AN ABSTRACT OF THE THESIS OF
Sonja E. Raven for the degree of Master of Science in Rangeland Resources
presented on May 3, 2004.
Title: Describing the Thermal Regimes of Two Northeastern Oregon Streams
Abstract approved:
Redacted for Privacy
arson
Michael M. Borman
Two tributaries of the south fork of the Burnt River, near Unity Oregon were studied
during the summers of 2000 and 2001 to determine water heating and cooling
patterns. Hourly temperature data were recorded for air, water, and soil parameters
at four elevations 150m apart on Barney and Stevens Creeks. Both creeks shared the
same geographic attributes but were found to heat quite differently. These
differences may be attributed to differences in discharge, exposure time, temperature
gradient, and vegetative cover. Lags between peak heating of air and water were
evident on both creeks, with slight differences. Lags between air and water during
the cooling cycle were not as evident, particularly on Stevens Creek. Although
differences between the creeks' thermal regimes were apparent, equations showed
strong association between maximum air temperature and maximum water
temperature, with minimum air temperature also showing strong association with
maximum water temperature on portions of Stevens Creek. Stepwise regression
analysis established a strong relationship (R2 0.69-0.9 1) between maximum air
temperature and peak water temperature change (the maximum temperature change
recorded in a 2 hour heating period between 5am-5pm). An equation to predict
water temperature change in a given day, based on derivatives of air and water
temperature also showed strong fit and good predictive ability on both creeks. Thus
while each creek was different and responded to its thermal environment, certain air
temperature variables had strong association, regardless of differences between
creeks.
© Copyright by Sonja E. Raven
May 3, 2004
All Rights Reserved
Describing the Thermal Regimes of Two Northeastern Oregon Streams
by
Sonja E. Raven
A THESIS
submitted to
Oregon State University
in partial fulfillment of
the requirements for the
degree of
Master of Science
Presented May 3, 2004
Commencement June 2005
Master of Science thesis of Sonja E. Raven presented on May 3, 2004
APPROVED:
Redacted for Privacy
-M.rofe sor, representing Rangeland Resources
Redacted for Privacy
Co-Major Professor, representing Rangeland Resources
Redacted for Privacy
Head of the Department of 'angeltnd Resources
Redacted for Privacy
Dean of the
te School
I understand that my thesis will become part of the permanent collection of Oregon
State University libraries. My signature below authorizes release of my thesis to any
reader upon request.
Redacted for Privacy
Sonja E. Raven, Author
ACKNOWLEDGEMENTS
So many people have been a part of this project from the beginning, and I am so
grateful to them.
Thank you to the professors at the University College of the Cariboo in Kamloops,
BC, for the incredible undergraduate education I received and their complete support
of my dream to pursue my Masters at OSU. Thanks to John Buckhouse for inviting
me to OSU, and his wonderful wife Vicki for being my Corvallis Mom. Gratitude
and thanks go to Mike Borman and Larry Larson, my two major professors, for
taking me on and giving me this wonderful opportunity as well as all their support,
guidance and understanding. Kudos and thanks to Dave Thomas, minor professor
and statistician par excellence. I will always be grateful for your tireless patience in
unraveling the mysteries of SAS and statistics for me - you are a saint. Special
thanks to George Taylor for being part of my committee and helping to support and
guide me. Thanks to Dr. James Males of Animal Sciences for serving as my Grad
School Representative. Thanks to the faculty and staff of the Department of
Rangeland Resources for helping me to further my education with excellence.
Special thanks to the graduate students of the Range Department. You were always
there for me, whether I needed to laugh or cry - and you are all pot-luckers of
superior talent!
The Burnt River Irrigation District and members of the surrounding community also
deserve special mention. To the district: your proactive attitude is a credit to you, as
is your Irrigation District Manager, Jerry Franke. To the district and community: I
so appreciated all the support and interest you showed in my project. You are a fine
group of people and I am privileged to have known and worked with you. Cheryl
Bradford of the USFS also deserves special recognition for her kindness to me and
her helpfulness with this project.
A huge loving thank you to Pat and Jerry Franke - my American parents. Thank you
for welcoming me into your lives and hearts. Your generosity, kindness and love are
gifts I treasure. And I will always be grateful for the installation of the shower at the
office!! But Jerry, I really think you need a few more cats!
For all my friends "back home" in Canada, thank you for your belief in and support
of me - I have missed you all. A special thank you to Kathy Lowson, for her
tremendous help in editing my masterpiece!! And for all my friends here in Oregon
- how lucky I am to have gotten to know you, and how I will miss you all when I'm
gone.
My family has been with me through thick and thin, and I would never have made it
without them. For my amazing sister Karen Raven - friend, confidante, shrink and
field tech extraordinaire. Our mother never would have believed we could live in a
24-foot travel trailer (for not one but two summers!) and live to tell about it! Those
two summers were the best. And hauling rocks for you was a privilege! For my
mother, Betty Raven, for her steadfast belief in me, and her wisdom when I really
needed it. For my father, Arne Raven, for his incredible unstinting love and support
of me - this is as much yours as mine, Dad. And finally, my wonderful husband
Wade Whibley, who helped me spread my wings and soar with his steadfast
devotion, constant emotional (and technical!) support, and unwavering belief in me.
TABLE OF CONTENTS
Page
INTRODUCTION
1
LITERATURE REVIEW
3
THE PROCESS OF HEAT TRANSFER
Meteorological, Hydrological and Other Variables
3
3
Meteorological Variables
4
Hydrological Variables
.6
Other Variables Influencing Heat Transfer in Streams
.7
THE THEORY OF EQUILIBRIUM TEMPERATURE
.9
PREDICTION OF STREAM TEMPERATURE
10
Solar Radiation
.10
Air Temperature
10
CONCLUSION
.12
STUDY AREA
MATERIALS AND METHODS
DATA COLLECTION
DATA ANALYSIS
16
.. 16
18
Objective 1: Determining the Influence of Elevation, MonthlYear, and
Stream on Thermal Environment
.19
Objective 2: Describing Diurnal Air and Water Cycles
19
Objective 3: Modeling Variables Strongly Associated with
Maximum Water Temperature
.20
Objective 4: Developing a Predictive Equation for Water
Temperature Change
.22
RESULTS AND DISCUSSION
.24
OBJECTIVE 1: DETERMINING THE INFLUENCE OF ELEVATION,
MONTH/YEAR, AND STREAM ON THERMAL ENVIRONMENT
24
Results
Elevation
24
24
TABLE OF CONTENTS (Continued)
Month
Page
27
Year
27
Stream
30
Discharge, Velocity and Area
32
Distances, Gradients and Exposure Times
33
Tributaries and Vegetation Differences
35
Discussion
.36
OBJECTIVE 2: DESCRIBING DIURNAL AIR AND WATER
CYCLES
Results
.40
40
Air Temperature Patterns
.40
Heating Cycle
40
Cooling Cycle
.. . .48
Water Temperature Patterns
Heating Cycle
.51
.51
.
Cooling Cycle
Discussion
.57
.
.60
OBJECTIVE 3: MODELING VARIABLES STRONGLY ASSOCIATED
WITH MAXIMUM WATER TEMPERATURE
62
Results
.62
Discussion
69
OBJECTIVE 4: DEVELOPING A PREDICTIVE EQUATION FOR
WATER TEMPERATURE CHANGE
72
Results
Discussion
SUMMARY AND CONCLUSIONS
OBJECT1VEI: DETERMINING THE iNFLUENCE OF ELEVATION,
MONTH/YEAR, AND STREAM ON THERMAL ENVIRONMENT
.72
77
..
.79
79
TABLE OF CONTENTS (Continued)
Page
OBJECTIVE 2: DESCRIBING DIURNAL AIR AND WATER
CYCLES
80
OBJECTIVE 3: MODELING VARIABLES STRONGLY ASSOCIATED
WITH MAXIMUM WATER TEMPERATURE
.81
.
OBJECTIVE 4: DEVELOPING A PREDICTIVE EQUATION FOR
WATER TEMPERATURE CHANGE
82
BIBLIOGRAPHY
.84
APPENDICES
90
LIST OF FIGURES
Figure
Page
Diagram of the process of heat transfer to and from a stream
.3
MapofStudyArea
14
Diagram of 450 v-notch weir
17
Mean air, water, and soil temperatures by months and years for Barney
Creek, 2000-2001
28
Mean air, water, and soil temperatures by months and years for Stevens
29
Creek, 2000-200 1
Mean air, water, and soil temperatures by elevations for each stream,
month, and year
2-hour period air temperature changes with average hourly air
temperature at beginning of each 2-hour period, for 1370-m
Stevens Creek July 2000
.
.31
.41
2-hour period temperature changes for air, Barney Creek
July2000
43
2-hour period temperature changes for air, Stevens Creek
July2000
44
2-hour period temperature changes for water, Barney Creek
July2000
52
2-hour period temperature changes for water, Stevens Creek
July2000
Scatterplots demonstrating fit of prediction model for Barney
Creek 2000
Scatterplots demonstrating fit of prediction model for Stevens
Creek 2000
. .73
74
LIST OF TABLES
Table
Maximum, minimum, and mean air temperature data, as well as
historical air temperatures for the Burnt River area
Page
15
Partitioning the influence of the fixed effects of stream, elevation, year, and
month, as well as their interactions, for each parameter
25
Regression coefficient estimates for elevation in °Celsius/150 m elevation
decrease
. .26
Discharge, velocity, and area for both creeks
33
Distances, gradients, and exposure times for water on both creeks
35
Frequencies of when peak air heating occurred across elevations on Barney
Creek, both years
46
Frequencies of when peak air heating occurred across elevations on Stevens
Creek, both years
47
Frequencies of when peak air cooling occurred across elevations on Barney
Creek, both years
49
Frequencies of when peak air cooling occurred across elevations on Stevens
Creek, both years
50
Frequencies of when peak water heating occurred across elevations on
Barney Creek, both years
55
Frequencies of when peak water heating occurred across elevations on
Stevens Creek, both years
. . .56
Frequencies of when peak water cooling occurred across elevations on
Barney Creek, both years
.58
Frequencies of when peak water cooling occurred across elevations on
Stevens Creek, both years
.59
Table of variables strongly associated with maximum water temperature,
for Barney Creek, year 2000
63
Table of variables strongly associated with maximum water temperature,
for Stevens Creek, year 2000
64
Summary of the models explaining maximum water temperature on
Barney Creek 2000-2001, including model R2, coefficients of the
variables, and their SEs
.67
LIST OF TABLES (Continued)
Table
Page
Summary of the models explaining maximum water temperature on
Stevens Creek 2000-200 1, including model R2, coefficients of the
variables, and their SEs
.68
Barney Creek 2000, showing fit of model, mean absolute deviation
in °C (with SDs in brackets), and variable coefficients (with SEs in
brackets)
75
Stevens Creek 2000, showing fit of model, mean absolute deviation
in °C (with SDs in brackets), and variable coefficients (with SEs in
brackets)
76
LIST OF APPENDICES
Appendix
Page
2-hour period temperature changes for air, Barney and Stevens Creeks, July
and August 2000 and July and August 2001
91
2-hour period temperature changes for water, Barney and Stevens
Creeks, July and August 2000 and July and August 2001
Tables of variables strongly associated with maximum
water temperature for both creeks in 2000 and 2001
Scatterplots showing the fit of predicted versus observed
values for both creeks in 2000 and 2001
Tables showing fit of the model to predict maximum water temperature
change, by elevation, for both creeks in 2000 and 2001
.. 100
109
...1 14
119
LIST OF APPENDIX FIGURES
Figure
Page
2-hour period temperature changes for air, Barney Creek
July2000
92
2-hour period temperature changes for air, Stevens Creek
July2000
.93
2-hour period temperature changes for air, Barney Creek
August2000
.94
2-hour period temperature changes for air, Stevens Creek
August2000
.95
2-hour period temperature changes for air, Barney Creek
July2001
..96
2-hour period temperature changes for air, Stevens Creek
July2001
..97
2-hour period temperature changes for air, Barney Creek
August2001
.98
2-hour period temperature changes for air, Stevens Creek
August2001
.99
2-hour period temperature changes for water, Barney Creek
July2000
101
2-hour period temperature changes for water, Stevens Creek
July2000
102
2-hour period temperature changes for water, Barney Creek
August2000
103
2-hour period temperature changes for water, Stevens Creek
August2000
104
2-hour period temperature changes for water, Barney Creek
July2001
105
2-hour period temperature changes for water, Stevens Creek
July2001
106
2-hour period temperature changes for water, Barney Creek
August2001
107
LIST OF APPENDIX FIGURES (Continued)
Figure
Page
2-hour period temperature changes for water, Stevens Creek
August2001
108
Scatterplots demonstrating fit of prediction model for Barney
Creek 2000
.115
Scatterplots demonstrating fit of prediction model for Stevens
Creek 2000
.116
Scatterplots demonstrating fit of prediction model for Barney
Creek200l
.117
Scatterplots demonstrating fit of prediction model for Stevens
Creek200l
118
LIST OF APPENDIX TABLES
Table
Page
Table of variables strongly associated with maximum water
temperature, for Barney Creek, year 2000
110
Table of variables strongly associated with maximum water
temperature, for Stevens Creek, year 2000
111
Table of variables strongly associated with maximum water
temperature, for Barney Creek, year 2001
112
Table of variables strongly associated with maximum water
temperature, for Stevens Creek, year 2001
113
Barney Creek 2000, showing fit of model, mean absolute deviation
in °C (with SDs in brackets), and variable coefficients
(with SEs in brackets)
120
Barney Creek 2001, showing fit of model, mean absolute deviation
in °C (with SDs in brackets), and variable coefficients
(with SEs in brackets)
121
Stevens Creek 2000, showing fit of model, mean absolute deviation
in °C (with SDs in brackets), and variable coefficients
(with SEs in brackets)
122
Stevens Creek 2001, showing fit of model, mean absolute deviation
in °C (with SDs in brackets), and variable coefficients
(with SEs in brackets)
123
DEDICATION
This thesis is dedicated to my father, Arne Raven, for giving me wings and showing
me how to use them. You are a fine man, and I am proud to be your daughter.
DESCRIBING THE THERMAL REGIMES OF TWO
NORTHEASTERN OREGON STREAMS
INTRODUCTION
Water quality issues continue to be at the forefront of ecological concerns in the state
of Oregon, particularly with regard to stream temperatures. In order to better
understand how streams heat in northeastern Oregon, a study was continued to make
a closer examination of the diurnal air and water cycles on two headwater streams in
the Burnt River drainage in Baker County.
Prior research had established general elevation effects of these two streams, but an
examination of how streams function over a day-long period was expected to lead to
a determination of some of the important variables having a strong relationship with
water temperature. Closer examination of the diurnal cycles would show how the air
and water temperatures fluctuate over the course of a 24-hour period, allowing us a
more detailed understanding of when and where peak heating or cooling occurred for
each parameter. Elevational influences could also be examined, to see if they
followed the same general pattern established in previous research by Meays (2000).
Our study of two creeks in northeastern Oregon will allow us to examine the diurnal
cycles of both creeks, and further determine variables most strongly related to water
temperature increase thus gaining a better understanding of the thermal regimes
governing these two streams.
2
LITERATURE REVIEW
THE PROCESS OF HEAT TRANSFER
The physics of stream heating are governed by four main processes: conduction,
convection, radiation, and advection. Conduction is the process by which heat
energy is transferred from molecule to molecule within or between substances
(Ahrens 1998). Convection is defined as "the transfer of heat by the mass movement
of a fluid (such as water and air)" (Abrens 1998). Kreith (1958) states that
convection is the most significant method of heat transfer between solids and liquids.
Radiation is the transfer of energy between two objects of differing temperatures that
are separated in space, with the energy flowing from the object having a higher
temperature to the one having a lower temperature (Strahler and Strahler 1992).
Radiation may either be longwave radiation (atmospheric radiation) or shortwave
radiation (solar radiation). Atmospheric radiation has wavelengths between 4-21
jtm, with the majority between 8-14tm, while solar radiation wavelengths range
between 0.2 - 1.2 jim, with most between 0.4 and 1 .2jim. Advection is "the transfer
of heat by horizontally moving air [or water]" (Ahrens 1998), and also the movement
of heat within the water itself. One example of advection is any heat within
precipitation that falls into a stream that may add heat to the stream (LeBlanc et al.
1997; Raphael 1962). These four processes are directly related to the first and
second laws of thermodynamics: Law 1: "The total energy of the universe is
constant.. .therefore energy carmot be created or destroyed" (Silberberg 1996); Law
2: "Every energy transfer or transformation increases the order of the universe"
(Campbell 1993); or to paraphrase Kreith (1958) energy moves from an area of
higher concentration to an area of lower concentration.
3
In terms of cooling, streams primarily lose heat through the process of evaporation,
which is the change in state of the water from liquid to gas (LeBlanc et al. 1997).
The energy required for this change in state may come from other sources, including
the air (Ahrens 1998). Not only is heat lost during the phase change from liquid to
gas, but any heat contained within the water that is evaporated is also removed
(Raphael 1962). Also, streams give off heat by radiating longwave radiation back to
the atmosphere (Sinokrot and Stefan 1993; Smith and Lavis 1975; Raphael 1962).
Figure 1. Diagram of the process of heat transfer to and from a stream.
S1PtE.t)
CI-UCT o
SR
Meteorological, Hydrological, and Other Variables:
The physics of stream heating are heavily influenced by meteorological,
hydrological, and physical variables. For example, conduction heat transfer with the
streambed is of primary importance in shallow streams flowing through exposed
areas, especially if the streambed has a low albedo (Brown 1969). Convective heat
4
transfer between air and water is influenced by wind speed, atmospheric pressure,
and any temperature gradient that might exist between the two (LeBlanc et al. 1997)
Advection plays a role when there are additions of cooler water (eg: precipitation,
groundwater inputs etc.) to a stream with warmer temperatures (LeBlanc et al. 1997;
Raphael 1962).
Meteorological Variables:
Radiation, in either longwave or shortwave form, can also have a significant impact
on water heating (LeBlanc et al. 1997; Sinokrot and Stefan 1994; Stefan and
Preud'homme 1993; Smith and Lavis 1975; Swift and Messer 1971). Beschta (1997)
and Mosley (1983) believed that longwave radiation from the atmosphere is
approximately equalized by back radiation from the stream during the course of an
entire diurnal cycle, thus they considered shortwave radiation to be the primary
source of heat to stream systems. However, not all shortwave radiation is absorbed
by the stream: air bubbles and suspended sediments serve to scatter some of the
radiation, while some is also reflected from the surface of the water itself (Raphael
1962). Also, differing wavelengths of light penetrate and heat water differently. For
example, while much of the shorter wavelength solar radiation is able to penetrate
the water column to a depth of 30 feet (Sellers 1965; Buttolphl955), it is not
associated with significant heating of the water (Hollaender 1956). It is the 7001,000 nanometer (nm) of near infrared radiation and >1,000 nm wavelength of
ambient radiation that is responsible for the majority of heating which occurs in the
top four inches of the water column (Sellers 1965; Buttolphl9S5).
Other meteorological variables influencing temperature dynamics within a stream
include: cloud cover, (Sinokrot and Stefan 1993; Dingman 1971) relative humidity
(Stefan and Preud'homme 1993;Dingman 1971) solar radiation (Beschta 1997;
LeBlanc et al. 1997; Larson and Larson 1996; Stefan and Preud'homme 1993;
5
Beschta and Taylor 1988; Mosley 1983; Feller 1981; Brown and Krygier 1970;
Levno and Rothacher 1967), air temperature (Constanz 1998; Mohseni et al. 1998;
Webb and Nobilis 1997; Larson and Larson 1996; Brown and Binkley 1994; McRae
and Edwards 1994; Ward 1985; Dingman 1971; Levno and Rothacher 1967), and
wind speed (Sinokrot and Stefan 1994; Stefan and Preud'homme 1993; Dingman
1971).
These meteorological variables influence the physics of stream heating in the
following ways: cloud cover directly impacts the amount of solar radiation reaching
the stream (Sinokrot and Stefan 1994; Mosley 1983; Pluhowski and Kantrowitz
1963). Relative humidity is tied with air temperature in influencing evaporative
processes - as the temperature of the air increases, the surrounding atmosphere is
able to hold more moisture, which makes possible an increase in the evaporative heat
transfer from stream to air (Mohseni et al. 1998). Solar radiation reaching and
absorbed by the stream obviously has a direct impact on the amount of heat radiated
to the stream, and air temperature is tied to the convective transfer of heat to the
stream. Amount of wind can also have a significant impact on the evaporative losses
from a stream (Dingman 1971).
Many researchers (Larson and Larson 1996; McRae and Edwards 1994; Sinokrot and
Stefan 1993; Needham and Jones 1959) have indicated that water temperatures
follow air temperatures closely, with peak water temperatures lagging behind peak
air temperatures. According to Edinger et al. (1968), meteorological variables are
largely responsible for this time lag. Stefan and Preud'homme (1993), found this lag
to be depth1discharge dependent, showing a lag of only four hours in shallow rivers
approximately 0.6 m deep, to a lag of up to a week in deeper (4.5 m) rivers. This
corroborated findings from the study by Edinger et aL (1968).
