008), Coupling terrestrial and atmospheric water dynamics to
improve prediction in a changing environment, Bull. Amer. Meteorol. Soc., 89(9),
1275-1279.
Manabe, S. (1969), Climate and ocean circulation. I. Atmospheric circulation and
hydrology of Earth's surface, Mon. Weather Rev., 97(11), 739-&.
Mills, R. (1990), Remote Sensing Science for the Nineties. IGARSS '90. 10th Annual
International Geoscience and Remote Sensing Symposium (Cat. No.90CH2825-8),
Remote Sensing Science for the Nineties. IGARSS '90. 10th Annual International
Geoscience and Remote Sensing Symposium (Cat. No.90CH2825-8), 3 vol. l+2488
pp-2483 vol. l+2488 pp.
166
Moncrieff, J., P. Jarvis, and R. Valentini (2000), Canopy Fluxes, in Methods in
Ecosystem Science, edited by O. Sala, R. Jackson, H. Mooney and R. Howarth, pp.
161-180, Springer-Verlag, New York.
Moran, M. S., R. L. Scott, T. O. Keefer, W. E. Emmerich, M. Hernandez, G. S. Nearing,
G. B. Paige, M. H. Cosh, and P. E. O'Neill (2009), Partitioning evapotranspiration in
semiarid grassland and shrubland ecosystems using time series of soil surface
temperature, Agric. For. Meteorol., 149(1), 59-72.
Mueller, B., et al. (2011), Evaluation of global observations-based evapotranspiration
datasets and IPCC AR4 simulations, Geophys. Res. Lett., 38.
Nadezhdina, N., et al. (2010), Trees never rest: the multiple facets of hydraulic
redistribution, Ecohydrology, 3(4), 431-444.
NAST, N. A. S. T. U. G. C. R. P. (2000), Climate Change Impacts on the United States:
The potential consequences of variability and change., edited, Cambridge University
Press, New York.
Nemani, R., L. Pierce, S. Running, and S. Goward (1993), Developing satellite-derived
estimate of surface moisture status, J. Appl. Meteorol., 32(3), 548-557.
Otterman, J., M. D. Novak, and D. O. Starr (1993), Turbulent heat-transfer from a
sparsely vegetated surface - 2-component representation, Bound.-Layer Meteor.,
64(4), 409-420.
Pitman, A. J. (2003), The evolution of, and revolution in, land surface schemes designed
for climate models, Int. J. Climatol., 23(5), 479-510.
167
Porporato, A., E. Daly, and I. Rodriguez-Iturbe (2004), Soil water balance and ecosystem
response to climate change, Am. Nat., 164(5), 625-632.
Porporato, A., P. D'Odorico, F. Laio, L. Ridolfi, and I. Rodriguez-Iturbe (2002),
Ecohydrology of water-controlled ecosystems, Adv. Water Resour., 25(8-12), 13351348.
Quinn, P., K. Beven, and A. Culf (1995), The introduction of macroscale hydrological
complexity into land surface-atmosphere transfer models and the effect on planetary
boundary layer development, J. Hydrol., 166(3-4), 421-444.
Raz-Yaseef, N., D. Yakir, G. Schiller, and S. Cohen (2012), Dynamics of
evapotranspiration partitioning in a semi-arid forest as affected by temporal rainfall
patterns, Agricultural and Forest Meteorology, 157, 77-85.
Reynolds, J. F., P. R. Kemp, and J. D. Tenhunen (2000), Effects of Long-Term Rainfall
Variability on Evapotranspiration and Soil Water Distribution in the Chihuahuan
Desert: A Modeling Analysis, Plant Ecology, 150(1/2), 145-159.
Rietkerk, M., M. C. Boerlijst, F. van Langevelde, R. HilleRisLambers, J. van de Koppel,
L. Kumar, H. H. T. Prins, and A. M. de Roos (2002), Self-organization of vegetation
in arid ecosystems, Am. Nat., 160(4), 524-530.
Rodriguez-Iturbe, I. (2000), Ecohydrology: A hydrologic perspective of climate-soilvegetation dynamics, Water Resour. Res., 36(1), 3-9.
Rodriguez-Iturbe, I., A. Porporato, L. Ridolfi, V. Isham, and D. R. Cox (1999),
Probabilistic modelling of water balance at a point: the role of climate, soil and
168
vegetation, Proc. R. Soc. London Ser. A-Math. Phys. Eng. Sci., 455(1990), 37893805.