6
Hydrological Variables:
Hydrological variables, particularly the discharge, velocity, depth, and size of
stream, as well as coldwater inputs such as those from groundwater sources or cooler
tributaries can also exert a significant influence on the heat dynamics of a stream
system.
Discharge, depth, and size of the water body are all closely correlated to the volume
of water to be heated. Streams with lower flows experience a lower thermal inertia
(Stefan and Preud'homme 1993), which means they are much more responsive to
small changes in the heat status of the surroundings. As such, they respond rapidly
to any increase or decrease in the ambient heat of the surroundings (Ward 1985;
Mosley 1983; Vugts 1974; Swift and Messer 1971). The less water in the stream, the
less energy it takes to raise its temperature (LeBlanc et al. 1997). Mathematical
relationships indicate that water temperature increases inversely to depthlflow
(Amaranthus et al. 1989). Brown and Krygier (1967) indicated that shallow streams
with low flows are likely to respond with more significant changes in thermal
regime, and Beschta and Taylor (1988) reported that "low flows are associated with
higher water temperatures." Ward (1985) found that the small streams he studied
showed rapid temperature changes of 3°C or greater per hour, strongly supporting
the hypothesis that discharge is particularly significant to heating changes within a
stream. A sensitivity investigation performed in the study by LeBlanc et al. (1997)
found that stream discharge had the greatest impact on stream temperature, which
supported findings by Smith and Lavis (1975). However, LeBlanc et al. (1997)
cautioned that the choice of variables to be analyzed can strongly influence the
sensitivity analysis results, and that different geographic areas may produce different
results.
Velocity of water directly affects the exposure time of water to heating variables as
the water moves downstream. If the water travels through a combination of shade
7
and exposed areas, the faster it moves the less time it is exposed to heating elements.
It stands to reason that the longer the water is exposed to warmer air, soil, or solar
inputs, the greater the likelihood of a temperature increase (Larson and Larson 1997;
Amaranthus et al. 1989).
Width of the channel is directly related to how much surface area is exposed to
heating components; the wider the channel the more area available to exchange heat
(LeBlanc et al. 1997).
Groundwater inputs can serve to dampen the effects of any temperature elevation
(Stringham et al. 1996; Meisner et al. 1988; Ward 1985; Mosley 1983; Vugts 1974;
Brown et al. 1971). Side channel inputs, depending on their temperature and
discharge relative to the main stem, can serve to reduce the impact of heat additions
to streams (Swift and Baker 1973; Brown et al. 1971).
Other Variables influencing heat transfer in streams:
Stream temperatures are also influenced to varying degrees by a variety of natural
factors. These include: vegetative and topographic cover (shade), aspect, elevation,
ambient soil temperature, and actual exposure time of the water to heating elements
(Mohseni and Stefan 1999; Constanz 1998; Amaranthus et al. 1989; Beschta and
Taylor 1988; Meisner et al. 1988; Fowler et al. 1979; Smith and Lavis 1975; Meehan
1970).
Shading of streams, either through vegetative cover or topographic features,
interrupts the passage of radiant energy from the sun, thus reducing solar heat
transfer to the stream (LeBlanc et al. 1997; Meisner 1990; Gray and Edington 1969;
Brown and Krygier 1967). It is important to note that water absorbs little radiation
from the visible spectrum (O.4-O.7j.im) while the top four inches of the water column
8
absorbs the radiation from the near-infrared (0.7-1.0 ixm) and ambient (>1gm length)
spectrum primarily (Buttolph 1955). Many researchers have attributed changes in
stream temperature regimes to the presence/absence of vegetation (Beschta 1997;
LeBlanc et al. 1997; Brown and Binkley 1994; McRae and Edwards 1994;
Amaranthus et al. 1989; Beschta and Taylor 1988; Ward 1985; Brown et al. 1971;
Meehan 1970; Brown and Krygier 1967). The presence of vegetation covering the
stream also reduces nighttime longwave radiation from the stream, which decreases
the nighttime cooling of streams (Swift and Messer 1971).
The orientation of a stream also has a direct effect on the amount of heat available to
reach the water (LeBlanc et al. 1997; Pluhowski 1970). Pluhowski (1970) found that
east-west oriented streams received 7-19% more solar radiation than north-south
oriented streams. LeBlanc et al. (1997) further clarified this phenomena using 3D
modeling, and determined that north-south streams received solar insolation of a
greater intensity, leading to higher maximum temperatures, but east-west streams
received solar insolation for a longer time period, meaning whatever maximum air
temperature they achieved remained for a longer period.
As one moves to higher elevations, the density of the surrounding air decreases
(Hidore and Oliver 1993), which means the air doesn't retain heat as effectively as it
does at lower elevations (Meisner et al. 1988). Also, air temperatures and vapour
pressures are lower (Hidore and Oliver 1993) which leads to more rapid loss of heat
due to evaporative processes as well as a decrease in absorption of radiation
(Meisner et al 1988). This then results in less accumulation of energy during the
day, and increases the speed of thermal cycling within the environment (Larson and
Larson 1997).
Soil temperatures themselves may play an important role, depending on the substrate
of the channel (Castro and Hornberger 1991; Meisner et al. 1988). As the water in
9
the channel is able to percolate into the channel bed and move subsurface, the
temperature of the soil, if lower than the water temperature, may be able to reduce
the temperature of the water (Constanz 1998).
THE THEORY OF EQUILIBRIUM TEMPERATURE:
Hydrological, meteorological, and physical variables all interact in heating a stream.
The concept of 'equilibrium temperature' (Dingman 1971; Edinger et al. 1968) helps
to weave these variables together, particularly the importance of any gradient that
exists between air and water temperatures.
Edinger et al. (1968) stated "a body of water exchanges heat with its surroundings by
short- and long-wave radiation, conduction, and evaporation." The gradient between
the air and the water temperature determines how rapidly heat will be transferred to
or from the water (Law 2 of Thermodynamics): a steep gradient results in much
faster heat transfer than a small gradient (Larson and Larson 1997). Dingman (1971)
indicated that surface water temperatures are constantly being driven to an
equilibrium state "where there is no net exchange of energy with the atmosphere."
This occurs at a rate that is proportional to the difference between the actual surface
water temperature and the equilibrium temperature (Dingmanl97l). Equilibrium
temperature is usually not equal to the ambient air temperature because not only is
the equilibrium temperature dependent on conductive and convective processes, but
it is also related to current relative humidity, wind, and the effects of solar inputs
(Dingman 1971). Thus ambient air temperature and surface water temperature form
the endpoints of the potential equilibrium temperature.
10
PREDICTION OF STREAM TEMPERATURE:
Solar Radiation:
Solar radiation has been thought by some to be the principal driver of the heat budget
for many streams. Brown's (1969) paper outlining a temperature prediction model
for small streams is widely cited, and is one of many energy budget models (LeBlanc
et al. 1997; Sinokrot and Stefan 1994; Sinokrot and Stefan 1993; Mosley 1983).
This particular model assumes that solar radiation is the primary source of heat
transfer to exposed streams, and the effect of air temperature is negligible (Brown
1969). For the particular streams surveyed in the coastal region of western Oregon,
with reach length of 2000 feet or less, this model worked well (Brown 1969).
LeBlanc et al. (1997) found in their study on Long Island, New York, that solar
radiation was the most important variable when attempting to determine predicted
water temperatures. This was then tied into the amount of vegetation present
(shading possibilities), with the statement that "water temperature is very sensitive to
stream shading" (LeBlanc et al. 1997). The benefits attributed to shade have been
stated frequently, with many studies relating the effects of vegetation removal to
increased stream temperatures (Dong et al. 1998; Brown and Binkley 1994; Lynch et
al. 1984; Feller 1981; Swift and Baker 1973; Swift and Messer 1971; Brown and
Krygier 1970; Brown 1969; Levno and Rothacher 1967). However, Levno and
Rothacher (1967) also stated that on heavily shaded streams, ambient air temperature
will be driving any temperature increase in the water, with their study showing
surface water temperature increases of up to 20F.
Air Temperature:
A number of studies show increases in stream temperatures primarily attributed to air
temperature alone. Needham and Jones' (1959) study in the Sierra Nevadas found
that air temperature had a significant impact on water temperatures in that location.
11
McRae and Edwards (1994), in their study of four streams in northern Wisconsin,
found that air temperature played the dominant role in determining the stream
thermal regime even when sites were located near sources of groundwater.
Conversely, the Webb and Nobilis' (1997) study in north central Austria found a
decreasing relationship between air and water temperatures as the impacts of
groundwater and shade became more apparent.
Air temperature is also particularly important in streams with low discharge
(Mohseni and Stefan 1999; Stefan an Preud'homme 1993; Smith and Lavis 1975);
Stefan and Preud'homme (1997) found an increasing linear relationship between
ambient air temperature and corresponding water temperature as discharge
decreased.
A number of studies (Mohseni and Stefan 1999; Mohseni et al. 1998; Stefan and
Preud'homme 1993) have looked at the mathematical relationship between air and
water temperatures. Stefan and Preud'homme (1993) concentrated on linear
equations to explain the relationship, focusing on temperatures between 0 and 200 C.
They achieved accuracy within 1°C on shaded low-flow streams, and within 2°C on
wider, higher-discharge streams. Mohseni et al. (1998) then developed a non-linear
model to predict stream temperatures from air temperatures. They discovered that
the relationship was better described by an S-shaped curve, with linearity between air
and water temperatures falling generally between 0 and 20°C, and sometimes as high
as 25°C. When they tested this model, they found 89% accuracy with 99%
confidence. Mohseni and Stefan (1999) explained that air and water temperatures
are non-linear (nearly flat) at temperatures <0°C because the presence of ice inhibits
heat transfer, and cooler groundwater temperatures dampen the response of the
stream to heat transfer. At higher temperatures (generally >20°C) the relationship is
also non-linear because evaporative cooling heavily counteracts the amount of heat
transfer to the stream (Mohseni and Stefan 1999).
12
It also appears that location of the study may be more important than previously
thought, as differing locales often produce much different results (eg: Webb and
Nobilis 1997; McRae and Edwards 1994; Vugts 1974). Ward's (1969) model, with
solar radiation as the primary heat source, was found to be accurate on short reaches
of streams in western Oregon, while Walker and Lawson (1977) in Australia had
good results using air temperature as the primary heat source in their predictive
model.
What seems most reasonable is that it is a combination of all heating variables, with
particular emphasis on solar inputs and ambient air temperatures that results in
temperature change in stream systems. Levno and Rothacher (1967) believed that
both of these variables were working together to alter temperatures in streams.
Sinokrot and Stefan (1994) corroborated this, with their sensitivity analysis showing
both air and solar radiation as the dominant factors. However, the amount of stream
exposed to direct sunlight can strongly influence which variable or factor dominates.
Levno and Rothacher (1967) found that under dense cover, air temperature tended to
dominate as the source of energy input to the stream.
CONCLUSION:
Each stream is subtly different from another, and it is important to be mindful of this
when attempting to determine what the most significant heating variable may be.
This study will allow us to compare the thermal regimes of two creeks in adjacent
drainages, one heavily forested, one not, and analyze the heating patterns of each. It
is hoped we can then determine the relationship between air temperature, in
combination with other variables, and the thermal regime of each of these creeks.
13
STUDY AREA
The study area is located in the Burnt River Drainage near Unity, in Baker County,
Eastern Oregon. The Burnt River Drainage encompasses approximately 2849 square
kilometers (Mangelson 2000), and supplies water for livestock, irrigation, and
recreation purposes. This study focuses on Barney and Stevens Creeks (Figure 2),
adjacent north-facing drainages that drain into the South Fork of the Burnt River.
Study site elevations range from 1370 meters to 2010 meters at the headwaters of
Stevens Creek, and from 1370 meters to 2192 meters at the headwaters of Barney
Creek.
There is a significant difference in the vegetation component of the two creeks.
Barney Creek was the site of a large wildfire in 1989, and has no mature standing
tree cover over its upper reaches. Shrub and tree cover increase at lower elevations
(below 1500 meters), but the stream is still exposed to direct solar radiation during
some part of the day. Trees at lower elevation are predominantly Ponderosa pine
(Pinusponderosa), shrubs include bearberry (Arctostaphylos uva-ursi), dogbane
(Apocynum spp.) with spiraea (Spiraea betul?folia) nearer the creek. Forbs and
grasses included monkey flower (Mimulus lewisii and Mimulus guttatus), strawberry
(Fragaria virginiana) and heart leaved arnica (Arnica cordfolia) with pine grass
(Calamagrostis rubescens), and some seeded Timothy (Phleum pratense) and
wheatgrasses (Agropyron spp.). Although immediately adjacent to the burned area,
Stevens Creek did not burn and is well vegetated along the entire length of its reach.
Its tree component consists mainly of Douglas fir (Pseudotsuga menziesii) with some
young larch (Larix spp.) in the understory and shrubs included spiraea (Spiraea
betulfolia). Forbs included prince's pine (Chimaphila umbellata) heart leaved
arnica (Arnica cordifolia) and some bunchberry (Cornus canadensis). Pine grass
(Calamagrostis rubescens) predominated the grass component.
14
Figure 2. Map of Study Area.
S
Legend
Station Location5
Stevens Creek
/'\/ Barney Creek
Label
De1iititioii/Eleation
Label
Definition/Elevation
STOI
STO2
STO3
STO4
Station . 370m
Station 2. I 525ni
Station 3. 1675m
Station 4, 1 830m
[3AMOiJTI-1
STHW
Headwaters of Stc cns ('r. 2u2 in
I3A3
Mouth of Barney Cr.
Station I. I 3Thrn
Stanon 2. I525m
First Side Channel.
Station 3. (i75rn
Second Side Channel. I 765m
BA)
BAu2
BSC
BAfl4
L3HW I
RH \\2
I
Station 4, I 83i1)iii
First Headwaters of Barney Cr 21
Second Hcadwaters of Barney Cr. 21
15
A study quantifying the amount of vegetative and topographic cover on both creeks
using Hemiview equipment determined that Barney Creek had approximately 20%
cover, while Stevens Creek had in excess of 80% cover (Krueger 2001, pers.
comm.).
Temperature data (in °F) taken from the Unity Ranger Station for the years 2000 and
2001, as well as historical averages spanning the period from 1948 to 2003 are
summarized in the following table:
Table 1. Maximum, minimum, and mean air temperature data, as well as historical
air temperatures for the Burnt River area.
Maximum (°F)
Minimum (°F)
Mean (°F)
July2000
91.3
44.2
68.2
August2000
93.1
43.6
68.6
July2001
90.6
43.6
67.3
August2001
94.5
45.3
70.1
Historical July
85.7
47.7
64.4
Historical August
84.8
45.2
63.4
As the data show, maximum air temperatures for the two years of the study were
generally much warmer than the historical average, as were mean air temperatures,
while minimum temperatures were generally one to two degrees Fahrenheit lower
than historical averages.
Precipitation data, also from the Unity Ranger Station, indicate that precipitation
levels were much lower than historical averages, with July and August of 2000
recording 0.00 and 0.73 inches respectively. July and August of 2001 recorded 0.28
and 0.00 inches respectively, compared to historical mean levels of 0.57 in July and
1.05 inches in August, with historical data spanning the time period from 1948 to
2003.
16
MATERIALS AND METHODS:
DATA COLLECTION
Air, soil, and water temperatures were recorded hourly during July and August of
2000 and 2001 using Stowaway ® XTI temperature loggers. Loggers were installed
at 1370, 1525, 1675, and 1830-rn elevations on both Barney and Stevens Creeks.
Air temperature loggers were enclosed in solar radiation shields and attached to the
north side of trees adjacent to the creek. Soil temperature loggers were enclosed in
submersible cases and buried 0.30 rn below ground, adjacent to the creek, in non-
saturated soils. Temperature loggers enclosed in submersible cases were placed in
the deepest free flowing water available at each site. Tributary water temperature
information was collected at 1766-rn and 1583-rn elevations on Barney Creek.
Comparisons made between tributaries and rnainstem indicated no difference in
temperatures. There were no tributaries on Stevens Creek.
Stream velocity and discharge for each site were estimated using a Pygmy® ST1
Flow Meter, at a depth of six-tenths of stream depth (measured from surface) and an
AquaCaic® 5000 (version AQCUG.7a). The Pygmy® meter measured flow, and the
AquaCaic® counted revolutions of the Pygmy® meter and calculated velocity and
discharge. As the 2000 season progressed, water levels in Stevens Creek became too
shallow to accurately collect flow data with the pygmy meter. Low runoff levels in
the spring of 2001 resulted in low flows on Stevens Creek, again resulting in the
pygmy meter being unable to record flow effectively. V-notch weirs were
constructed of 1.9 cm thick Plexiglas (Figure 3). The weirs were generally 0.60 m
wide (depending on the width of the creek at each station) and 0.45 m high, with a
90-degree angle v-notch in the center. These were placed in the creek at an
17
established cross section, leveled, and the edges and bottom dammed with plastic
sacking and mud. The water backed up behind the obstruction and flowed through
the 90-degree v-notch. When the water level equilibrated, a measurement using a
staff gauge was taken. This measurement was then converted to cubic feet per
second using a standard conversion Table (USBR 1997, p. A9). Further conversion
to cubic meters per second was done using the conversion factor of 0.028317; all
measurements were recorded in metric.
Figure 3: Diagram of 45° v-notch weir.
level
Reach distances between stations were calculated from Digital Ortho Photo Quad
(DOPQ) maps using the following methods. Station points were recorded with a
Trimble® Pro-XR (model TSC1) GPS unit and overlayed using ArcView Version
3.0 on to DOPQ maps (Scale: 1:24,000). The maps were obtained from the USGS.
Each creek was digitized as a separate overlay. ArcView 3.0 was then used to
calculate reach distances on the DOPQ maps.
18
DATA ANALYSIS
Stowaway® temperature data were recorded hourly and analyzed for July and
August 2000 and 2001. Data were imported into SAS (version 8.0), where each
reading was rounded to the nearest hour. Times from :00 to :29 were rounded back,
and times from :30 to :59 were rounded forward. Daily maximum, minimum, and
mean temperatures and times for each, as well as total net heating or cooling (per
day) for each parameter (air, water, and soil) were then calculated. Days when the
data loggers were downloaded were removed from the datasets, as we would not
have a complete data set for that day. Each month had a total of two days missing
except for July 2000, when there were three download days. The other exception
was Stevens Creek at 1370-rn elevation during August of 2001. This station dried up
on the 26th so there are seven days missing for that month.
Analysis of daily heating and cooling cycles was performed on air and water
parameters only, as diurnal fluctuations within the soil parameter were too small
(generally <1.25°C) to facilitate meaningful analysis. A full diurnal cycle was set
from Sam to Sam, with the heating portion consisting of 5am to 5pm, and the cooling
cycle consisting of 5pm to Sam. When temperatures from the previous day were
required for maximum water temperature analysis, we altered the daily cycle to read
from 5pm to 5pm. In this way we were able to obtain minimum temperatures from
the previous day, so that we could evaluate their influence on maximum water
temperatures the following day.
Each heating and cooling cycle was broken down into 2-hour increments, to further
pinpoint time of day when peak heating and cooling occurred, and to determine
whether periods within each cycle were the same. Peak heating change was
identified as the maximum temperature change recorded within a 2-hour period of
the heating cycle. Peak cooling change was identified as the maximum amount of
cooling during a 2-hour period of the cooling cycle. Time of peak (the 2-hour period
19
in which maximum heating or cooling occurred), temperature at peak, as well as
temperature change during the peak 2-hour period were also calculated.
Statistical tests were performed using the SAS (Version 8.0) software packages.
Analysis of Variance (ANOVA) was performed using PROCMIXED, with days as
blocking factor, so as to alleviate problems of serial correlation. PROCREG was
used for stepwise regressions. Statistical significance was set at the 0.05 level.
Objective 1: Determining The Influence of Elevation, MonthlYear, and Stream on
Thermal Environment.
ANOVA was run to partition the fixed effects of month, year, stream elevation, and
their interactions. By determining whether there was a month, year, stream, or
elevation effect we were able to structure analysis of subsequent hypotheses
accordingly. That is, stream, elevation, and month were analyzed separately for each
year, as these effects were generally found to be significant. A linear regression then
tested mean daily air, water, and soil temperatures as a function of elevation. All
R2s were adjusted for day-to-day variation, meaning that after day-to-day variation
was accounted for, the parameter 'elevation' explained X% of the remaining
variability.
Objective 2: Describing Diurnal Air and Water Cycles
ANOVA (PROCMTXED) with mean separation and days as blocking factor was
used to test whether each 2-hour period within the air heating and cooling cycle
yielded the same amount of temperature change. Heating and cooling cycles were
analyzed separately for each elevation, each month of each year, and each creek.
The same procedure was then followed to test whether 2-hour periods within the
20
heating and cooling cycle for water resulted in equal amounts of temperature change
Fisher's LSD was used to determine which periods were different from each other.
ANOVA (PROCMIXED) with days as blocking factor, was used to test temperature
change during the peak periods (described as the 2-hour period of greatest positive or
negative temperature change) for heating and cooling to determine whether the
amount of heating and cooling was the same at each elevation. Water and air heating
and cooling cycles were analyzed separately for each month and year, on each creek.
ANOVA (PROCMIXED) with days as blocking factor was used to analyze whether
timing of peak temperature change differed across elevations. Time of first peak was
defined as (the 2-hour period during which the greatest positive [for heating] and
negative for cooling] temperature change occurred. Fisher's LSD was used to
determine which elevations were different from each other. Water and air heating
and cooling cycles were analyzed separately for each month and year, on each creek.
Objective 3: Modeling Variables Strongly Associated with Maximum Water
Temperature.
Stepwise regression was performed to determine which variables were most strongly
related to maximum water temperature. Each elevation on each creek, for each
month and year were analyzed separately. "Current day" for this objective is defined
as the 24-hour period between 5pm and 5pm. "Previous day" includes data from
5am to 5pm of the day before the current day. Variables were chosen based on
examination of current water temperature literature, as well as variables of interest to
the author.
21
The pool of variables to choose from included:
> maximum air temperature from the current day (5pm-5pm) - the highest
temperature reached during the current day;
> air temperature change during the peak 2-hour heating period of the current
day (Spm-5pm) amount of air temperature change within the 2-hour period
that experienced the highest temperature of the current day;
> minimum air temperature from the previous day - the lowest air temperature
from the previous day;
> total air cooling from the previous day - the net amount of air temperature
change experienced during the cooling portion of the diurnal cycle (5pm5 am);
minimum water temperature from the previous day - the lowest water
temperature from the previous day;
total water heating from the previous day - the total amount of water
temperature change experienced during the heating portion of the diurnal
cycle (5am-5pm);
total water cooling from the previous day - the total amount of water
temperature change experienced during the cooling portion of the diurnal
cycle (5pm-5am);
In order to utilize data from the previous day, we altered the "day cycle" from the
normal 5am-Sam cycle, to Spm-5pm. This allowed us to have the minimum
temperature data included within a one-day cycle, as maximum water temperature
was always achieved prior to 5pm, as were maximum air temperature and peak air
heating.
22
Objective 4: Deriving a Predictive Equation for Water Temperature Change.
Stepwise regression using PROCREG with selection option R2, was performed to
determine variables for a predictive equation for water temperature change. We
chose a number of air and water characteristics from the pooi of data collected, from
which the SAS stepwise regression program could select the best subsets that would
predict the net daily change in water temperature. Variables were chosen based on
examination of current water temperature literature, as well as variables of interest to
the authors.
"Current day" for this objective is always defined as the 24-hour period between 5am
and 5am.
Variables for the main pool included:
> net air temperature change during the 24-hour period (5am-5am) of the
current day ("airday"); - how much the air temperature changed over the
course of the current day (5am 5am)
> maximum air temperature reached during the current day ("maxair");
> minimum air temperature reached during the current day ("minair");
> total net air temperature change during the 24-hour period of the previous day
("airday_1 ") - how much the air temperature changed over the previous day
(Sam-Sam)
> maximum air temperature reached during the day before - from 5am to Sam
of the previous day ("maxairl ")
> minimum air temperature reached during the day before - from Sam to 5am
of the previous day ("minairl ")
> maximum water temperature recorded during the current day ("maxwat")
> minimum water temperature recorded during the current day ("minwat")
23
maximum water temperature reached during the day before - from 5am to
5am of the previous day ("maxwatl")
> minimum water temperature reached during the day before - from 5am to
5am of the previous day ("minwati")
> water temperature change from the day before - from 5am to 5am of the
previous day ("watday 1") - how much the water temperature changed over
the course of the previous day
The diurnal cycle was still considered to run from 5am to 5am, however, data from
the previous day's cycle were also included so the important variables from the
previous day could also be selected. To select the best subset or variables to describe
daily net water temperature change, we balanced a Cp statistic of 4 with the
smallest BIC (Bayesian Information Criteria) statistic. The choice of a small BIC
statistic was used because there was a belief that the model involved some, but not
all of the variables. Using a Cp statistic as well, allowed us to account for the trade
off between bias from excluding potentially important variables, and the extra
variance from including too many variables. This was performed at each elevation
on each creek for each year. These subsets of variables were scrutinized and the five
most common among all elevations for both creeks and both years were selected.
Thus "maxair", "airday", "minair", as well as "minwat", and "watday" from the
previous day were retained and the others dropped. We then ran a multiple linear
regression to test the model and determine fit. Using cross validation, we determined
what the predicted value of a case would be when that case was left out of the model.
We then plotted actual versus predicted values, to determine accuracy of the model
in predicting various water temperature changes over a one-day period (Sam-Sam).
24
RESULTS AND DISCUSSION:
OBJECTIVE 1: TO DETERMiNE THE INFLUENCE OF ELEVATION,
MONTH/YEAR, AND STREAM ON THERMAL ENVIRONMENT.
Results:
Elevation:
Partitioning of data set variation by month, year, stream, and elevation determined
temperatures were strongly associated with elevation differences (Table 2).
Linear regressions were developed to determine the association between mean daily
air, water, and soil temperatures with elevation (Table 3). All model slopes indicate
that temperature (°C) values increase with every 150 m decrease in elevation.
Stevens Creek had a greater rate of temperature increase per 150 m decrease in
elevation than Barney Creek. Differences in the rate of temperature increase ranged
from 0.2-0.4°C for air temperatures, 0.9-1.0°C for water temperatures, and 1.0-1.5°C
for soil temperatures. These results are comparable to those found by Meays (2000).
Table 2. Partitioning the influence of the fixed effects of stream, elevation, year, and month, as well as all their interactions,
for each parameter.
WATER
Effect
Stream
Elevation
streambyelevation
Year
AIR
Denom DF F-value
Pr>F
Denom DF
763
6895.77
<.0001
763
763
<.0001
763
5836.04
467.50
120
yearbystream
yearbyelevation
yearbystreambyelevation
SOIL
Pr>F
Denom DF
<.0001
763
F-value
2731.90
2266.13
<.0001
<.0001
763
88.21
<.0001
10.56
0.0015
120
0.18
0.6686
763
0.02
0.8953
763
66.65
<.0001
763
1.11
0.3440
763
5.65
0.0008
758
763
3.99
0.0078
763
1.09
0.3512
758
Month
120
2.12
0.1479
120
4.30
120
month by stream
763
5.92
0.0152
763
0.28
0.0404
0.5947
758
monthbyelevation
763
21.47
<.0001
763
6.64
0.0002
758
month by stream by elevation
763
0.37
0.0090
763
4.67
0.0030
758
yearbymonth
120
0.62
0.5465
120
0.5771
0.62
0.4312
763
0.1185
0.0790
0.31
763
2.47
3.09
120
year by month by stream
758
60.75
<.0001
yearbymonthbyelevation
yearbymonthbystreambyelevation
763
1.23
0.2980
763
2.26
0.0802
758
1.20
0.3073
763
0.44
0.7223
763
0.42
0.7407
758
0.70
0.5511
WATER
Variance Components
(Error)
Parameter
Day (year by month)
Residual
AIR
Variance Components
(Error)
Estimate
2.1064
0.2234
Parameter
Day (year by month)
Residual
SOIL
Variance Components
(Error)
Estimate
10.8245
0.1401
Parameter
Day (year by month)
Residual
Estimate
.8499
.0541
Pr>F
758
F-value
200021.80
<.0001
758
11644.90
<.0001
758
7917.33
<.0001
120
0.56
0.4541
758
12.69
0.0004
68.89
71.69
<.0001
24.37
70.52
135.37
44.05
<.0001
<.0001
<.0001
<.0001
<.0001
26
Table 3. Regression coefficient estimates for elevation in°Celsius/1 50 m elevation
decrease.
Barney
Stevens
2000
(July)
2000
(July)
'slope
water
air
soil
1.46
0.66
0.73
SE
0.03
0.05
0.09
p-value
<.0001
<.0001
<.0001
R2*
0.97
0.71
0.42
Barney
'slope
water
1.29
air
soil
0.74
0.85
SE
0.03
0.05
0.08
p-value
<.0001
<.0001
<.0001
R2*
0.95
0.71
0.54
1.03
1.70
'slope
water
air
soil
Barney
Stevens
2001
2001
(July)
SE
0.06
0.04
0.11
p-value
<.0001
<.0001
<.0001
R2*
0.95
0.91
0.73
2.11
0.96
2.11
R2*
SE
0.05
0.04
0.11
p-value
<.0001
SE
0.06
0.03
0.10
p-value
<.0001
<.0001
<.0001
0.95
0.96
0.79
SE
0.06
0.03
0.09
p-value
<.0001
<.0001
<.0001
0.95
0.92
0.86
<.0001
<.0001
0.95
0.89
0.82
(July)
'slope
water
air
soil
2.36
Stevens
2000
(August)
2000
(August)
1.40
0.70
0.38
SE
0.03
0.04
0.09
p-value
R2*
<.0001
<.0001
<.0001
0.95
0.76
0.20
Barney
'slope
water
air
soil
2.42
1.12
1.84
R2*
Stevens
2001
2001
(August)
(August)
'slope
1
'slope
water
air
soil
water
1.29
air
soil
0.67
0.51
SE
0.03
0.05
0.08
p-value
<.0001
<.0001
<.0001
R2*
0.95
0.69
0.32
'slope
water
air
2.21
soil
2.01
0.97
R2*
Slope = temperature change in °C/150 m decrease in elevation
= proportion of remaining variation (after accounting for day-to-day variation)
accounted for by elevation, ie: elevations accounts for X% of the remaining
variation.
27
Month:
Month was not a significant source of variation for water temperature (Table 1).
Month x stream and month x elevation were both significant. Mean monthly air and
soil temperatures were different. Mean monthly temperatures were greater in August
than in July for both parameters over both years and across both creeks (Figures 2
and 3). The month x elevation interaction was significant for air and soil as well as
the month x stream interaction for soil.
Year:
Results from the analysis of variance (Table 1) indicated that yearly differences were
not significant for air, but air in combination with stream, and air in combination
with elevation were highly significant. Soil followed a similar pattern; year was not
significant alone, but the interaction of year and stream, and year and elevation were
statistically significant. The opposite pattern emerged for water: yearly differences
were statistically significant, but not the interaction of year and stream or year and
elevation. Figures 4 and 5 summarize the mean temperatures for air, water, and soil
across months and years for Barney and Stevens Creeks.
Figure 4. Mean air, water, and soil temperatures by month and years for Barney Creek, 2000-2001.
Mean temperatures, Barney Creek 1525m
Mean temperatures, Barney Creek 137Gm
2000-2001
2000-2001
20 -
20
18 -
18
16
(I)
14
o rrean water temperature
12
0
C) rrC)an soil temperature
a)
10
C)
C)
0)
C)
0 rrwari water temperature
o sean air temperature
10
o swan soil temperature
-
C)
6
O
C)
O rrean air temperature
161412-
6-
o 4a,
4
2Ju y 2000
August 2000
Ju y 2001
July 2000
August 2001
August 2000
2000-2001
2000-2001
C)
0
August 2001
Mean temperatures, Barney Creek 183Gm
Mean temperatures, Barney Creek 1675m
(0
July 2001
Month and Year
Month and Year
20
18
18 -
16
1614-
o rrean water terrperature
12
W
0 rrean air temperature
10
rrwan soil temperature
0)6
141210-
O mean water temperature
0 sean air temperature
(0
G)
-
C,
6-
C) mean soil temperature
C)
C)
0
4
-
2
2
July 2000
August 2000
July 2001
Month and Year
August 2001
July 2000
August 2000
July 2001
August 2001
Month and Year
00
Figure 5. Mean air, water, and soil temperatures by month and years for Stevens Creek, 2000-2001.
Mean temperatures, Stevens Creek 1370m
Mean temperatures, Stevens Creek 1525m
2000-2001
2000-2001
20 -
18 -
18 -
16
1614a' 12ta
0
10
a'
(a
14
El rrwan water terijerature
(a
12 -
tO swan water teriperature
0 rman air teriperature
0
10
o swan air terrperature
sean soil tenperature
a'
-
6-
a'
04
0
(a
a'
a'
0)
a)
0
20I
8642-
sean soil terrperature
0
July 2000
August 2000
July 2001
August 2001
July 2000
Month and Year
August 2000
July 2001
August 2001
Month and Year
Mean temperatures, Stevens Creek 1675m
Mean temperatures, Stevens Creek 1830m
2000-2001
2000-2001
18
16
16
14 -
14
12-
(a 12
O sean water terrperature
(a
(J 10
o sean air terrperature
0
(a
a'
swan soil terTgjerature
a,
0)
(a
a'
a'
0)
a'
a)
0
0
2
o sean water terrperature
a' 10-
o swan air teriperature
86-
swan soil terrperature
42-
0
July 2000
August 2000
July 2001
Month and Year
August 2001
July 2000
August 2000
July 2001
Month and Year
August 2001
30
Stream:
Results from the analysis of variance (Table 1) indicated that stream differences
were significant for each of the three parameters. In all cases, mean air temperatures
across elevations on Barney Creek were greater than mean air temperatures on
Stevens Creek by 1-2°C. Mean air temperatures always exceeded mean water
temperatures on both creeks by 1.5-2°C across all elevations, with a general trend of
larger differences occurring at higher elevations and smaller differences at lower
elevations. A notable difference between streams was evident at the upper three
elevations of Stevens Creek, where mean air temperatures exceeded mean water and
soil temperatures by a wider margin than found on Barney Creek (Figure 6). Mean
water temperatures were greater on Barney Creek than at corresponding elevations of
Stevens Creek, by 1-4°C. The difference between Barney and Stevens Creek mean
air and water temperatures decreased in the downstream direction. Mean soil
temperatures on Barney Creek followed the same pattern as air and water, exceeding
soil temperatures on corresponding elevations of Stevens Creek by 3-4°C, with the
exception of 1370-rn. In this case, the mean soil temperatures on Stevens Creek
exceeded those at the same elevation of Barney Creek by approximately 0.5-1°C.
Differences between the two creeks' mean soil temperatures were fairly constant at
all elevations. Mean soil temperatures exceeded mean water temperatures across all
elevations on Stevens Creek, and this was also the case for Barney Creek, with the
exception of 1370-rn elevation, where mean water temperatures were greater than
mean soil temperatures.
Figure 6. Mean air, water, and soil temperatures by elevations for each stream,
month, and year.
Mean air temperature
19 -
a--- Barney July 2000
18
Barney August 2000
-A-- Barney July 2001
-
Barney August 2001
-.--- Stevens July 2000
Stevens August 2000
-.-- Stevens July 2001
-.-- Stevens August 2001
1413
12
1830
1675
1525
1370
6evation
Mean water temperature
17
e--- Barney July 2000
16
Barney August 2000
15
a-- Barney July 2001
14
Barney August 2001
-.-- Stevens July 2000
Stevens August 2000
-.--- Stevens July 2001
C)
10
Stevens August 2001
9
8
7
1830
1675
1525
1370
Elevation
Mean soil temperature
a-- Barney July 2000
-
0
Barney August 2000
a-- Barney July 2001
a--- Barney August 2001
-.- Stevens July 2000
14 -
Stevens August 2000
-.--- Stevens July 2001
-.--- Stevens August 2001
1830
1675
1525
Elevation
1370
32
Discharge, Velocity, and Area
Table 4 summarizes discharge, velocity, and cross-sectional area at permanent
locations along both creeks. Calculations are based on 7 to 14 measurements
(measurement number varies with stream width) taken across the permanent cross-
section locations. Discharge on Barney Creek ranged from 1.8 to 4.6 times greater
than Stevens Creek in 2000. Lower flows occurred on both streams in 2001 and
Barney Creek discharge was 5.5 to 11 times greater than Stevens Creek. Stevens
Creek discharge in 2000 was 2.8 to 7.6 times larger than in 2001. Barney Creek
discharge was 1.2 to 2.3 times greater in 2000 than 2001. Velocities decreased in
2000 and 2001 as the season progressed from mid July to late August. Barney Creek
was generally wider than Stevens Creek, with the greatest difference in widths seen
at the lowest elevations.
33
Table 4. Discharge, velocity, and area for both creeks.
2000
Stevens 1370-rn
Date
Discharge
(m'/s)
Stevens 1525-rn
2
Area(m)
Stevens 1675-rn
Stevens 1830-rn
Velocity
Discharge
Area
Velocity
Discharge
Area
Velocity
Discharge
Area
Velocity
(m/s)
(ms/s)
(m)
(mis)
(oils)
(m)
(mis)
(mIs)
(m-)
(oils)
7/15/00
0.0084
0.04
0.2100
0.0149
0.07
0.2129
0.0107
0.05
0.2140
0.0038
0.06
0.0633
7/28/00
0.0112
0.09
0.1244
0.0078
0.05
0.1560
0.0099
0.04
0.2475
0.0021
0.05
0.0420
8/09/00
0.0029
0.03
0.0967
0.0063
0.06
0.1050
0.0077
0.03
0.2567
0.0017
0.04
0.0425
0.0048
0.05
0.0960
8/21/00
2000
Date
insufficient flow
Barney 1370-rn
Discharge
Area (or)
(m3/s)
insufficient flow
Barney 1675-rn
Barney 1525-rn
insufficient flow
Barney 1830-rn
Velocity Discharge
Area
Velocity
Discharge
Area
Velocity
Discharge
Area
Velocity
(oils)
(nr)
(oils)
(m3/s)
(m)
(oi/s)
(nT/s)
(m)
(oi/s)
(m3ls)
7/13/00
0.0669
0.19
0.3521
0.0368
0.13
0.2831
0.0318
0.10
0.3180
0.0115
0.05
7/29/00
0.0600
0.16
0.3750
0.0328
0.13
0.2523
0.0268
0.09
0.2978
0.0074
0.03
0.2467
8/10/00
0.0231
0.09
0.2567
0.0238
0.10
0.2380
0.0194
0.07
0.2771
0.0060
0.03
0.2000
8/22/00
0.0299
0.13
0.2300
0.0143
0.07
0.2043
0.0361
0.16
0.2256
0.0029
0.01
0.2900
2001
Date
Stevens 1370-rn
Discharge
(m'/s)
Area (ni)
Stevens 1675-rn
Stevens 1525-rn
Velocity Discharge
(mis)
(mis)
0.2300
Stevens 1830-rn
Area
Velocity
Discharge
Area
Velocity
Discharge
Area
Velocity
(nr)
(m's)
(mis)
(m)
(m's)
(m7s)
(m2)
(m/s)
7/13/01
0.0033
0.08
0.0407
0.0052
0.04
0.1377
0.0030
0.12
0.0259
7/26/01
0.0020
0.07
0.0305
0.0039
0.03
0.1132
0.0023
0.11
<0.0013
0.06
<0.0205
0,0202
<0.0243
<0,0013
0.05
8/15/01
insufficient flow
0.0018
0,02
0,0783
<0.0013
0.08
<.0160
<0.0013
0.04
<0.0308
8/27/01
dry
<0.0013
0.01
<0.1327
<0.0013
0.12
<.0105
<0.0013
0.04
<0.0316
2001
Date
Barney 1370-rn
Discharge
(mis)
Area (or)
Barney 1525-rn
Barney 1675-rn
Barney 1830 rn
Velocity Discharge
Area
Velocity
Discharge
Area
Velocity
Discharge
Area
Velocity
(oi/s)
(rn')
(oils)
(mis)
(m)
(oils)
(m3/s)
(rn')
(m/s)
(m3/s)
7/05/01
0.0423
0.24
0.1763
0.0298
0,15
0.1987
0.0326
0.11
0.2964
0.0105
0.05
0,2100
7/14/01
0.0388
0.27
0.1437
0.0375
0.17
0.2206
0.0300
0.12
0.2500
0.0079
0.05
0,1580
7/27/01
0.0224
0.24
- 0,0933
0.0320
0,16
0.2000
0.0234
0.10
0.2340
0.0056
0.03
0.1867
8/16/01
0.0198
0.24
0.0825
eqpt failure
0.0127
0.08
0.1588
0.0042
0.03
0,1400
8/28/01
0.0102
0.07
0.1457
insufficientflow
0.0194
0.08
0.2425
insufficientflow
note: velocity calculated using the equation V=Q/A, with Q=discharge, and A=area
Distances, Gradients, and Exposure Times:
Table 5 summarizes the distance within each stream section, the gradient associated
with each section, and approximate exposure time for the water based on the stream
velocities presented in the previous table. Overall, Barney Creek had a longer total
distance than Stevens Creek. Section distances between the upper three elevations
34
were shorter for Barney Creek and much longer between the two lower elevations.