Ryel, R. J., A. J. Leffler, M. S. Peek, C. Y. Ivans, and M. M. Caldwell (2004), Water
conservation in Artemisia tridentata through redistribution of precipitation,
Oecologia, 141(2), 335-345.
Sala, O. E., and W. K. Lauenroth (1982), Small rainfall events: An ecological role in
semiarid regions, Oecologia, 53(3), 301-304.
Salvucci, G. D. (2001), Estimating the moisture dependence of root zone water loss using
conditionally averaged precipitation, Water Resour. Res., 37(5), 1357-1365.
Sanchez-Mejia, Z. M., and S. A. Papuga (in review), Observations of a two-layer soil
moisture influence on surface energy dynamics and planetary boundary layer
characteristics in a semiarid shrubland, Water Resour. Res.
Sanchez-Mejia, Z. M., S. A. Papuga, and J. B. Swetish (in review), Quantifying the
influence of deep soil moisture on ecosystem albedo: the role of vegetation, Water
Resour. Res.
Santanello, J., M. Friedl, and W. Kustas (2005), An empirical investigation of convective
planetary boundary layer evolution and its relationship with the land surface, Journal
of Applied Meteorology, 44(6), 917-932.
Santanello, J., C. Peters-Lidard, and S. Kumar (2011), Diagnosing the Sensitivity of
Local Land-Atmosphere Coupling via the Soil Moisture-Boundary Layer Interaction,
Journal of Hydrometeorology, 12(5), 766-786.
169
Santanello, J., C. Peters-Lidard, A. Kennedy, and S. Kumar (2013), Diagnosing the
Nature of Land-Atmosphere Coupling: A Case Study of Dry/Wet Extremes in the US
Southern Great Plains, Journal of Hydrometeorology, 14(1), 3-24.
Scheffer, M., M. Holmgren, V. Brovkin, and M. Claussen (2005), Synergy between
small- and large-scale feedbacks of vegetation on the water cycle, Glob. Change
Biol., 11(7), 1003-1012.
Scott, R. L., W. L. Cable, and K. R. Hultine (2008), The ecohydrologic significance of
hydraulic redistribution in a semiarid savanna, Water Resour. Res., 44(2), 12.
Scott, R. L., T. E. Huxman, W. L. Cable, and W. E. Emmerich (2006), Partitioning of
evapotranspiration and its relation to carbon dioxide exchange in a Chihuahuan
Desert shrubland, Hydrol. Process., 20(15), 3227-3243.
Seneviratne, S. I., T. Corti, E. L. Davin, M. Hirschi, E. B. Jaeger, I. Lehner, B. Orlowsky,
and A. J. Teuling (2010), Investigating soil moisture-climate interactions in a
changing climate: A review, Earth-Sci. Rev., 99(3-4), 125-161.
Shuttleworth, W. J. (1993), Evaporation, in Handbook of Hydrology, edited by D.
Maidement, p. 53pp, McGraw-Hill, New York.
Small, E. E., and S. A. Kurc (2003), Tight coupling between soil moisture and the surface
radiation budget in semiarid environments: Implications for land-atmosphere
interactions, Water Resour. Res., 39(10), 14.
Stull, R. (1988), An Introduction to Boundary Layer Meteorology, edited, p. 666, Kluwer
Acad. Publ., Boston, London.
170
Trenberth, K., A. Dai, R. Rasmussen, and D. Parsons (2003), The changing character of
precipitation, Bulletin of the American Meteorological Society, 84(9), 1205-+.
Villegas, J. C., D. D. Breshears, C. B. Zou, and D. J. Law (2010), Ecohydrological
controls of soil evaporation in deciduous drylands: How the hierarchical effects of
litter, patch and vegetation mosaic cover interact with phenology and season, J. Arid.
Environ., 74(5), 595-602.
Wang, L., P. D'Odorico, J. P. Evans, D. J. Eldridge, M. F. McCabe, K. K. Caylor, and E.
G. King (2012), Dryland ecohydrology and climate change: critical issues and
technical advances, Hydrol. Earth Syst. Sci., 16(8), 2585-2603.
Weltzin, J. F., et al. (2003), Assessing the response of terrestrial ecosystems to potential
changes in precipitation, Bioscience, 53(10), 941-952.
Western, A. W., R. B. Grayson, G. Bloschl, G. R. Willgoose, and T. A. McMahon
(1999), Observed spatial organization of soil moisture and its relation to terrain
indices, Water Resources Research, 35(3), 797-810.