However, approximate exposure time for Barney Creek remained shorter than
Stevens Creek throughout the study due to the greater velocities associated with
larger stream discharge. In 2000, exposure times on Barney Creek ranged from 67
minutes to 244 minutes (approximately 1 to 4 hours). Slightly longer exposure
times (due to lower discharge and velocities) occurred in 2001, ranging from 93
minutes to 251 minutes (approximately 1.5 to 4.5 hours). In contrast, Stevens Creek
2000 exposure times varied from 148 minutes to 514 minutes (approximately 3 to 9
hours). The much lower discharge in Stevens Creek in 2001, coupled with minimal
velocities, had a profound impact on exposure times. They ranged from 324 minutes
to greater than 3624 minutes (approximately 5.5 hours to 2.5 days). All exposure
times are considered to be a rough approximation since they are based on 'typical'
stream cross-sections and single point-in-time velocity measurements. However,
they do offer some insight into the exposure times that occurred during the study.
35
Table 5. Distances, gradients, and exposure time for water on both creeks.
2000
Date
S1370-S1525
Distance (us)
S1525-S1675
Gradient
Exposure
(%)
Time1 (mm)
Distance (us)
S1675-S1830
Gradient
Exposure
(%)
Time1 (rain)
Distance (m)
Gradient
Exposure
(%)
Time' (mm)
7/15/00
2677
5.60
210
2285
6.56
178
1296
11.57
341
7/28/00
2677
5.60
286
2285
6.56
154
1296
11.57
514
8/09/00
2677
5.60
425
2285
6.56
148
1296
11.57
508
8/25/00
2677
5.60
465
2285
6.56
too shallow
1296
11.57
too shallow
2000
Date
B1370-B1525
Distance (m)
B1525-R1675
Gradient
Exposure
(%)
Tinie' (mm)
Distance (m)
B1675-111830
Gradient
Exposure
(%)
Time1(min)
Distance (rn)
Gradient
Exposure
(%)
Time' (mm)
7/15/00
2995
5.01
176
2155
6.96
t76
1168
12.84
85
7/29/00
2995
5.01
198
2155
6.96
198
1168
12.84
79
8/10(00
2995
5.01
210
2155
6.96
210
1168
12.84
97
8/22/00
2995
5.01
244
2155
6.96
244
1168
12.84
67
Gradient
Exposure
(%)
Time1 (mm)
2001
Dale
S1370-S1525
Distance (us)
S1525-S1675
Gradient
Exposure
(%)
Time' (miii)
Distance (m)
S1675-S1830
Distance (m)
Gradient
Exposure
(%)
Time' (mm)
7/t3/01
2677
5.60
324
2285
6.56
1472
1296
11.57
>1054
7/26/01
2677
5.60
394
2285
6.56
1888
1296
11.57
>889
8/15/01
2677
5.60
570
2285
6.56
>3627
1296
11.57
>701
8/27/01
2677
5.60
>336
2285
6.56
>2380
1296
11.57
>684
Gradient
Exposure
(%)
Tim& (mm)
Distance (m)
2001
Date
B1370-B1525
Distance (m)
81675-B1830
131525-B1675
Gradient
Exposure
(%)
Time' (miii)
Distance (us)
Gradient
Exposure
(%)
Time' (rain)
7/05/01
2995
5.01
251
2155
6.96
121
tt68
12.84
93
7/14/01
2995
5.01
226
2155
6.96
144
1168
12.84
123
7/27/01
2995
5.01
250
2155
6.96
153
1168
12.84
104
8/16(01
2995
5.01
eqpt failure
2155
6.96
226
1168
12.84
139
8/28/01
2995
5.01
too shattow
2155
6.96
148
1168
12.84
toO shallow
Exposure time was calculated by dividing distance between stations by the velocity
recorded at the higher elevation station, and then dividing the result by 60 to obtain
minutes.
Tributaries and Vegetation D?fferences:
The Stevens Creek drainage contained no tributaries. Barney Creek had two
tributaries. Discharge from the Barney Creek tributary located at 1 768-m elevation
was roughly equal to main stem discharge. The second Barney Creek tributary,
located at 1530-rn elevation, had approximately 1/3 the discharge of the main stem.
36
No measurable increase in main stem temperature could be attributed to either
Barney Creek tributary. This result was in agreement with results observed by
Meays (2000).
Hemiview data gathered during the 2001 field season determined that view-to-sky
differed dramatically on the two creeks. Generally, Stevens Creek had less than 20%
view-to-sky due to topographic and vegetation obstruction. Barney Creek had 80%
or more view-to-sky with minimal topographic or vegetation obstruction (Krueger
2001, unpublished data [pers. comm j)
Discussion:
Elevation has been associated with the thermal patterns of the environment and
stream temperature (Larson and Larson 1997; McRae and Edwards 1994; Meisner et
al.1988; Ward 1985; Raphael 1962). A portion of the increased air temperature
associated with decreasing elevation can be associated with the adiabatic process
where atmospheric compression always results in warming (Ahrens 1993). The air
density decreases associated with increasing elevation, decrease the effect of the
atmosphere (absorption and re-radiation of long-wave radiation) and modify thermal
patterns at the earth's surface (Hidore and Oliver 1993). As a result, air and ground
temperatures are generally cooler at higher elevations and warmer at lower
elevations. Similarly, stream temperatures have also been shown to increase
naturally as water flows downstream through a warming environment with a greater
increase seen in creeks with lower discharge (Zweinecki and Newton 1999). The
values and patterns in our data were consistent with those found by Meays (Meays
2000).
37
Stevens Creek drainage generally showed greater increases in temperature per 150 m
decrease in elevation than that found in Barney Creek drainage, for all parameters.
LeBlanc et al. (1997) indicate that the abundance and type of vegetation cover will
impact the pattern of energy retention within a watershed by modifying the pattern of
radiation absorption and emission. Given the observed differences in vegetation
structure found between Stevens and Barney Creek drainages one should anticipate
that as temperature patterns migrate toward a thermal equilibrium that differences
within the drainages would become apparent. Stream temperature is also influenced
by exposure time, the time that a parcel of water is exposed to a thermal environment
before leaving an area. In July 2000 and 2001, the approximate exposure time for
Barney Creek was 50% and 20% of exposure time observed on Stevens Creek.
When exposure time is taken into account, Barney and Stevens Creek were warming
0.5°C/h in July 2000 and 0.4°C/h in July 2001. Exposure time likely influenced
August 2000 and 2001 stream temperature patterns. However, exposure times in
those months exceed the length of the daily heating cycle, making a direct
comparison inappropriate.
The findings that monthly mean temperatures were different for air infers that
weather patterns differ from month to month. Similarly, we see a difference in mean
monthly soil temperatures, but not in mean monthly water temperatures. The
interaction of month and stream was significant for water and soil, but not for air.
This indicates that mean monthly weather patterns were behaving similarly on both
streams, that is, the streams were experiencing the same weather patterns. This is
understandable, given that both creeks share the same aspect, orientation, and
watershed. Mean monthly water and soil temperature patterns were different on each
creek, likely attributable to stream characteristics and vegetation structure
differences found within each drainage.
38
Yearly mean air temperatures were not found to be different, and the month and year
interaction was insignificant. Weather patterns were similar both years and months
in the sense that August air temperatures were warmer than July air temperatures in
both years. Response of water and soil showed similar monthly patterns each year
(August temperatures generally wanner than July temperatures); however mean
water temperatures were different year to year. A portion of the water temperature
variation is likely attributable to differences in discharge and exposure time. While
both streams shared the same general geographical attributes (aspect, orientation, and
locale), they were very different in their discharge, depth, and cover characteristics.
These are the likely reasons that the fixed effect of stream was significant.
Mean air temperatures on Barney Creek were greater than those on Stevens Creek.
This is likely due to drainage wide differences in the drainage basin and vegetation
structure, with Barney Creek being more exposed than Stevens Creek. However, the
rate of air heating with decreasing elevation was similar between the two creeks.
Mean water temperatures on Barney Creek exceeded those at corresponding
elevations on Stevens Creek. Given that mean air temperatures were higher on
Barney Creek, and findings that water temperatures tend to follow air temperatures
closely (Sinokrot and Stefan 1993; Stefan and Preud'homme 1993; Needham and
Jones 1959), it stands to reason that mean water temperatures on Barney Creek
would exceed those found on Stevens Creek. It is of interest to note that the rate of
temperature change per 150 m drop in elevation was greater on Stevens Creek
between 1830-1675-rn elevations and between 1525-1370-rn elevations than on
Barney Creek. This can be attributed to differences in discharge, exposure times,
and temperature differences within the thermal environment. Discharge on Stevens
Creek was generally at most half that of Barney Creek. Streams with lower flows
tend to respond more rapidly to increases in temperature (Le Blanc et al. 1997;
Adams and Sullivan 1989; Beschta and Taylor 1988; Mosley 1983; Smith and Lavis
39
1974; Swift and Messer 1971). Exposure times on Stevens Creek, while rough
approximations, indicated that water parcels on Stevens Creek were exposed to the
thermal environment within individual stream sections for longer periods of time
than equivalent water parcels on Barney Creek. In addition, temperature differences
between air and water were greater on Stevens Creek than on Barney Creek (see
Appendix 1). Raphael (1962) noted that greater temperature differences between the
air and the water (with air being warmer than water) were associated with greater
water heating. It stands to reason then that the increased exposure time and
temperature gradient between air and water would be associated with greater water
temperature increases for Stevens Creek.
Mean soil temperatures on Barney Creek generally exceeded those found at the same
elevations on Stevens Creek, with the exception of 1 370-rn. Here, Barney Creek
actually dropped in temperature between 1525 and 13 70-rn, while Stevens Creek
increased dramatically (by 4°C or greater) between these two elevations. The decline
in mean soil temperature seen at Barney 1370-rn elevation may be attributable to the
presence of additional vegetation structure in the lower portion of the drainage. The
higher mean soil temperatures seen on the upper portion of the Barney Creek
drainage may be a result of greater exposure to direct solar insolation particularly at
1525-rn, where the bowl-shaped topography would serve to concentrate heat. It has
been shown that soil temperatures in non-shaded areas tend to be greater than those
areas having vegetative cover (Pluhowski and Kantrowitz 1963). Vegetative cover,
coupled with the limited air movement common in these areas serves to limit energy
inputs to the ground surface (Fowler et al. 1979). Soil temperatures generally were
lower than air temperatures at corresponding elevations, except at Stevens Creek
1370-rn, where soil temperatures actually exceeded air temperatures by
approximately 0.5-1°C.
40
OBJECTiVE 2: DESCRIBING DIIJRNAL AIR AND WATER CYCLES.
Results:
Air Temperature Patterns:
Daily heating and cooling patterns were divided into 12-hour cycles containing 6
two-hour periods for analysis. The analysis for elevation and month indicated that 2hour periods within the heating and cooling cycles for air were different (p<O.0001)
on each creek. Figure7 shows a typical 24-hour cycle for air temperature, broken
into 2-hour periods, with the corresponding average hourly air temperature at the
beginning of each 2-hour period overlaid as a line graph.
Heating Cycle:
The four elevations on Barney and Stevens Creek had differing amounts of peak
temperature change in each month studied (p<O.0001). Figures 8 and 9 show the 2hour period temperature changes for air in July of 2000 for each creek (graphs of
results for both creeks and years are contained within Appendix 2). On Barney
Creek the peak period of heating occurred between 7-9am with decreasing amounts
of heat accumulation occurring over the remainder of the 12-hour heating period.
The pattern was consistent regardless of month or elevation. Air heating was
greatest at the 1675-rn site where temperature increases were 1°C or more higher
than all other sites.
On Stevens Creek the peak period of heating occurred one period later (9-1 1am) than
on Barney Creek and the greatest heating occurred at the lowest elevation.
Differences observed between Barney and Stevens Creeks appeared to be related to
differences in air movement and surface heating. The open exposure of the Barney
Creek basin favored a rapid warming of the ground surface, concentration of direct
Figure 7. 2-hour period air temperature changes with average hourly air temperature at beginning of each 2-hour period,
for 1370-rn elevation Stevens Creek July 2000.
Peak period temperature changes, and
corresponding average hourly air temperatures for
Stevens Creek at 1370-rn in July 2000.
r
30.00 -
i
25.00 U)
C)
I-U)
Peak air temperature change per
2-hour period
Average hourly air temperature at
beginning of each 2-hour period
20.00 15.00 -
C)
Cu
C,
0
U)
10.00
a)
E a)
C)
V
5.00 0.00
-5.00
F-
-10.00 -
14)
'-'0)
c)
'-
-
N-
-
0)
C)
'4)
r-
'4)
-
Time of day
-
('1
th
c)
-
'4)
c)
42
solar radiation near the 1675-rn site during summer, and a basin-wide pattern of air
circulation that favored afternoon and evening breezes. By contrast, the forest
canopy of Stevens Creek modified basin heating patterns by dampening surface air
circulation patterns, elevating solar interception from the ground surface into the tree
canopy, and providing different absorption characteristics. These differences
appeared to cause peak heating to occur later in the day on Stevens Creek and the
amount of heating within the basin tended to increase with decreasing elevation.
Figure 8. 2-hour period temperature changes for air, Barney Creek July 2000. Temperature changes without a common letter
(black) are significantly different (p<O.05). Mean amount of peak heating/cooling without a common letter (starred, blue) are
significantly different (p<O.O5) across elevations.
Barney Creek 1525m, July 2000
Barney Creek 1370m, July 2000
12
Mean amount of peak, heating: 7.46°C A*
10
a)
(
6
Q
4
.-,
A
8
)
a)
a)
2
(I)
a)
a)
C)
a)
(0
0)
-6
-8
Mean amount of peak, cooling: -7.45 °C E*
6
0
-4
.
C
0
-2
Mean amount of peak, heating: 10.59°C B*
8
Mean amount of peak, cooling: -6.89°C E F*
B
A
10
LII
C,
.21
12
(C
F
E
F
E
F
B
4
2
0
-2
(7,
(0
C.)
E
period (time of day)
a)
a)
2
Wa)
U
(P
(0
(0
-6
-8
-10
(0
0)
0
E
E
F
0
0
Mean amount of peak, cooling: -6.38°C F*
B
4
LC)
0
a--C
-2
.21
-4
L
Mean amount of peak, heating: 7.90°C A*
A
8
C5
C
(7,
-4
C)
C
B
4
U)
.21
10
Mean amount of peak, cooling: -7.61 °C F.*
O
.-.
Barney Creek 1830m, July 2000
12
Mean amount of peak. heating: 10.27°C B*
6
6
-8
C
C
(0
C.)
E
G
F
G
-10
period (time of day)
E
G
Barney Creek 1675m, July 2000
C)
E
F
-8
-10
H
A
10
c
0
period (time of day)
a)
B
-6
G
-10
12
C
ID
-4
E
B
period (time of day)
E
E
Figure 9. 2-hour period temperature changes for air, Stevens Creek July 2000. Temperature changes without a common letter
(black) are significantly different (p<O.05). Mean amount of peak heating/cooling without a common letter (starred, blue) are
significantly different (p<O.05) across elevations.
Stevens Creek 1370m, July 2000
12
A
Stevens Creek 1525m, July 2000
12
Mean amount of peak, heating: 10.43°C A*
10
a)
Ca
2
.-
B
B
=
(C'
-4
4-
,)
(C'
-4
Ca
c
o
C.)
(0
7'
O
PS)
-4
PS)
PS)
,)
Mean amount of peak, cooling: -4.78°C F*
B
6
G
E
E
2
a)
p
B
C
LIE]
0
-2
G
-
-1
(C'
(C'
-4
E
F
E
H
Stevens Creek 183Gm, July 2000
12
Mean amount of peak, heating: 6.98°C B*
10
a)
Mean amount of peak, cooling: -5.48°C FC*
A
0)
a)
0)
8
(a
6
C
A
6
B
4
-c
C
C
0
D
(7,
--4
(i.)
(C'
C.)
(7
;7'
(0
-4
PS)
Co
G
-8
0) 0
-2
-4
E
G
-
L;I
F
E
E
(a
Mean amount of peak, cooling: -5.85°C EC*
A
B
B
4
2
LiE]
0
-6
H
G
period (time of day)
Mean amount of peak, heating: 6.75°C B*
10
Ca
(f.)
(C'
F
Stevens Creek 1675m, July 2000
-4
--4
-10
12
-2
'
(7
G
period (time of day)
2
(i.)
-8
-10
ca
C
D
-6
F
F
-8
p
C.)
F
-6
A
U0 4
C
D
E
0
-2
0
C
o
Mean amount of peak, cooling: -6.32°C E*
6
4
Mean amount of peak, heating: 7.49°C B*
10
8
-6
C
C
C
(0
(0
((5)
(5)
(7'
(0
((7'
-4
D
F
F
-8
-10
-10
period (time of day)
period (time of day)
E
D
D
45
Tables 6 and 7 illustrate the timing and frequency of peak heating for air on Barney
Creek and Stevens Creek across both years. LSD significance indicates that timing
of peak heating was different across certain elevations within a month, unless
otherwise stated. If not statistically significant (ie: p-value not less than 0.05), pvalues are given.
46
Table 6. Frequencies of when peak air heating occurred across elevations on Barney
Creek, both years.
July
2000
1370-rn
1525-rn
frequency
Barney
time
(%)
1675-rn
frequency
time
(%)
time
92.86
3.57
7-9
8571
9-11
64.29
25.0
5-7
7-9
9-11
11-13
10.71
11-13
7.14
7.14
11-13
7-9
A
1830-rn
frequency
AB
frequency
time
(%)
3.57
BC
..
5-7
7-9
(%)
35.71
60.71
13-15
3.57
C
.
.::1
,.
August
2000
1370-rn
1525-rn
frequency
Barney
time
7-9
(%)
9-11
72.41
17.20
11-13
10.34
A
1675-rn
frequency
time
7-9
9-11
(%)
65.52
34.48
A
1830-rn
frequency
time
7-9
frequency
(%)
time
(%)
100.00
5-7
7-9
3.45
93.10
3.45
9-11
B
B
July
2001
1370-rn
1525-rn
frequency
Barney
tirne
7-9
9-11
11-13
13-15
1
(%)
1675-rn
frequency
time
(%)
time
5-7
7-9
77.78
7-9
88.89
7.41
7.41
7.41
11-13
13-15
3.70
A
9-11
7.41
A
1830-rn
frequency
11-13
15-17
(%)
3.70
85.19
3.70
3.70
3.70
frequency
time
5-7
7-9
(%)
3.70
85.19
9-11
7.41
13-15
3.70
A
A
August
2001
1370-rn
1525-rn
frequency
Barney
time
7-9
9-11
1
A
(%)
62.96
37.04
1675-rn
frequency
time
1830-rn
frequency
frequency
(%)
time
(%)
time
(%)
7-9
92.59
5-7
7-9
5-7
7-9
9-11
7.41
9-11
3.70
92.59
3.70
3.70
92.59
3.70
B
B
9-11
B
note: letters different from each other indicate difference in time of first peak across
elevations within a month, not across months (p<O.O5) (time of first peak is defined
as the first time in the cycle where the maximum amount of heating was recorded)
47
Table 7. Frequencies of when peak air heating occurred across elevations on Stevens
Creek, both years.
July
2000
1370-rn
1525-rn
frequency
Stevens
time
7-9
9-11
11-13
(%)
28.57
64.29
3.57
1675-rn
frequency
time
7-9
9-11
11-13
(%)
time
17.86
75.00
7.14
5-7
7-9
9-11
11-13
A
A
1830-rn
frequency
(%)
3.57
25.00
64.29
7.14
A
.. .)
.
...
frequency
time
5-7
7-9
9-11
11-13
13-15
A
..
(%)
3.57
14.29
75.00
3.57
3.57
.
August
2000
1370-rn
1525-rn
Stevens
time
7-9
9-11
1
(%)
20.69
79.31
A
1675-rn
frequency
frequency
time
7-9
9-11
11-13
(%)
10.34
86.21
3.45
1830-rn
frequency
time
7-9
9-11
11-13
(%)
20.69
75.86
3.45
A
AB
frequency
time
(%)
9-11
11-13
B
93.10
6.90
July
2001
1370-rn
1525-rn
frequency
Stevens
time
7-9
9-11
11-13
13-15
(%)
22.22
62.96
1675-rn
frequency
time
(%)
7-9
9-11
29.63
62.96
13-15
15-17
3.70
3.70
11.11
3.70
A
AB
1830-rn
frequency
frequency
time
5-7
7-9
9-11
11-13
13-15
(%)
3.70
time
(%)
48.15
29.63
3.70
7-9
9-11
3.70
74.07
3.70
18.52
14.81
A
11-13
13-15
B
August
2001
1370-rn
1525-rn
frequency
Stevens
time
(%)
time
7-9
9-11
8.70
91.30
7-9
9-11
A
1675-rn
A
(%)
14.81
85.19
1830-rn
frequency
frequency
time
7-9
9-11
A
(%)
18.52
81.48
frequency
time
7-9
9-11
11-13
(%)
14.81
81.48
3.70
A
note: letters different from each other indicate difference in time of first peak across
elevations within a month, not across months (p<O.O5) (time of first peak is defined
as the first time in the cycle where the maximum amount of heating was recorded)
48
Cooling Cycle:
A consistent pattern of temperature change during the cooling cycle was not apparent
across years and months on Barney Creek: elevation differences were present in
2000 but not in 2001. in general, cooling was greatest at the middle elevations, with
the largest amount of cooling at 1675-rn in July and at 1525-rn in August for both
years.
The period of greatest cooling occurred between 19-2 1 h at all elevations
with the exception of the highest elevation studied. The timing of peak cooling at
1830-rn elevation was almost always different from the lower elevations, beginning
between 17-19 h and extending into the 19-21 h period. Higher elevations with an
open exposure are likely to be subject to more rapid thermal cycling (Larson and
Larson 1997). Air density decreases with elevation, reducing the ability of the air
mass to absorb and emit longwave radiation (Meisner et al. 1998). Tables 8 and 9
illustrate the timing and frequency of peak cooling on each creek.
Similar to the pattern seen on Barney Creek, Stevens Creek showed the amount of
temperature change during peak periods to be different across elevations in 2000
(July p-value = 0.001, August p-value <0.000 1) but not in 2001 (July p-value
0.10,
August p-value = 0.12). Air cooling on Stevens Creek generally began between 1921h, but often continued into later periods. Figure 9 shows the amount of air
temperature change seen during the cooling cycle for Stevens Creek 2000. Cooling
at the two lowest elevations was greatest between 19-21h. Cooling at the upper two
elevations began between 17-19 h and continued through 19-2 1 h.
49
Table 8. Frequencies of when peak air cooling occurred across elevations on Barney
Creek, both years.
July
2000
1370-rn
1525-rn
frequency
Barney
time
17-19
19-21
21-23
(%)
7.14
75.00
17.86
A
1675-rn
frequency
time
17-19
19-21
21-23
23-1
A
(%)
3.57
78.57
14.29
3.57
1830-rn
frequency
time
17-19
19-21
21-23
(%)
7.14
75.0
17.86
A
frequency
time
17-19
19-21
21-23
(%)
35.71
60.71
3.57
B
August
2000
1370-rn
1525-rn
frequency
Barney
time
17-19
19-21
(%)
3.45
93.10
time
19-21
21-23
23-1
1675-rn
frequency
(%)
93.10
6.90
1830-rn
frequency
time
21-23
(%)
13.79
75.86
6.90
1-3
3.45
17-19
19-21
frequency
time
(%)
17-19
79.31
19-21
17.24
3.45
21-23
3.45
A
A
A
B
July
2001
1370-m
1525-rn
frequency
Barney
1830-rn
frequency
frequency
time
(%)
time
(%)
time
(%)
time
(%)
17-19
19-21
21-23
11.11
17-19
19-21
21-23
11.11
17-19
19-21
25.93
70.37
17-19
19-21
37.04
55.56
23-1
3.70
23-1
3.70
3-5
3.70
66.67
22.22
- 23-1
1
1675-rn
frequency
A
77.78
7.41
3.70
A
A
A
August
2001
1370-rn
1525-rn
frequency
Barney
time
(%)
time
17-19
19-21
21-23
96.30
3.70
1675-rn
frequency
19-21
21-23
(%)
22.22
74.07
3.70
A
frequency
time
(%)
time
(%)
17-19
19-21
21-23
7.41
17-19
85.19
3.70
3.70
21-23
59.26
37.04
3.70
23-1
A
1830-rn
frequency
A
19-21
B
note: letters different from each other indicate difference in time of first peak across
elevations within a month, not across months (p<O.O5) (time of first peak is defined
as the first time in the cycle where the maximum amount of cooling was recorded)
50
Table 9. Frequencies of when peak air cooling occurred across elevations on
Stevens Creek, both years.
July
2000
1370-rn
1525-rn
frequency
Stevens
time
17-19
19-21
21-23
23-1
1-3
1
(%)
10.71
57.14
25.00
3.57
3.57
A
1675-rn
frequency
time
17-19
19-21
21-23
(%)
17.86
60.71
17.86
23-1
3.57
AB
1830-rn
frequency
time
17-19
19-21
21-23
(%)
35.71
53.57
10.71
BC
frequency
time
17-19
19-21
21-23
(%)
53.57
42.86
3.57
C
August
2000
1370-rn
1525-rn
frequency
Stevens
time
19-21
21-23
1675-rn
1830-rn
frequency
frequency
frequency
(%)
time
(%)
time
(%)
time
(%)
10.34
89.66
17-19
19-21
21-23
6.9
89.66
3.45
17-19
19-21
44.83
51.72
17-19
19-21
37.93
58.62
1-3
3.45
1-3
3.45
A
A
A
A
July
2001
1370-rn
1525-rn
frequency
Stevens
time
21-23
(%)
25.93
48.15
18.52
1-3
7.41
17-19
19-21
A
1675-rn
frequency
time
17-19
19-21
21-23
23-1
(%)
22.22
59.26
14.81
3.7
time
17-19
19-21
21-23
23-1
1-3
A
1830-rn
frequency
(%)
40.74
48.15
3.70
3.70
3.70
A
frequency
time
17-19
19-21
(%)
48.15
48.15
3-5
3.70
A
August
2001
1370-rn
1525-rn
frequency
Stevens
time
17-19
19-21
1
21-23
AB
(%)
43.48
52.17
4.35
1675-rn
frequency
time
17-19
19-21
21-23
A
(%)
22.22
74.07
3.70
1830-m
frequency
time
17-19
19-21
21-23
B
(%)
55.56
40.74
3.70
frequency
time
17-19
19-21
21-23
(%)
66.67
29.63
3.70
B
note: letters different from each other indicate difference in time of first peak across
elevations within a month, not across months (p<O.05) (time of first peak is defined
as the first time in the cycle where the maximum amount of cooling was recorded)
51
Both creeks were similar in times of peaks and amounts of cooling, however Stevens
Creek appeared to have cooling more spread out over the 4-hour period from 19-23 h
than did Barney Creek. This is likely associated with the insulative effect of the
closed forest canopy. In other studies, forest vegetation has been shown to delay
radiation losses overnight (Le Blanc et al. 1997; Swift and Messer 1971). The higher
elevations on both creeks also showed similar patterns of earlier evening cooling
compared to lower elevations.
Water Temperature Patterns
Water temperature patterns were similar across both creeks. Peak amounts of
heating decreased with increasing elevation. The heating period typically was
concentrated into two periods while cooling was extended over longer periods of
time.
Heating Cycle:
Water temperature change during peak heating periods was different across all
elevations on Barney and Stevens Creeks for both years (p-values <0.000 1). Figures
10 and 11 show the 2-hour period temperature changes for water in July of 2000 for
each creek (graphs of results for both creeks and years are contained within
Appendix 3).
Figure 10. 2-hour period temperature changes for water, Barney Creek July 2000. Temperature changes without a common
letter (black) are significantly different (p<O.O5). Mean amount of peak heating/cooling without a common letter (starred,
blue) are significantly different (p<O.O5) across elevations.
Barney Creek 1370m, July 2000
Barney Creek 1525m, July 2000
6
A
Mean amount of peak, heating: 4.40°C A*
A
a,
Mean amount of peak, cooling: -2.46°C E*
C.,
sc-)
B
C
t!a,
.2
Co
-1
r.')
(°)
-2
(5
E
E
-3
G
F
F
D
E
F
B
0
C
0
1
E
-i
01
C.)
-2
-3
H
H
period (time of day)
period (time of day)
Barney Creek 1830m, July 2000
Barney Creek 1675m, July 2000
6
6
Mean amount of peak, heating: 2.94°C (:*
5
4,
C)
a,
Mean amount of peak, cooling: -0.80°C F*
(5
A
(.)
01
-4
-4
0)
Mean amount of peak, cooling: -2.43°C E*
C
3
(l)
0)
C
A
4
.c
B
Mean amount of peak, heating: 3.82°C B*
5
Q,
o
3
Mean amount of peak, heating: 2.32°C D°
5
Mean amount of peak, cooling: -1.39°C C
4
A
3
A
B
C
C
C
C
(0
(0
C..)
(;')
01
-4
£
G
C
F
F
£
E
E
E
C
period (time of day)
period (time of day)
F
D
F
D
D
Figure 11. 2-hour period temperature changes for water, Stevens Creek July 2000. Temperature changes without a common
letter (black) are significantly different (p<O.05). Mean amount of peak heating/cooling without a common letter (starred,
blue) are significantly different (p<O.O5) across elevations.
Stevens Creek 1370m, July 2000
6
a)
5
C
4
C)
-c
A
3
C.,
a)
Wi
C)
- Wa
.i
5
C
4
-C
B
C.)
2Q
B
-.
a)
C)
Mean amount of peak, cooling: -1.55°C E *
B
Stevens Creek 1525m, July 2000
6
Mean amount of peak, heating: 4.22°C A*
Mean amount of peak, healing: 1.36°C B*
Mean amount of peak, cooling: -0.60°C F*
3
A
E
F
C
D
1'.)
'__-1
.
03
2
D
n
(()
(0
E
E
E
E
D
-3
C.)
C)
Mean amount of peak, cooling: -0.49°C F*
A
C
E
E-C
4
C
F
G
G
F
G
Mean amount of peak, cooling: -0.34°C G*
0
A
B
B
B
.-
D
--4
2
F
-C
3
-1
E
0)
Mean amount of peak, heating: 0.91 °C ('*
5
a)
G)
.
('3
Stevens Creek 183Gm, July 2000
6
4
-
-
period (time of day)
Mean amount of peak, heating: 1.05°C BC*
5
O2
a)
E a)1
(i.)
('3
G
Stevens Creek 1675m, July 2000
6
-C
(
-4
period (time of day)
C,
C
(C)
E
E
-4
a)
--4
-4
01
a)
i3
B
C
E0
1
(0
-1
C
-'I
(0
-;J
(.0
(11
(0
(()
(
-
('3
('3
C.)
F
E
)
(7
a)
a)1
-1
D
A
C
E
=
a)Q
((fl
4
-4
(0
()
((Y(
.;-4
(
F
F
F
-4
period (time of day)
N,)
((.)
F
F
F
-4
N)
-
0)
CM
3
,,
a)
F
E
period (time of day)
F
F
F
54
On Barney Creek the greatest amount of water heating occurred at 1370-rn elevation
and declined approximately 1°C with each 150 m increase in elevation, with the
lowest amount of heating occurring at 1830-rn. On Stevens Creek the greatest
amount of heating occurred at 1370-rn elevation (4-5°C) with higher elevations
having temperature changes of 1-1.5°C.
The time of peak heating on Barney Creek was different across elevations in August
(2000 and 2001 p-values <0.0001), but not July (2000 p-value
0.067;
2001 p-value = 0.306). In July the time of peak heating was concentrated between 911 h with a tendency for heating at the 1830-rn and 1370-rn elevations to shift into
the 11-13 h period. In August the pattern continued with heating at 1830 and 1370m elevations concentrated in the 11-13 h period; mid-elevations remained in the 9-11
hperiod.
The time of peak heating on Steven Creek did not differ across elevation with the
exception of August 2001 (p-value = 0.006). During the majority of the study, the
time of peak heating occurred between 11-13 h across all elevations. The August
2001 exception resulted from an increased frequency for the time of peak heating at
the lowest elevation to shift into the 13-15 h period. Tables 10 and 11 illustrate the
timing and frequency of peak heating for water on Barney Creek and Stevens Creek
across both years. LSD significance indicated that timing of peak heating was
different across certain elevations within a month, unless otherwise stated. If not
statistically significant (ie: p-value not less than 0.05), p-values are given.
55
Table 10. Frequencies of when peak water heating occurred across elevations on
Barney Creek, both years.
July
2000
1370-rn
1525-rn
frequency
Barney
time
(%)
time
1830-rn
1675-rn
frequency
(%)
frequency
time
(%)
11-13
32.14
64.29
9-11
11-13
75.00
21.43
9-11
11-13
64.29
28.57
15-17
3.57
15-17
B
3.57
15-17
7.14
9-11
A
AB
frequency
time
7-9
9-11
11-13
13-15
15-17
(%)
3.57
28.57
57.14
7.14
3.57
A
August
2000
1370-rn
1525-rn
frequency
Barney
time
(%)
time
9-11
3.45
96.55
9-11
11-13
11-13
1
A
1675-rn
frequency
(%)
79.31
20.69
B
1830-rn
frequency
frequency
time
time
9-11
(%)
79.31
11-13
20.69
11-13
13-15
9-11
(%)
10.34
86.21
3.45
A
B
July
2001
1370-rn
1525-rn
frequency
Barney
time
9-11
11-13
13-15
(%)
48.28
44.83
6.90
1675-rn
frequency
time
9-11
11-13
13-15
(%)
75.86
17.24
6.9
A
A
1830-rn
frequency
frequency
time
(%)
time
9-11
11-13
13-15
72.41
17.24
9-11
10.34
A
11-13
13-15
A
(%)
55.17
34.48
10.34
August
2001
1370-rn
1525-rn
frequency
Barney
time
9-11
11-13
13-15
1
1
A
1675-rn
frequency
(%)
time
3.45
93.10
3.45
7-9
9-11
11-13
13-15
B
(%)
3.45
75.86
17.24
3.45
1830-rn
frequency
time
(%)
frequency
time
7-9
9-11
11-13
B
55.17
44.83
9-11
11-13
13-15
(%)
3.45
17.24
72.41
6.90
A
note: letters different from each other indicate difference in time of first peak across
elevations within a month, not across months (p<0.O5) (time of first peak is defined
as the first time in the cycle where the maximum amount of heating was recorded)
56
Table 11. Frequencies of when peak water heating occurred across elevations on
Stevens Creek, both years.
July
2000
1370-rn
1525-rn
frequency
Stevens
time
(%)
1675-rn
tirne
(%)
1830-rn
frequency
frequency
time
(%)
frequency
tirne
7-9
9-11
11-13
13-15
15-17
3.57
89.29
3.57
3.57
1
11-13
92.86
15-17
7.14
AB
9-11
10.71
9-11
11-13
13-15
15-17
B
64.29
21.43
3.57
11-13
(%)
3.57
3.57
92.86
A
August
2000
1370-rn
1525-m
frequency
time
Stevens
11-13
(%)
100
A
1
July
-
2001
1370-rn
tirne
11-13
13-15
Stevens
time
9-11
11-13
13-15
1
(%)
96.55
3.45
A
(%)
3.45
89.66
6.9
A
frequency
time
11-13
13-15
1675-rn
frequency
time
9-11
11-13
13-15
A
(%)
96.55
3.45
A
1525-rn
frequency
1830-m
1675-rn
frequency
(%)
6.90
86.21
6.90
frequency
tirne
11-13
13-15
A
1830-rn
frequency
time
9-11
11-13
13-15
(%)
93.10
6.90
(%)
3.45
93.10
3.45
A
frequency
time
9-11
11-13
13-15
A
(%)
17.24
79.31
3.45
Augu
2001
1370-rn
1525-rn
frequency
Stevens
time
(%)
time
9-11
11-13
13-15
A
1
79.17
20.83
1675-rn
frequency
11-13
13-15
B
(%)
3.45
89.66
6.90
1830-rn
frequency
time
(%)
frequency
time
9-11
11-13
B
100
11-13
(%)
6.90
93.10
B
note: letters different from each other indicate difference in time of first peak
across elevations within a month, not across months (p<O.05) (time of first peak is
defined as the first time in the cycle where the maximum amount of heating was
recorded)
57
Cooling Cycle.
Peak water temperature change during the cooling cycle on Barney Creek was
different across all elevations and both years (p-value <0.0001). The amount of peak
cooling decreased with increasing elevation and resulted in a high-to-low elevation
difference of 1-1.5°C. Cooling was generally concentrated in the 17-19 h and
19-2 1 h periods with no difference in the amount of cooling taking place during
either period. Differences in August 2000 and 2001 show a tendency for the time of
peak cooling to become more concentrated in the 19-21 h period at both the 1830-rn
and 1370-rn elevations.
The amount of peak water cooling on Stevens Creek decreased with increasing
elevation and range from 0.3 to 1.7°C. The cooling was typically spread across
several periods with each period having similar amounts of temperature loss, making
a determination of peak cooling meaningless.
Tables 12 and 13 illustrate the timing and frequency of peak cooling for water on
Barney Creek and Stevens Creek across both years. LSD significance indicates that
timing of peak cooling was different across certain elevations within a month, unless
otherwise stated. If not statistically significant (ie: p-value not less than 0.05), pvalues are given.
58
Table 12. Frequencies of when peak water cooling occurred across elevations on
Barney Creek, both years.
July
2000
1370-rn
1525-rn
frequency
Barney
tirne
17-19
19-21
21-23
1-3
1675-rn
frequency
time
(%)
7.14
85.71
3.57
3.57
17-19
19-21
21-23
(%)
53.57
39.29
7.14
time
17-19
19-21
21-23
23-1
A
1830-rn
frequency
(%)
67.29
7.14
21.43
7.14
AB
B
frequency
time
17-19
19-21
21-23
23-1
1-3
(%)
42.86
25.00
21.43
7.14
3.57
A
August
2000
1370-rn
1525-rn
frequency
Barney
time
19-21
1675-rn
frequency
(%)
100
time
(%)
time
17-19
68.97
31.03
17-19
19-21
1830-rn
frequency
19-21
(%)
89.66
10.34
frequency
time
17-19
19-21
21-23
A
B
July
2001
0
C
-
1370-rn
1525-rn
frequency
Barney
(%)
37.93
58.62
3.45
1675-rn
frequency
1830-rn
frequency
frequency
time
(%)
time
(%)
time
(%)
time
(%)
17-19
20.69
17-19
17-19
72.41
6.90
19-21
55.17
37.93
17-19
19-21
21-23
41.38
51.72
6.90
21-23
6.90
21-23
41.38
34.48
20.69
3.45
21-23
19-21
19-21
23-1
A
A
A
A
August
2001
1370-rn
1525-rn
frequency
Barney
time
(%)
time
17-19
19-21
1
1
A
100.00
1675-rn
frequency
19-21
BC
(%)
51.72
48.28
1830-rn
frequency
time
17-19
19-21
B
(%)
68.97
31.03
frequency
time
17-19
19-21
(%)
44.83
51.72
23-1
3.45
C
note: letters different from each other indicate difference in time of first peak across
elevations within a month, not across months (p<O.O5) (time of first peak is defined
as the first time in the cycle where the maximum amount of cooling was recorded)
59
Table 13. Frequencies of when peak water cooling occurred across elevations on
Stevens Creek, both years.
July
2000
1370-rn
1525-rn
frequency
Stevens
time
(%)
17-19
42.86
21.43
25.00
7.14
3.57
19-21
21-23
23-1
1-3
A
1675-rn
frequency
time
1830-rn
frequency
time
(%)
(%)
frequency
time
17-19
19-21
21-23
23-1
1-3
3-5
10.71
35.71
21-23
35.71
21-23
23-1
42.86
1-3
3-5
39.29
17.86
7.14
23-1
1-3
3-5
10.71
B
(%)
28.57
28.57
17.86
10.71
7.14
7.14
A
B
August
2000
1370-rn
1525-rn
frequency
Stevens
time
19-21
21-23
(%)
68.97
31.03
time
3-5
1830-rn
frequency
time
(%)
19-21
21-23
23-1
1-3
A
1675-rn
frequency
41.38
37.93
21-23
23-1
1-3
6.90
13.79
3-5
B
(%)
3.45
41.38
34.48
17.24
3.45
frequency
time
(%)
17-19
13.79
34.48
34.48
10.34
3.45
3.45
19-21
21-23
23-1
1-3
3-5
A
B
July
2001
1370-rn
1525-rn
frequency
Stevens
time
17-19
19-21
21-23
23-1
(%)
24.14
34.48
27.59
13.79
time
19-21
21-23
23-1
1-3
3-5
1
1675-rn
frequency
A
B
2001
1370-rn
1525-rn
3.45
24.14
20.69
34.48
17.24
19-21
21-23
23-1
1-3
3-5
B
(%)
6.90
34.48
17.24
24.14
17.24
frequency
time
(%)
13.79
17-19
19-21
21-23
48.28
31.03
1-3
6.90
A
-
frequency
Stevens
time
(%)
August
1830-rn
frequency
time
17-19
19-21
21-23
23-1
A
(%)
12.50
4.17
45.83
37.50
1675-rn
frequency
time
19-21
21-23
23-1
1-3
3-5
B
(%)
6.90
48.28
17.24
13.79
13.79
1830-rn
frequency
time
19-21
21-23
23-1
1-3
3-5
C
(%)
3.45
20.69
17.24
20.69
37.93
frequency
time
17-19
19-21
21-23
23-1
1-3
3-5
A
(%)
10.34
48.28
13.79
20.69
3.45
3.45
note: letters different from each other indicate difference in time of first peak across
elevations within a month, not across months (p<O.O5) (time of first peak is defined
as the first time in the cycle where the maximum amount of cooling was recorded.)