Yepez, E. A., R. L. Scott, W. L. Cable, and D. G. Williams (2007), Intraseasonal
variation in water and carbon dioxide flux components in a semiarid riparian
woodland, Ecosystems, 10(7), 1100-1115.
Zeng, N., J. D. Neelin, K. M. Lau, and C. J. Tucker (1999), Enhancement of interdecadal
climate variability in the Sahel by vegetation interaction, Science, 286(5444), 15371540.
171
Table 1. Shallow layer soil moisture threshold and  (exponential time constant), and
deep layer soil moisture threshold.
shallow
deep
Ecosystem
threshold
0.1229
0.1013
Bare
 threshold
3.5
0.1251
0.1015
Canopy
 threshold
3.6
0.1142
0.1003

2.7
172
Table 2. Coefficients of 2nd order polynomial empirically-derived relationships
173
Table 3. Parameters used in two-layer bucket model based on vegetation and soil similar
to our study site. The selection process describes the source from which the parameter
was derived.
Symbol
Layer depth
ZE
ZT
Soil parameters
n
Ksat
B
s*
sfc
sh
sw
Value
Selection Process
0-20 cm
20-60 cm
Estimated from root profiles
Estimated from root profiles
0.43
80 cm d-1
4.9
0.46
0.56
0.075
0.085
[Cosby et al., 1984]
[Laio et al., 2001]
[Laio et al., 2001]
[Laio et al., 2001]
[Laio et al., 2001]
[Laio et al., 2001]
[Laio et al., 2001]
Atmospheric loss parameters
Ecosystem Emax
0.45 cm d-1
Ecosystem Tmax
0.5 cm d-1
Bare Patch Emax
0.5 cm d-1
Bare Patch Tmax
0.4 cm d-1
Canopy Patch Emax
0.4 cm d-1
Canopy Patch Tmax
0.55 cm d-1
[Moran et al., 2009;
Reynolds et al., 2000;
Villegas et al., 2010]
174
Table 4. Number of days for each Case applying two-layer soil moisture conceptual
framework to classify observed and model data. Ecosystem refers to ecosystem soil
moisture idem for bare and canopy.
Ecosystem
Observed
Model
Bare
Observed
Model
Canopy
Observed
Model
Case 1
Case2
Case3
Case4
608
587
13
18
144
216
247
274
604
557
15
15
145
151
248
372
670
669
6
74
106
242
246
110
175
List of Figures
Figure 1. Planetary boundary layer height PBLh (m) estimated as a function of a) shallow
soil moisture (sh) and atmospheric stability.
Figure 2. Planetary boundary layer height PBLh (m) estimated as a function of a) shallow
soil moisture (sh) and atmospheric stability (ML) for a) Case 1, b) Case 2, c) Case 3, and
d) Case 4.
Figure 3. Planetary boundary layer height PBLh (m) estimated as a function of a) shallow
soil moisture (sh) and albedo (), and b) deep soil moisture (d) and albedo.
Figure 4. Modified two-layer bucket model
Figure 5. Two-layer bucket soil moisture model performance compared to observed data.
Figure 6. The variability in (a) planetary boundary layer height (PBLh), (b) shallow soil
moisture, (c) deep soil moisture, and (d) albedo with respect to Case (1, 2, 3, 4).
Figure 7. Observed versus model PBLh as a function of a) shallow soil moisture and
albedo, b) shallow and deep soil moisture, c) shallow and deep soil moisture and albedo.
176
Figure 8. Land surface and PBLh dynamics based on modeling approach using
CURRENT average annual rainfall (8.3 mm) and storm frequency (10 days) as input. E
refers to ecosystem, B to bare, and C to canopy.
Figure 9. Land surface and PBLh dynamics based on modeling approach using a 20%
INCREASE in average annual rainfall (9.96 mm) and storm frequency (10 days) as input.
E refers to ecosystem, B to bare, and C to canopy.
Figure 10. Land surface and PBLh dynamics based on modeling approach using a 20%
DECREASE in average annual rainfall (6.64 mm) and storm frequency (10 days) as
input. E refers to ecosystem, B to bare, and C to canopy.
177
Figure 1.
178
Figure 2.
179
Figure 3.
180
Figure 4.
181
Figure 5.
182
Figure 6.
183
Figure 7.
184
Figure 8.
185
Figure 9.
186
Figure 10.