60
Discussion:
In this study, peak water heating lagged behind peak air heating in a maimer similar
to that reported in other studies (Stefan and Preud'homme 1993; Walker and Lawson
1977; Needharn and Jones 1959). On Barney Creek this margin was between two
and four hours, with the 4-hour lag being the most prevalent at 1370 and 1830-rn.
When 2-hour lags were evident, peak air heating occurred between 7-9 h and peak
water heating occurred between 9-11 h. The 4-hour lag showed peak air heating
between 7-9 h and peak water heating between 11-13 h. This difference was
associated with the greater amount of air temperature change being observed during
the 7-9 h period at the two middle elevations. A larger increase in air temperature at
that time would result in a greater temperature gradient between air and water and a
faster rate of water heating (Larson and Larson 1997; LeBlanc et al. 1997; Mosley
1983; Dingman 1971; Raphael 1962). On Stevens Creek, peak water heating tended
to lag peak air heating by two hours. In this case, peak air heating occurred between
9-11 h (2 hours later than Barney Creek), and peak water heating occurred between
11-13 h.
Data from the cooling cycle on Barney Creek indicated that air and water were most
commonly experiencing maximum cooling during the same time period (no lag)
across all elevations. Exceptions could be found at both of the mid elevations where
water was observed entering a peak cooling period before air, and at the 1370-rn
elevation where water cooling was observed lagging behind air cooling. Data from
Stevens Creek appeared more widely scattered with no real pattern of cooling
evident.
As stated earlier, most of the patterns of heating and cooling appeared to be
attributable to a number of factors, including the drainage basin exposure, patterns of
air movement, and the amount of exposure time. On these two creeks, air
temperatures were responsive to weather patterns and the absorption characteristics
61
of their respective drainage basins, with rapid changes often evident. Water
temperatures are buffered from rapid fluctuations due to mass and specific heat
characteristics (Ward 1985), and relate more to the daily trends of air temperature
than patterns of rapid fluctuation. In this study the data support the findings of
several researchers that small size streams that lack thermal buffers are responsive to
the surrounding thermal environment (Stefan and Preud'homme 1997; Sinokrot and
Stefan 1993; Smith and Lavis 1974; Swift and Messer 1971).
62
OBJECTIVE 3: MODELING VARIABLES STRONGLY ASSOCIATED WITH
MAXIMUM WATER TEMPERATURE:
Results:
Stepwise regression was used to determine variables most strongly related to
maximum water temperature. Results for 2000 are summarized in Tables 14 and 15.
Summaries for both creeks in 2001 are contained in Appendix 4.
A reminder that for this objective, "current day" is defined as the 24-hour period
between 5pm and 5pm. "Previous day" includes data from 5am to 5pm of the day
before the current day.
Definitions of each variable to enter the model are as follows:
> "max air temp. current day": the highest air temperature reached on the
current day;
> "peak air heating change, current day"; the change in air temperature during
the 2-hour period of the heating cycle with the greatest heating;
> "mm. air temp, previous day": the lowest air temperature reached on the
previous day;
> "air cooling, previous day": the net air temperature change during the cooling
portion of the diurnal cycle (Spm-Sam);
"mm. water temp. previous day": the lowest water temperature recorded
during the previous day;
> "water cooling previous day": the total water temperature change during the
cooling portion of the diurnal cycle (Spm-5am).
63
Table 14. Table of variables strongly associated with maximum water temperature,
for Barney Creek, year 2000.
Barney 1370-rn, July 2000
Variable
max air temp, current day
water cooling, prey, day
mm. water temp, prey, day
Barney 1370-rn, August 2000
Variable
max air temp, current day
mm. water temp, prey, day
Barney 1525-rn, July 2000
Variable
max air temp, current day
air cooling, prey, day
Barney 1525-rn, August 2000
Variable
max air temp. current day
mm water temp, prey, day
Barney 1675-rn, July 2000
Variable
max air temp, current day
air cooling, prey, day
peak air heating change, current day
Barney 1675-rn, August 2000
Variable
max air temp, current day
mm water temp, prey, day
Barney 1830-rn, July 2000
Variable
max air temp, current day
peak air heating change, current day
air cooling, prey, day
Barney 1830-rn, August 2000
Variable
max air temp, current day
mm water temp, prey, day
Partial R2
Prob F
.8414
.0446
.0243
<.0001
.0077
.0264
Partial R2
Prob F
.7803
.0796
<.0001
.0011
Partial R2
.8472
Prob F
.0296
<.0001
.0313
Partial R2
Prob F
.8200
.0723
<.0001
.0005
Partial R2
Prob F
.8648
<.0001
.0052
.04 12
.0169
.043 8
Partial R2
Prob F
.8248
.0926
<.000 1
<.000 1
Partial R2
Prob F
.8954
.0351
.0304
<.0001
.0030
.0006
Partial R2
Prob F
.8264
.1149
<.000 1
<.0001
64
Table 15. Table of variables strongly associated with maximum water temperature,
for Stevens Creek, year 2000.
Stevens 1370-rn, July 2000
Variable
max air temp, current day
nun water temp, prey, day
peak air heating change, current day
Stevens 1370-rn, August
2000
.
Vanable
max air temp, current day
mill water temp, prey, day
Stevens 1525-rn, July 2000
Variable
max air temp, current day
mm water temp, prey, day
water cooling, prey, day
Stevens 1525-rn, August
2000
.
Variable
mm air temp, prey, day
max air temp, current day
peak air heating change, current day
mm water temp, prey, day
REMOVED mm air temp, prey, day
water cooling, prey, day
Stevens 1675-rn, July 2000
Variable
max air temp. current day
mm water temp, prey, day
Stevens 1675-rn, August
Variable
mm air temp, prey, day
max air temp, current day
Stevens 1830-rn, July 2000
Variable
max air temp, current day
mm water temp, prey. day
Stevens 1830-rn, August
Variable
mm air temp, prey, day
max air temp, current day
mm water temp, prey, day
REMOVED mm air temp, prey, day
water cooling, prey, day
Partial R2
Prob F
.8712
.0387
.0260
<.0001
.0056
.0082
Partial R2
Prob F
.7670
.1372
<.0001
<.0001
Partial R2
Prob F
.9048
.0796
.0031
<.000 1
<.000 1
.
Partial R
.0349
2
Prob F
.02 16
<.0001
<.0001
.0087
.0244
.0799
<.0001
Partial R2
Prob F
.9054
.0723
<.0001
Partial R2
Prob F
.8535
.0824
<.000 1
<.000 1
Partial R2
Prob F
.9006
.0779
<.0001
Partial R2
Prob F
.8147
.1088
.0235
.0070
.0186
<.0001
<.0001
.0040
.0948
.0039
.8272
.1092
.0 166
.009 8
.0056
<.000 1
<.000 1
On Barney Creek, maximum air temperature from the current day entered the model
first in every case, for both years. The coefficients of determination were higher in
July for both years than in August, with August 2001 exhibiting the lowest R2s
(0.4670-0.6683). Improvements in model fit by the entry of a second variable in July
for both years accounted for an additional 3-4% of total variation. In August 2000,
65
maximum temperature from the current day was always followed by the minimum
water temperature from the previous day, which improved the R2 by 7 - 12%. In
August 2001, only the upper two elevations had a second variable enter: at 1675-rn
air cooling from the previous day improved fit by 8%, and at 1830-rn minimum
water temperature from the previous day improved fit by 15%. Model improvement
through the addition of a third variable was minimal in both years.
The Stevens Creek models for July 2000 and 2001 also selected maximum air
temperature from the current day as the first model variable. Minimum water
temperature from the previous day was the second variable selected but only
improved fit by 4 - 8%. Any variables to enter third resulted in an insignificant
increase in R2 (less than 3%). August 2000 (1370-rn) had maximum air temperature
from the current day enter first followed by minimum water temperature from the
previous day. Models developed for the upper three elevations in August 2000 had
minimum air temperature from the previous day enter first, with maximum air
temperature from the current day entering second. Any variables to enter third in
Stevens Creek models resulted in an insignificant increase in R2 (less than 3%).
Generally, models for both creeks selected one variable as an index of the heating
cycle and one variable for the cooling cycle. The models generally indicated that
maximum water temperature was strongly related to maximum air temperature from
the current day. Minimum water temperature was a common second variable
selected in model development.. The exceptions to the selection of maximum air
temperature as first variable were found on the upper three elevations of Stevens
Creek in August of 2000 and the two middle elevations in 2001. Tn these cases,
minimum air temperature from the previous day dominated as first to enter, but the
second variable to enter also greatly improved fit of the model and was often
maximum air temperature of the current day.
66
An interesting pattern to note is that coefficients of determination for the first
variable to enter the model were consistently lower in August compared to July, in
both years, across all elevations of both creeks. On Stevens Creek, the reduction in
fit for August compared to July, of the first variable to enter the model was between
4-15%, with one instance of 33%, while on Barney Creek, it was between 20-36%.
Concurrently, there was an increase in impact of the second variable to enter, which
strengthened the R2 by at least 7% and as much as 35%.
Tables 16 and 17 summarize the variable coefficients that explain maximum water
temperature. Models developed for Barney and Stevens Creeks (16 per creek, 32 in
total) had coefficients of determination >0.8 with the exception of three models on
Barney Creek, which range from 0.46 to 0.71, and two models on Stevens Creek
which were .079 and 0.68.
67
Table 16. Summary of the models explaining maximum water temperature on
Barney Creek 2000-2001, including model R2, coefficients of the variables, and their
SEs.
B1370
Intercept
Max air
Coefficient
and (SE)
2.36 (1.68)
0.50 (0.05)
-0.54 (0J8)
Coefficient
and (SE)
2.84 (1.46)
0.50 (0.05)
-0.17 (0.07)
Coefficient
and (SE)
0.67 (1.10)
0.48 (0.05)
-0.19 (0.06)
Coefficient
and (SE)
2.91 (0.70)
0.41 (0.04)
0.23 (0.07)
Coefficient
and (SE)
9.91 (1.14)
0.47 (0.04)
Coefficient
and (SE)
9.32 (1.07)
0.45 (0.04)
Coefficient
and (SE)
7.28 (0.83)
0.37 (0.03)
July 2001
Variable
Coefficient
and (SE)
R2 = 0.8918
Intercept
Max air
Peak air heating change
6.84 (0.60)
0.42 (0.04)
-0.24 (0.11)
July 2000
Variable
R2 = 0.8860
Intercept
Max air
Water cooling prey, day
B1525
July 2000
Variable
R2 = 0.8768
Intercept
Max air
Air cooling prey, day
B1675
July 2000
R2 = 0.9060
Variable
B1830
July 2000
= 0.9305
Variable
B1370
July 2001
Rz= 0.8240
Variable
B1525
July 2001
R2 0.8285
Variable
B1675
July 2001
R2 0.8456
Variable
Intercept
Max air
Air cooling prey, day
Intercept
Max air
Peak air heating change
Intercept
Max air
Intercept
Max air
151830
August 2000
Variable
Coefficient
and (SE)
R2 = 0.8599
Intercept
Max air
Minwat prey. day
9.12 (1.13)
0.30 (0.05)
0.38 (0.10)
131370
B1370
August 2001
R2 = 0.6090
Variable
Coefficient
and (SE)
10.39 (0.78)
0.25 (0.04)
0.35 (0.09)
Coefficient
and (SE)
6.36 (0.70)
0.26 (0.04)
0.45 (0.09)
Coefficient
and (SE)
4.51 (0.54)
0.22 (0.03)
0.55 (0.08)
Coefficient
and (SE)
Intercept
Max air
11.91 (1.80)
0.38 (0.06)
B1525
August 2001
R2 = 0.4670
Variable
151525
August 2000
Variable
R2 = 0.8923
Intercept
Max air
Minwat prey, day
131675
August 2000
Variable
R2 = 0.9174
Intercept
Max air
Minwat prey, day
151830
August 2000
Variable
R2 = 0.9413
Intercept
Max air
Minwat prey, day
Intercept
Max air
151675
August 2001
R = 0.7173
B1830
August 2001
R2 = 0.8166
Variable
Intercept
Max air
Air cooling prey, day
Variable
Intercept
Max air
Minwat prey, day
Coefficient
and (SE)
12.93 (1.94)
0.31 (0.07)
Coefficient
and (SE)
5.11(1.76)
0.36 (0.05)
-0.15 (0.06)
Coefficient
and (SE)
5.24 (1.01)
0.27 (0.03)
0.35 (0.08)
68
Table 17. Summary of the models explaining maximum water temperature on
Stevens Creek 2000-2001, including model R2, coefficients of the variables, and
their SEs.
S1370
July 2000
Variable
R2 = 0.9099
Intercept
Max air
Minwat prey, day
S1525
July 2000
= 0.9844
Variable
Intercept
Max air
Minwat prey. day
S1675
July 2000
Variable
R2 = 0.9777
Intercept
Max air
Minwat prey. day
S1830
July 2000
R2
0.9785
S1370
July 2001
= 0.7906
S1675
July. 2001
R2 = 0.8941
1.75 (0.30)
0.21 (0.02)
0.49 (0.06)
Variable
Coefficient
and (SE)
Intercept
Max air
Minwat prey, day
1.56 (0.26)
0.12 (0.01)
0.58 (0.06)
Variable
Intercept
Max air
S1525
July 2001
= 0.9115
Coefficient
and (SE)
2.22 (1.15)
0.46 (0.14)
0.44 (0.14)
Coefficient
and (SE)
1.44 (0.31)
0.26 (0.02)
0.50 (0.05)
Coefficient
and (SE)
Variable
Intercept
Max air
Peak airheating change
Variable
Intercept
Max air
Mm air prey. day
S1830
July 2001
Variable
R2 = 0.8244
Intercept
Max air
Peak air heating change
S1370
August 2000
Variable
R2 = 0.9042
Intercept
Max air
Minwat prey, day
S1525
August 2000
R2 = 0.9368
Variable
S1675
August 2000
R2 0.9360
Variable
S1830
August 2000
R2
0.923 5
Intercept
Mm air prey. day
Max air
Intercept
Mm air prey. day
Max air
Variable
Intercept
Minairprev. day
Max air
Coefficient
and (SE)
8.89 (1.10)
0.42 (0.04)
S1370
August 2001
R2 = 0.6875
Coefficient
and (SE)
6.22 (0.45)
0.44 (0.03)
-0.55 (0.08)
Coefficient
and (SE)
6.58 (0.36)
0.15 (0.02)
0.18 (0.04)
Coefficient
and (SE)
5.88 (0.22)
0.16 (0.02)
-0.17 (0.04)
SI525
August 2001
R2 = 0.8955
Variable
Intercept
Max air
Water cooling prey, day
Mm air prey. day
Variable
Intercept
Mm air prey. day
Max air
S1675
gust 2001
R2 = 0.8789
Variable
S1830
August 2001
R2 = 0.9005
Variable
Intercept
Mm air prey, day
Max air
Intercept
Max air
Minwat prey. day
Coefficient
and (SE)
6.45 (0.92)
0.30 (0.04)
0.45 (0.08)
Coefficient
and (SE)
4.08 (0.70)
0.39 (0.05)
0.22 (0.03)
Coefficient
and (SE)
5.36 (0.51)
0.33 (0.03)
0.14 (0.02)
Coefficient
and (SE)
5.58 (0.29)
0.14 (0.02)
0.09 (0.02)
Coefficient
and (SE)
3.39 (3.00)
0.34 (0.07)
-0.86 (0.24)
0.22 (0.10)
Coefficient
and (SE)
6.61 (0.70)
0.35 (0.03)
0.16 (0.02)
Coefficient
and (SE)
6.25 (0.63)
0.28 (0.03)
0.14 (0.02)
Coefficient
and (SE)
3.24 (0.41)
0.10 (0.01)
0.50 (0.05)
69
The coefficients indicate that an increase of 1°C in the current day maximum air
temperature was associated with an increase in maximum water temperature that
ranged from 0.22°C to 0.50°C on Barney Creek, and from 0.10°C to 0.46°C on
Stevens Creek when other model variables were held constant. Where minimum air
temperature from the previous day resulted in the best fit, (Stevens Creek) a 1°C
increase in minimum air temperature from the previous day was associated with an
increase in the maximum water temperature of between 0.14°C to 0.39°C, holding
all other variables constant.
Discussion:
Air temperature has been shown to have a strong relationship with water temperature
(Mohseni et al. 1998; Brown and Binkley 1994; Sinokrot and Stefan 1994; Sinokrot
and Stefan 1993; Smith and Lavis 1974; Levno and Rothacher 1967; Needham and
Jones 1959), particularly in streams with lower discharge (Stefan and Preud'homme
1993; Smith and Lavis 1974). Stefan and Preud'homme (1993) explained that water
responds to air more directly in lower flow streams because the thermal inertia of the
water is relatively low due to the relatively little amount of water present. As seen
with our study, maximum air temperature was always strongly related to maximum
water temperature on Barney Creek, being the first variable to enter our model, and
having a large coefficient of determination. With Stevens Creek, maximum air
temperature was first to enter at the lowest elevation, and usually second to enter at
the upper elevations in August. In terms of maximum and minimum temperatures,
Barney Creek was generally warmer than Stevens Creek, with maximum air
temperatures within 0.14 to 2.96°C, minimum air temperatures within -1.03 to
1.67°C, and minimum water temperatures within -0.042 to 2.65°C of Stevens Creek.
Maximum water temperature was fairly similar at 1 370-m, with Barney Creek
between 1.85 to 2.67°C warmer than Stevens Creek. However, at the three upper
elevations of Barney Creek, maximum water temperatures ranged from 5.5 to 8.17°C
70
greater than those on Stevens Creek at corresponding elevations. These lower
maximum water temperatures on Stevens Creek may account for the increased
importance of minimum water temperature from the previous day in the model, and
indicate the potential influence of a thermal buffer affecting these three stations. The
buffer may be related to over-night cooling, the temperature, volume, and location of
the water source, and/or the velocity of the stream flow at the higher elevations.
Another variable of importance was minimum air temperature from the previous day,
which often entered first in August of both years for the two middle elevations of
Stevens Creek. This may be due to the inverse relationship between peak air heating
and elevation seen on Stevens Creek; as elevation increased, peak air heating values
decreased, thus the influence of the minimum air temperature would be greater at
higher elevations. This inverse relationship was not seen on Barney Creek, and we
did not see minimum air temperature from the previous day playing a significant role
in the model statements for Barney Creek.
Differences between July and August models on both creeks, including changes in
the influence of the second variable to enter may be attributable to differences in
meteorological patterns from month to month. Model coefficients indicate
maximum air temperature has a very similar relationship with water temperature on
both creeks, regardless of creek differences in vegetation cover, discharge, exposure
time, and other parameters. That weather influences affect the water temperature of
both creeks in a like fashion would seem likely, as both creeks share the same aspect
and watershed.
Maximum air temperature had a very strong relationship to maximum water
temperature on all elevations of Barney Creek and at the lowest elevation of Stevens
Creek. The upper three elevations of Stevens Creek however, showed that minimum
71
air temperature from the day before dominated as the variable most strongly related
to maximum water temperature. Our findings were consistent with other researchers
who have shown that air temperature has a strong relationship with water
temperature (Mohseni et al. 1998; Brown and Binkley 1994; Sinokrot and Stefan
1994; Sinokrot and Stefan 1993; Smith and Lavis 1974; Levno and Rothacher 1967;
Needham and Jones 1959), particularly in streams with lower discharge (Stefan and
Preud'homme 1993; Smith and Lavis 1974). The significance of minimum air
temperature on the upper three elevations was likely attributable to the influence of
groundwater seeps. Groundwater inputs have been shown to depress diurnal
fluctuations in stream temperatures (Walker and Lawson 1977; Smith and Lavis
1974; Vugts 1974), with water temperatures thus not tracking air temperatures as
closely (Webb and Nobilis 1997). Coefficient values for model variables indicated a
smaller increase in minimum air temperature from the previous day affected the
same response in maximum water temperature as a larger increase in maximum air
temperature, holding all other variables constant.
72
OBJECTIVE 4: DERIVING A PREDICTIVE EQUATION FOR WATER
TEMPERATURE CHANGE:
Results:
A reminder that "current day" for this objective is always defined as the 24-hour
period between 5am and Sam. "Previous day" for this objective is the 24-hour
period from 5am to Sam of the day prior to the current day.
The five variables selected from the pool of 11 variables outlined in the methods
section included:
> "airday": net air temperature change during the 24-hour period (5am-Sam) of
the current day - how much the air temperature changed over the course of
the current day;
> "maxair": maximum air temperature reached during the current day;
> "minair": minimum air temperature reached during the current day;
> "minwat_1" minimum water temperature from the previous day;
> "watdayl ": water temperature change from the day before - how much the
water temperature changed over the course of the previous day.
Fixed stream models were developed using the five selected variables. The
coefficients of determination associated with the models ranged from 0.68 to 0.91.
Figures 12 and 13 show the fit of predicted versus observed values for both creeks in
2000, with dashed lines indicating fit within 0.5°C and solid outer lines indicating fit
within 1°C. Graphs for both creeks in 2001 are contained in Appendix 5.
Figure 12. Scatterplots demonstrating fit of prediction model for Barney Creek 2000; dashed lines indicate agreement within
0.5° C, solid outer lines indicate agreement within 1° C.
-
Scatterplot of predicted vs observed values testing
fit of prediction model for Barney 1370m, yr 2000.
Scatterplot of predicted vs observed values testing
fit of prediction model for Barney 1525m, yr 2000.
predicted
predicted
__-
Scatterplot of predicted vs observed values testing
fit of prediction model for Barney 1675m, yr 2000.
__-
Scatterplot of predicted vs observed values testing
fit of prediction model for Barney 1830m, yr 2000.
V
w
>
w
U)
.0
0
predicted
predicted
Figure 13. Scatterplots demonstrating fit of prediction model for Stevens Creek 2000; dashed lines indicate agreement within
0.5° C, solid outer lines indicate agreement within C.
Scatterplot of predicted vs observed values testing
fit of prediction model for Stevens 1370m, yr 2000.
predicted
Scatterplot of predicted vs observed values testing
fit of prediction model for Stevens 1675m, yr 2000.
predicted
Scatterplot of predicted vs observed values testing
fit of prediction model for Stevens 1525m, yr 2000.
predicted
____---Scatterplot of predicted vs observed values testing
fit of prediction model for Stevens 1830m, yr 2000.
predicted
75
Tables of coefficients for the models for each creek in 2000 (Tables 18 and 19) are
provided to indicate how each variable in the model had a differing influence across
elevations and creeks, along with the corresponding R2 and absolute deviation for
each model. Tables for both creeks in 2001 are contained in Appendix 5.
Table 18. Barney Creek 2000, showing fit of model, mean absolute deviation in °C
(with SDs in brackets), and variable coefficients (with SEs in brackets).
Elevation (m)
R2 (%)
Mean absolute deviation (°C)
Variable
Coefficient
Barney 1370
87.58
0.3977 (0.2777)
Barney 1525
79.31
0.4 146 (0.3350)
Intercept
airday
maxair
minair
watday 1
minwat 1
Intercept
airday
maxair
minair
watday 1
Barney 1675
74.51
0.4096 (0.3274)
Barney 1830
75.37
0.38 10 (0.2856)
0.5667 (0.6721)
0.3388 (0.0340)
0.0493 (0.0238)
0.0962 (0.0720)
-0.0314(0.1182)
-0.2567 (0.096 1)
0.9665 (0.7063)
0.1981 (0.0330)
0.0879 (0.0309)
0.1535 (0.0497)
-0.2725 (0.1142)
-0.4680 (0.0823)
1.5077 (0.6896)
0.1434 (0.0337)
0.0511 (0.0335)
0.1936 (0.0532)
-0.3605 (0.1326)
-0.4951 (0.0905)
1.4141 (0.5667)
0.1244 (0.0324)
0.0601 (0.0301)
0.1557 (0.0480)
-0.3571 (0.1337)
-0.4670 (0.0878)
minwati
Intercept
airday
maxair
minair
watday 1
minwatl
Intercept
airday
maxair
minair
watday_1
minwati
76
Table 19: Stevens Creek 2000, showing fit of model, mean absolute deviation in °C
(with SDs in brackets), and variable coefficients (with SEs in brackets).
Elevation (m)
Stevens 1370
Stevens 1525
Stevens 1675
R2 (%)
68.66
89.43
82.70
Mean absolute deviation (°C)
Variable
0.6 144 (0.8436)
Intercept
airday
maxair
minair
watday 1
minwat 1
Intercept
airday
maxair
minair
watday_1
minwat 1
Intercept
airday
maxair
minair
0.2296 (0.2000)
0.2436 (0.1837)
watdayl
minwat 1
Stevens 1830
83.01
0.1192 (0.1179)
Intercept
airday
maxair
minair
watday_1
minwat 1
Coefficient
-0.1274 (0.7465)
0.1926 (0.0374)
0.1633 (0.03 87)
0.0783 (0.0437)
-0.2784 (0.1168)
-0.4840 (0.101 1)
0.9519 (0.3737)
0.2292 (0.0247)
0.0538 (0.0184)
0.1780 (0.0447)
-0.1884 (0.1043)
-0.3823 (0.0633)
1.0889 (0.3978)
0.1599 (0.0243)
0.0281 (0.0189)
0.1443 (0.0387)
-0.1851 (0.1239)
-0.3360 (0.0702)
1.3629 (0.3681)
0.0885 (0.0146)
0.0 148 (0.0101)
0.0709 (0.02 18)
-0.1650 (0.1301)
-0.3398 (0.0811)
77
Discussion:
Three variables used in model development were descriptive of air temperature
characteristics. Increases in those variables were associated with an increase in
water temperature change. This result agreed with observations by Sinokrot and
Stefan (1993), Mohseni and Stefan (1989), and Smith and Lavis (1974) Levno and
Rothacher (1967), Needham and Jones (1959), that identified a positive association
between air and water temperature. In this study, net air temperature change was
dominant among the three air temperature variables. Its influence tended to be
greatest at lower elevations and decreased at higher elevations as the stream source
was approached. The positive association between air and water temperature was
consistent across years and streams, however the magnitude of the association varied
and appeared to reflect differences in weather patterns and/or other local
environmental conditions, such as air circulation patterns. Stefan and Preud'homme
(1993) and Smith and Lavis (1974) provided some support for this observation
noting that streams having lower discharge will have lower thermal inertia and tend
to be more strongly influenced by air temperature and other local environmental
factors.
The remaining two variables used in model development were descriptive of water
temperature characteristics during the previous day. These attributes were negatively
associated with water temperature change. Water temperature change during the
previous day was dominant in model development and generally appeared to be
greatest at higher elevations and decreased at lower elevations as stream and water
temperatures moved toward thermal equilibrium. These attributes appeared to be
descriptive of changes that were taking place in the buffering capacity (temperature)
of the stream. In this case an increase in both water temperature change and
78
minimum temperature during the previous day appeared to be reflecting adjustments
in stream temperature patterns to a warming summertime environment.
80
tend to follow air temperatures closely (Sinokrot and Stefan 1993; Stefan and
Preud'homme 1993; Needham and Jones 1959).
Mean soil temperatures on Barney Creek generally exceeded those found at the same
elevations on Stevens Creek, with the exception of 1 370-rn, which may be
attributable to the presence of additional vegetation structure in the lower portion of
the Barney Creek drainage. Upper elevations of Barney Creek experienced greater
exposure to solar insolation, particularly at 1525-rn elevation, where the bowl-shaped
topography served to concentrate heat, while at 1370-rn elevation there was some
vegetation structure present. It has been shown that soil temperatures in non-shaded
areas tend to be greater than those areas having vegetative cover (Pluhowski and
Kantrowitz 1963) which, coupled with the limited air movement common in these
areas, serves to limit energy inputs to the ground surface (Fowler et al. 1979).
OBJECTIVE 2: DESCRIBING DIURNAL AIR AND WATER CYCLES.
Studies have shown that water temperatures track air temperatures with a lag of
hours or days, depending on discharge (Stefan and Preud'homme 1993; Walker and
Lawson 1977; Needham and Jones 1959). Both creeks in our study corroborated
these findings. Barney Creek had a time lag of 2-4 hours, with 4 hours most
prevalent at either extreme of elevation. Peak air heating was between 7-9 h and
peak water heating fell between 9-11 h for the 2-hour lag, and 11-13 h for the 4-hour
lag. Stevens Creek showed a 2-hour lag with peak air heating two hours later than
on Barney Creek (9-11 h) and peak water heating between 11-13 h. Lags between
air and water peak cooling periods weren't particularly evident on either creek.
Middle elevations on Barney Creek at times showed water cooling commencing one
period before air cooling, and occasionally air cooling began one period prior to
water cooling at lower elevations. No lags or patterns in cooling were apparent on
Stevens Creek.
81
Most of the patterns of heating and cooling appeared to be attributable to a number
of factors, including the drainage basin exposure, patterns of air movement, and the
amount of exposure time. On these two creeks, air temperatures were responsive to
weather patterns and the radiation absorption characteristics of their respective
drainage basins, with rapid changes often evident. Water temperatures are buffered
from rapid fluctuations due to mass and specific heat characteristics (Ward 1985),
and relate more to the generalities of air temperatures than patterns of rapid
fluctuation. In this study the data support the findings of several researchers that
smaller streams have less thermal buffering capacity and are more responsive to the
surrounding thermal environment than larger streams (Stefan and Preud'homme
1997; Sjnokrt and Stefan 1993; Smith and Lavis 1974; Swift and Messer 1971).
OBJECTIVE 3: MODELING VARIABLES STRONGLY ASSOCIATED WITH
MAXIMUM WATER TEMPERATURE.
Derivatives of air temperature were seen to have strong association with maximum
water temperature, and strong predictive ability of water temperature change. Our
findings were consistent with other researchers who have shown that air temperature
has a strong association with water temperature (Mohseni et al. 1998; Brown and
Binkley 1994; Sinokrot and Stefan 1994; Sinokrot and Stefan 1993; Smith and Lavis
1974; Levno and Rothacher 1967; Needham and Jones 1959), particularly in streams
with lower discharge (Stefan and Preud'homme 1993; Smith and Lavis 1974).
Maximum air temperature showed a very strong association with maximum water
temperature on all elevations of Barney Creek and at the lowest elevation of Stevens
Creek. At the upper three elevations of Stevens Creek however, minimum air
temperature from the day before dominated as the variable most strongly related to
maximum water temperature. Each of these two air temperature variables help to
82
establish the thermal gradient on a particular stream. Whichever one is more
dominant in terms of influencing the thermal gradient will be seen to have the
strongest relationship with maximum water temperature at a particular elevation.
OBJECTIVE 4: DEVELOPING A PREDICTIVE EQUATION FOR WATER
TEMPERATURE CHANGE.
Five variables were identified and tested in their ability to predict water temperature
change in the current day. They included: net air temperature change for the current
day, maximum air temperature for the current day, minimum air temperature for the
current day, minimum water temperature from the day before, and water temperature
change from the day before. R2s ranged from 0.69-0.9 1. The strong fit of our model
across both creeks and all elevations as well as its ability to predict accurately within
0.5 - 1°C indicated that these variables are useful in determining the potential water
temperature change in a given day in this system.
Air temperature variables have a direct effect on the thermal cycling evident on a
particular creek. What is generally evident is that maximum air temperature has a
greater influence on thermal cycling at lower elevations, and minimum air
temperature tends to have a greater effect at higher elevations. Thus these two
variables have a strong impact on water temperature change.
Water temperature variables have a strong influence over water temperature change,
particularly minimum water temperature, as it is the 'start point' for any temperature
change. That is, a lower initial water temperature results in an increased thermal
gradient and a greater rate of temperature change. This is consistent with our
findings that as minimum water temperature increases, water temperature change
decreases.
83
Thus while each creek is different and responds to its thermal environment, certain
variables have strong association, regardless of differences between creeks.
84
BIBLIOGRAPHY
Adams, T., and K. Sullivan. 1989. The physics of forest stream heating; a simple
model. Timber, Fish and Wildlife. TFW-WQ3-90-007.
Ahrens, C.D. 2001. Essentials of Meteorology: An Invitation to the Atmosphere
(3th
edition). Brooks/Cole-Thomson Learning, Canada.
Ahrens, C.D. 1998. Essentials of Meteorology: An Invitation to the Atmosphere
(21 edition). Wadsworth Publishing Company, Belmont, CA.
Amaranthus, M., H. Jubas, and D. Arthur. 1989. USDA Forest Service Gen.
Tech. Rep. PSW-109. p. 75-78.
Beschta, R.L. 1997. Riparian shade and stream temperature: An alternative
perspective. Rangelands, 19(2) :25-28.
Beschta, R.L., and R.L. Taylor. 1988. Stream temperature increases and land use
in a forested watershed. Water Resources Research. 24(1):19-25.
Brown, G.W. 1969. Predicting temperatures of small streams. Water Resources
Research. 5(1):68-75.
Brown, G.W., and J.T. Krygier. 1970. Effects of clear-cutting on stream
temperature. Water Resources Research. 6(4): 1133-1139.
Brown, G.W., and J.T. Krygier. 1967. Changing water temperatures in small
mountain streams. J. of Soil and Water Conservation. November-December,
pp 242-244.
Brown, G.W., G.W. Swank and J. Rothacher. 1971. Water temperature in the
Steamboat drainage. USDA Forest Service, Pacific Northwest Forest and Range
Experiment Station, Research Paper PNW-119. pp. 15.
Brown, T.C., and D. Binkley. 1994. Effect of management on water quality in
North American forests. USDA Forest Service, Rocky Mountain Forest and Range
Experiment Station, Gen. Tech. Rept. RM-248. Fort Collins, CO, USA pp 1-27.
85
Buttoiph, L.J. 1955. Practical applications and sources of ultraviolet energy. in
Hollaender, A. 1955. Radiation Biology, Vol. II: Ultraviolet and Reflected
Radiation. McGraw-Hill Book Company Inc. New York, Toronto, London. (pp 7778, 115-116).
Campbell, N.A. 1993. Biology. The BenjaminlCummings Publishing Co.,
Redwood City, CA. pp. 1190+.
Castro, N.M., and G.M. Hornberger. 1991. Surface-subsurface water interactions
in an alluviated mountain stream channel. Water Resources Research. 27(7):16131621.
Constanz. J. 1998. Interaction between stream temperature, streamfiow, and
groundwater exchanges in alpine streams. Water Resources Research. 34(7):16091615.
Dingman, S.L. 1971. Equilibrium temperatures of water surfaces as related to air
temperature and solar radiation. Water Resources Research. 8(1):42-48.
Dong, J., J. Chen, K.D. Brosofske, and R.J. Naiman. 1998. Modeling air
temperature gradients across managed small streams in western Washington. Journal
of Environmental Management. 53:309-321.
Edinger, J.E., D.W. Duttweiler, and J.C. Geyer. 1968. The response of water
temperatures to meteorological conditions. Water Resources Research. 4(5):11371143.
Feller, M.C. 1981. Effects of clearcutting and slashburning on stream temperature
in southwestern British Columbia. Water Resources Bulletin. 17(5):863-867.
Fowler, W.B., J.D. Helvey, and C. Johnson. 1979. Baseline climatic and
hydrologic relationships for the High Ridge evaluation area. USDA Forest Service,
Pacific Northwest Forest and Range Experiment Station, Gen. Tech. Rept. PNW-91.
pp 1-16.
Gray, J.R.A., and J.M. Edington. 1969. Effect of woodland clearance on stream
temperature. J. Fish. Res. Bd. Canada. 26:399-403.
Heede, B.H., and J.N. Rinne. 1990. Hydrodynamic and fluvial morphologic
processes: Implications for fisheries management and research. North American
Journal of Fisheries Management. 10(3):249-267.
86
Hidore, J.J., and J.E. Oliver. 1993. Climatology: An Atmospheric Science.
MacMillan Publishing Co., New York, NY. pp 423.
Hollaender, A. 1956. Radiation Biology. Vol. 1-3. McGraw Hill, New York.
Hoitby, L.B. 1988. Effects of logging on stream temperatures in Carnation Creek
British Columbia, and associated impacts of the Coho salmon (Oncorhyncus
kisutch). Can. J. Fish. Aquat. Sci. 45:502-5 13.
Kreith, F. 1958. Principles of Heat Transfer. Haddon Craftsmen, Inc., Scranton
Pennsylvania. pp 553.
Krueger, W.C. 2001 pers. comm.. Dept. of Rangeland Resources, 200A Strand Ag
Hall, Oregon State University, Corvallis, OR (541-737-1615) Unpublished Data.
Larson, L.L, and P.A. Larson. 1997. The natural heating and cooling of water.
Rangelands. 19(6):6-8.
Larson, L.L, and Larson, S.L. 1996. Riparian shade and stream temperature: A
perspective. Rangelands, 18(4): 149-152.
LeBlanc, R.T., R.D. Brown, and J.E. FitzGibbon. 1997. Modeling the effects of
land use change on the water temperature in unregulated urban streams. Journal of
Environmental Management. 49:445-469.
Levno, A., and J. Rothacher. 1967. Increases in maximum stream temperatures
after logging in old-growth Douglas fir watersheds. USDA Forest Service, Pacific
Northwest Forest and Range Experiment Station, Portland OR. Research Note,
PNW-65,pp. 12.
Lynch, J.A., G.B. Rishel, and E.S. Corbett. 1984. Thermal alterations of streams
draining clearcut watersheds: Quantification and biological implications.
Hydrobiologia. 111:161-169.
Mangelson, K.A. 2000. Burnt River Basin water temperature modeling study
(Final Report; Draft). US Department of the Interior, Bureau of Reclamation
Technical Service Center, Denver, CO. USA.
McRae, G., and C.J. Edwards. 1994. Thermal characteristics of Wisconsin
headwater streams occupied by beaver: Implications for Brook trout habitat.
Transactions of the American Fisheries Society. 123:641-646.
87
Meays, C. 2000. Elevation, thermal environment, and stream temperatures on
headwater streams in Northeastern Oregon. Masters Thesis, Oregon State
University, Corvallis, OR, USA.
Meehan, W.R. 1970. Some effects of shade cover on stream temperature in
Southeast Alaska. USDA Forest Service, Pacific Northwest Forest and Range
Experiment Station, Research Note, PNW-1 13. pp. 9.
Meehan, W.R., M.A. Brusven, and J.F Ward. 1987. Effects of artificial shading
on distribution and abundance ofjuvenile Chinook salmon (Oncorhynchus
tshawytscha). Great Basin Naturalist. 47(l):22-31.
Meisner, J.D. 1990. Effect of climatic warming on the southern margins of the
native range of Brook trout, Salvelinusfontinalis. Can. J. Fish. Aquat. Sci. 47:10651070.
Meisner, J.D., J.S. Rosenfeld, and H.A. Regier. 1988. The role of groundwater in
the impact of climate warming on stream salmonines. Fisheries. 13(3):2-8.
Mohseni, 0., and H.G. Stefan. 1999. Stream temperature/air temperature
relationship: a physical interpretation. Journal of Hydrology. 218:128-141.
Mohseni, 0., H.G. Stefan, and T.R. Erickson. 1998. A nonlinear regression
model for weekly stream temperatures. Water Resources Research. 34(10):26852692.
Mosley, M.P. 1983. Variability of water temperatures in the braided Ashley and
Rakaia rivers. New Zealand Journal of Marine and Freshwater Research. 17:33 1342.
Needham, P.R., and A.C. Jones. 1959. Flow, temperature, solar radiation and ice
in relation to activities of fishes in Sagehen Creek, California. Ecology. 40(3)465474.
Pluhowski, E.J. 1970. Urbanization and its effect on the temperature of streams on
Long Island, New York. Geological Survey Professional Paper 627-D.
Pluhowski, E.J. and I.H. Kantrowitz. 1963. Influence of land-surface conditions
on ground-water temperatures in southwestern Suffolk County, Long Island, New
York. U.S. Geological Survey Professional Paper 475-B. p. B186-188.
Raphael, J.M. 1962. Prediction of temperature in rivers and reservoirs. Journal of
the Power Division. P02:157-181.
88
Sellers, W.D. 1965. Physical Climatology. The University of Chicago Press.
Chicago, Illinois. (pp 37-38).
Silberberg, M. 1996. Chemistry: The Molecular Nature of Matter and Change.
Mosby-Year Book Inc. St Louis, MO.
Sinokrot, B.A., and H.G. Stefan. 1994. Water temperature sensitivity to weather
and bed parameters. Journal of Hydraulic Engineering. 120(6):722-737.
Sinokrot, B.A., and H.G. Stefan. 1993. Stream temperature dynamics:
Measurements and modeling. Water Resources Research. 29(7):2299-23 12.
Smith, K., and M.E. Lavis. 1975. Environmental influences on the temperature of
a small upland stream. Oikos. 26:228-236.
Stefan, H.G., and E.B. Preud'homme. 1993. Stream temperature estimation from
air temperature. Water Resources Bulletin. 29(1):27-45.
Strahier, A.H., and A.N. Strahier. 1992. Modern Physical Geography
Edition). John Wiley and Sons, Inc. New York, NY.
(4th
Stringham, T.K., J.C. Buckhouse and W.C. Krueger. 1996. Stream temperature
as related to subterranean irrigation, in Water Temperature Research in Rangeland
Management, Rangeland Science Series Report #3. Department of Rangeland
Resources, Oregon State University, Corvallis, Oregon.
Swift, L.W. (Jr.), and S.E. Baker. 1973. Lower water temperatures within a
streamside buffer strip. USDA Forest Service, Southeastern Forest Experiment
Station, Asheville, NC. Research Note SE-193. pp. 7.
Swift, L.W. (Jr.), and J.B. Messer. 1971. Forest cuttings raise temperatures of
small streams in the southern Appalachians. Journal of Soil and Water
Conservation. May-June 1971:111-116.
US Department of the Interior, Bureau of Reclamation. 1997. Water
Measurements Manual. US Government Printing Office, Denver, CO.
Vugts, H.F. 1974. Calculation of temperature variation of small mountain streams.
Journal of Hydrology. 23:267-278.
Walker, J.H., and J.D. Lawson. 1977. Natural stream temperature variations in a
catchment. Water Research. 11:373-377.
89
Ward, J.V. 1985. Thermal characteristics of running waters. Hydrobiologia.
125 :3 1-46.
Webb, B.W., and F. Nobilis. 1997. Long-term perspective on the nature of the airwater temperature relationship: A case study. Hydrological Processes. 11:137-147.
Zwieniecki M.A. and M. Newton. 1999. Influence of streamside cover and stream
features on temperature trends in forested streams of western Oregon. Western
Journal of Applied Forestry 14(2): 106-113.
90
APPENDICES
91
APPENDIX 1.
2-hour period temperature changes for air, Barney and Stevens Creeks, July and
August 2000 and July and August 2001.
Figure 14. 2-hour period temperature changes for air, Barney Creek July 2000. Temperature changes without a common letter
(black) are significantly different (p<O.05). Mean amount of peak heating/cooling without a common letter (starred, blue) are
significantly different (p<O.O5) across elevations.
Barney Creek 1525m, July 2000
Barney Creek 1370m, July 2000
12
10
A
Mean amount of peak, cooling: -6.89°C EF*
B
4
2
'-
.
r,)
12
10
8
Mean amount of peak, heating: 7.46°C A*
0)
=
C
00
4
2C1)
D
LI[
(n
93
-4
-6
-8
Mean amount of peak, heating: 10.59°C B*
Mean amount of peak, cooling: -7.45°C E*
Ca
D
0
A
(31
-4
E'-
(0
(J
(0
C'.)
E
F
E
F
E
-2
-4
B
B
C
D
0
(;fl
F'.)
0)
(31
-2
-4
G
E
E
F
-8
-10
H
G
period (time of day)
Barney Creek 1830m, July 2000
12
A
Mean amount of peak, heating: 10.27°C B*
6
0
C
4
Q)
a)
LI
A
8
0)
B
Mean amount of peak, heating: 7.90°C A*
10
'
Mean amount of peak, cooling: -7.61 °C E*
(0
(0
E
Barney Creek 1675m, July 2000
0
-
F
period (time of day)
8
6
4
2
c
-6
F
-10
12
10
B
Mean amount of peak, cooling: -6.38°C F*
B
C
C
2
Wa)
(0
N)
(-.3
D
-6
-8
E
D
D
-4
-6
-8
-10
-10
period (time of day)
C.)
E
G
G
E
E
F
period (time of day)
NJ
Figure 15. 2-hour period temperature changes for air, Stevens Creek July 2000. Temperature changes without a common
letter (black) are significantly different (p<O.O5). Mean amount of peak heating/cooling without a common letter (starred,
blue) are significantly different (p<O.05) across elevations.
Stevens Creek 1370m, July 2000
12
A
Stevens Creek 1525m, July 2000
12
Mean amount of peak, heating: 1043°C A*
10
8
a)
Mean amount of peak, cooling: -6.32°C E*
B
B
6
0c5'
4
cl)
G)
2
a)
0
-2
Mean amount of peak, heating: 7.49°C B*
10
C
D
D
E
.-
p
F')
'.0
(.)
0)
8
u3
6
4
- C)
2
C
=E
-4
a)
J&)
CO
(0
F
-6
F
-8
G
E
O)
CO
Mean amount of peak, cooling: -4.78°C F*
B
B
C
ELIJ
(;j
(0
-8
E
H
Mean amount of peak, heating: 6.98°C B*
10
a)
Mean amount of peak, cooling: -5.48°C FC*
A
8
A
6
B
4
C
C
A
-,o
0
D
U)
w
B
4
2
Mean amount of peak, cooling: -5.85°C E6*
B
6
c
0
-6
H
E
F
Stevens Creek 1830m, July 2000
12
Mean amount of peak, healing: 6.75°C 8*
10
-4
G
period (time of day)
Stevens Creek 1675m, July 2000
-2
F')
-10
12
2
F')
F
G
period (time of day)
G)
(0
C_J,
-6
F
G
-10
G)
F.)
(i.)
(0
4
E
C
D
0
-2
U'
A
C
C
C
0
(0
CD
(0
(0
F')
E
G
(0
C-.)
G
F
E
.2
E
-4
-6
H
F')
F
F
D
-8
-8
-10
-10
period (time of day)
period (time of day)
E
D
D
Figure 16. 2-hour period temperature changes for air, Barney Creek August 2000. Temperature changes without a common
letter (black) are significantly different (p<O.05). Mean amount of peak heating/cooling without a common letter (starred,
blue) are significantly different (p<O.05) across elevations.
Barney Creek 137Dm, August 2000
12
a
Cl)
B
8
B
4
2
10
Mean amount of peak, cooling: -7.50°C E*
D
'
0)
8
Ca
6
C
D
B
2
(0
()
0)
-4
01
E-
(131
-
(0
-2
Mean amount of peak, cooling: -8.63 °C F*
C
F
-8
F
(0
01
0)
(0
-
period (time of day)
Barney Creek 183Dm, August 2000
A
12
Mean amount of peak, heating: 11.73 °C C
10
8
Ca
6
Cn
Ca
"- 0)
2
Mean amount of peak. heating: 9.53°C B*
a)
Mean amount of peak, cooling: -7.33°C E*
0)
C
00 4
B
C
D
6
ca
C
D
w
B
C
C
D
2
E
F
0
-2
(0
(0
N)
-4
Ca
A
10
Mean amount of peak, cooling: -7.37°C E*
4
-6
-8
-10
±j
F')
(0
(1°
°)
G
E
F
0
E
F')
(0
(7,
0.)
F
H
-4
01
-
-6
-8
ri
=
c
C..)
(;)
)
C.)
F
G
H
-10
period (time of day)
E
F
period (time of day)
Barney Creek 1675m, August 2000
C
F
H
-10
12
0
C
-8
C
-10
0)
0
-6
-6
a)
D
E
Q
a)
Mean amount of peak, heating: 9.75 °C B*
A
00 4
C
E-2
C)
Mean amount of peak, heating: 8.58°C A*
A
10
00
Barney Creek 1525m, August 2000
12
period (time of day)
0
0
E
E
D
Figure 1 7. 2-hour period temperature changes for air, Stevens Creek August 2000. Temperature changes without a common
letter (black) are significantly different (p<O.O5). Mean amount of peak heating/cooling without a common letter (starred,
blue) are significantly different (p<O.05) across elevations.
Stevens Creek 137Dm, August 2000
12
A
Mean amount of peak, heating: 12.30°C A*
10
a,
C)
0&
=a,
Cl)
E
2
F
Mean amount of peak, cooling: -5.28°C F*
C)
E
D
(0
(0
(0
4
6
a
B
o5 4
0
4
(C
N)
a,
a,
C)
caiI-
p
N)
F
E
G
F
C
0
E
C;,'
'74
(0
F
-6
G
-8
-10
B
00 4
(a
-4
-6
'
E
(0
-
D
U)
a,
n
-
C.)
(0
C.)
(0
B
6
C
0c3 4
C
D
Mean amount of peak, cooling: -6.17°C E*
8
C)
C
a
B
6
E'
E
F
(;7
7.1
(0
N)
r-i
1
C;.)
N)
(0
=
(f,)
(0
E
F
D
H
G
2
E
(I,'
.iI-
-4
-6
-8
-8
-10
-10
period (time of day)
0
E
E
- C)
C.)
C.)
F
H
r1
Mean amount of peak, heating: 8.09°C ('*
A
10
Mean amount of peak, cooling: -5.55°C F*
8
0-D -2
C,'
Stevens Creek 183Dm, August 2000
12
Mean amount of peak, heating: 8.77°C BC*
A
10
Q)G)
H
C.)
period (time of day)
Stevens Creek 1675m, August 2000
12
2
G
ri
C.)
-10
period (time of day)
C)
fin
(0
N)
C.)
-4
.
G
B
0
-
E
H
2
0w
OD -2
I')
C.)
H
Mean amount of peak, heating: 9.83 °C B*
a,
C
C
6
A
10
Mean amount of peak, cooling: -6.53°C E*
B
Ca
Stevens Creek 1525m, August 2000
12
(0
(0
T
C.)
(Cr.
(,'
7.1
(0
N)
r1
r-1
C.)
N)
(0
C.)
F
H
'
H
period (time of day)
G
F
E
F
'C
Figure 1 8. 2-hour period temperature changes for air, Barney Creek July 2001. Temperature changes without a common
letter (black) are significantly different (p<O.O5). Mean amount of peak heating/cooling without a common letter (starred,
blue) are significantly different (p'<O.05) across elevations.
Barney Creek 1370m, July 2001
a)
0)
C
c
O
6
D
E
C
C
C
0
D
0
-2
PH
C,,)
0)
-4
N,)
-
(0
-6
r1
F
G
-8
-10
H
I
P1
0)
(in
F
E
G
F
B
B
4
i-n
-4
Mean amount of peak, cooling: -6.11 °C EF*
6
0
Mean amount of peak, heating: 8.77°C B*
8
B
2
- 0)
A
10
Mean amount of peak, cooling: -5.55°C E*
A
8
4
Barney Creek 1525m, July 2001
12
Mean amount of peak, heating: 6.90°C A*
12
10
C
2
C
0
El
-2
-4
C()
(0
0)
-4
G
A
8
-2
-4
-6
-8
-10
(0
N)
Li
N.)
N)
-
C,)
'J
)
(()
D
0
0
E
-8
period (time of day)
Barney Creek 1830m, July 2001
12
Mean amount of peak, heating: 10.25°C ('*
10
0
CD
E
Barney Creek 1675m, July 2001
2
-
-;'1
-10
12
4
((0
-6
period (time of day)
6
C
Mean amount of peak, cooling: -.6.76°C F*
B
C
B
C
D
C
D
0)
8
ca
6
C
05" 4
B
C
Mean amount of peak, cooling: -5.37°C E*
A
B
B
C
C
2
n
(0
C,,
-1
Mean amount of peak, heating: 6.01 °C A*
10
a,
(0
C,)
())
Cfl
(0
'-4
n ri
±J
(0
H
H
period (time of day)
)
E
F
G
0'
G
F
LI
-
P1
C,,)
C)
(in
E
0
E
Wa)
0--0
-2
-4
(0
0)
C
Li
El
((,)
((0
(7
-'4
N)
[-
El
D
D
0
C,,)
C.)
-6
-8
-10
Li
N)
E
F
period (time of day)
Figure 19. 2-hour period temperature changes for air, Stevens Creek July 2001. Temperature changes without a common
letter (black) are significantly different (p<O.O5). Mean amount of peak heating/cooling without a common letter (starred,
blue) are significantly different (p<O.O5) across elevations.
Stevens Creek 1370m, July 2001
Stevens Creek 1525m, July 2001
12
A
12
Mean amount of peak, heating: 7.91 °C A*
10
10
C
w
0)
C
Mean amount of peak, cooling: -5.73°C E*
0)
C
6
00 40
2
00 4
E
D
-
= LI
Co
T
()
C7
-
(0
r')
F'.)
H
2
0)
0
B
0
G
C.)
-4
E
G
of peak, cooling: -4.42 °C FG*
B
C
E
F
(i.)
T&)
H
Mean amount
E
7,
F'.)
G
H
G)
of peak, heating: 6.05 °C B*
A
6
c
0
E
A
Mean amount
El
Eli
C,)
CD
y
(0
(7,
F
-6
(3
-8
H
H
F
G
E
F
-10
period (time of day)
period (time of day)
Stevens Creek 1830m, July 2001
Stevens Creek 1675m, July 2001
12
12
Mean amount of peak, heating: 6.03 °C B*
C
c
0
Mean amount of peak, cooling: -4.62°C C
0)
A
6
0
C
2
E
B
C
C
0
CD
°o
w
6
2
-6
CD
C
C
0
D
El
CD
CD
-7
<C
-4
B
LIII
CD
.2
B
4
El
0
Mean amount of peak, cooling: -4.88°C EC*
A
0)
A
B
00 4
Mean amount of peak. heating: 6.14°C B*
10
10
N.)
-
C.,)
F
H
H
G
F
E
F
-4
-6
CD
(;)
(7fl
07
4
G
(5
-10
-10
period (time of day)
ED
()
(7)
CD
C.,.)
-8
-8
LI
Co
period (time of day)
F
E
F
E
0
E
Figure 20. 2-hour period temperature changes for air, Barney Creek August 2001. Temperature changes without a common
etter (black) are significantly different (p<O.05). Mean amount of peak heating/cooling without a common letter (starred,
blue) are significantly different (p<O.O5) across elevations.
Barney Creek 1370m, August 2001
Barney Creek 1525m, August 2001
w
2
- C)
Mean amount of peak, cooling: -7.22 °C E*
B
6
00 4
Eu,
.-.
A
8
)
8
C)
C
(0
C
D
F
o3
E
(0
(7,
0)
II
P
(i,)
cO
(7,
G
G
G
H
H
'H
r
Mean amount of peak, cooling: -7.30°C E*
6
B
C
4
2
-4
-6
-8
-10
.
a,
Mean amount of peak, heating: 8.90°C B*
A
10
10
C
-C
12
Mean amount of peak, heating: 7.45 °C A*
12
0
0
ri
0
-2
(7,
0)
0)
-4
F
-6
F
A
_ a,
a,a,
D
-2
B
C)
8
Ca
6
0
B
C
C
1"
ci.)
(7,
D
4
2
(7,
<0
.4
I..)
<0
-6
-8
a
Mean amount of peak, cooling: -7.05 °C E*
0)
a
Barney Creek 1830m, August 2001
12
Mean amount of peak, heating: 7.93 °C AB*
10
6
00
4
Eu
period (time of day)
Mean amount of peak. heating: 11.53°C C*
8
Wa,
0)
E
F
G
-10
period (time of day)
E
F
H
Barney Creek 1675m, August 2001
10
'
0)
E
G
-8
-10
period (time of day)
12
D
0
0
E
E
E-"
-2
(0
-6
.
D
-4
-8
-10
A
Mean amount of peak, cooling: -7.01 °C E*
B
C
D
C
D
E
E
F
L
(0
(7,
F)
7,
(C
(7,
--4
_(i.)
0)
(7,
0)
<0
I
K
PT
J
period (time of day)
H
F
I
G
G
H
Figure 21. 2-hour period temperature changes for air, Stevens Creek August 2001. Temperature changes without a common
letter (black) are significantly different (p<0.O5). Mean amount of peak heating/cooling without a common letter (starred,
blue) are significantly different (p<0.O5) across elevations.
Stevens Creek 1370m, August 2001
Stevens Creek 1525m, August 2001
12
A
12
Mean amount of peak. heating: 11.43°C A*
10
'
8
0)
Mean amount of peak, cooling: -6.33 °C E*
B
2
0
0--c
a)
C
D
C
C
D
E:J
C;,)
CD
0)
1')
CM
C')
E
-6
6
U
4
2
G
-8
E
F
B
C
E
0)
CO
C;)
(0
0
C,)
-4
,
I-
6
2
E,o
0)
0-t
E-"
-2
-6
Stevens Creek 1830m, August 2001
12
Mean amount of peak, heating: 7.60°C ('*
10
Mean amount of peak, cooling: -5.69°C F*
B
A
Mean anlount of peak, cooling: -5.89°C EF*
8
0)
B
°0 4
C
D
D
(;T
P
C,)
(,)
0)
F
B
00 4
..-
F
G
period (time of day)
Mean amount of peak. heating: 8.57°C BC*
8
0)
C-)
F
Stevens Creek 1675m, August 2001
A
N)
H
period (time of day)
10
9,
CM
-8
-10
-10
12
0
-6
E
G
B
- 0)
p
('3
CD
CM
Ca
C,
= LI
-2
Mean amount of peak, cooling: -5.62°C F*
0)
B
6
Mean amount of peak, heating: 9.12°C B*
A
10
c,
0)
CD
9,
CD
G
C')
G
2
GQ)
C')
E
F
E
D
=
LI
C,)
Cr
C,)
F
,
E
F
E
E
-4
-6
-8
-8
-10
-10
period (time of day)
C
D
C.)
()1
(0
9,
F',)
'3
CD
J
period (time of day)
H
C
F
F
H
G
C
100
APPENDIX 2.
2-hour period temperature changes for water, Barney and Stevens Creeks, July and
August 2000 and July and August 2001.
Figure 22. 2-hour period temperature changes for water, Barney Creek July 2000. Temperature changes without a common
letter (black) are significantly different (p<O.05). Mean amount of peak heating/cooling without a common letter (starred,
blue) are significantly different (p<O.O5) across elevations.
Barney Creek 1525m, August 2000
Barney Creek 1370m, August 2000
A
6
a)
5
C)
B
Mean amount of peak, heating: 5.51 °C A*
A
C)
Mean amount of peak, cooling: -2.93°C E*
C
B
-C
-c
O_ 3
C
2
C
0
2
E
ID
a)
F
F
E
0-w 0
LI
Qa) Q
Ev
-.4
.i ___-1
(0
(0
C'.)
CD
-2
F
-3
LI
-4
(0
(7,
C-)
-2
H
H
F
Ca
C..)
Mean amount of peak, cooling: -1.95°C F*
B
a)
5
C
4
0)
Mean amount of peak. heating: 2.37°C D
Mean amount of peak, cooling: -1.35°C G*
Ca
o
3
C
B
3
C.)
C
A
0
U)
C
a)
0
E
Et,
a)-
-
a)
-1
-2
-3
(0
a)
CM
C)
E-0
a)-
('7,
()
-1
a)
CD
F
I
H
G
F
E
Cc
E
F
LI
C;n
.4
-.-j
(0
((7
C..)
--4
(0
-
-3
H
I
-4
-4
period (time of day)
G
I
Barney Creek 1830m, August 2000
6
Mean amount of peak. heating: 3.37°C C*
A
J
period (time of day)
Barney Creek 1675m, August 2000
4
C-)
CD
H
-4
6
C)
-4
J
period (time of day)
5
(7,
-3
G
G
-4
a)
Mean amount of peak, cooling: -2.87°C E*
4
°.. 3
Ca
4-
Mean amount of peak, heating: 4.60°C B*
6
a)
period (time of day)
H
F
Figure 23. 2-hour period temperature changes for water, Stevens Creek July 2000. Temperature changes without a common
letter (black) are significantly different (p<O.05). Mean amount of peak heating/cooling without a common letter (starred,
blue) are significantly different (p<O.05) across elevations.
Stevens Creek 1370m, July 2000
6
Mean amount of peak, heating: 4.22°C A*
A
0)
=
4
C,
3
'L)
1
a
=&)
Ev
A
E
F
a,)
D
__-1
3
SO.)
0'O
.
Mean amount of peak, cooling: -0.60°C F*
4
0
B
C
.
0)
(a
B
Mean amount of peak. heating: 1.36°C B*
5
a,
Mean amount of peak, cooling: -1.55°C E*
B
Stevens Creek 1525m, July 2000
6
2
E
i3
E
E
E
E
C
0
0)
a,
(5
(0
(71
Ni
-
Ni
-
E
F
G
0)
-
E
F
0
-3
-4
c-n
F
G
G
G
-4
period (time of day)
period (time of day)
Stevens Creek 1675m, July 2000
6
Stevens Creek 1830m, July 2000
6
Mean amount of peak. heating: 1.05°C BC*
5
0)
((fl
((s)
(0
-1
(.0
B
C
C,
Mean amount of peak, cooling: -0.49°C F*
4
Mean amount of peak, heating: 0.91 °C C
5
a,
Mean amount of peak, cooling: -0.34°C G*
4
CD
o
3
C
E
A
3
A
B
B
B
a,
w1
D
E
01
((b)
Ni
Ni
E
F
0)
F
E
)
-
a,
F
E
(0
(0
01
C
D
o.w0
A
-
(i,)
(II
-'
--4
-(0
-(0
Ni
Ni
-
Ni
F
F
F
.4
-4
period (time of day)
Ni
-
()
C.)
F
F
0)
71
0)
F
F
period (time of day)
-
F
Figure 24. 2-hour period temperature changes for water, Barney Creek August 2000. Temperature changes without a
common letter (black) are significantly different (p<O.O5). Mean amount of peak heating/cooling without a common letter
(starred, blue) are significantly different (p<O.O5) across elevations.
Barney Creek 1525m, August 2000
Barney Creek 1370m, August 2000
A
6
a
C,
5
Mean amount of peak, heating: 5.51 °C A*
B
A
Mean amount of peak, cooling: -2.93°C E*
u,_._
o__ 3
Q-, 0
.2
U)
2
C).CØ)
0
D
F
E
(0
-.4
-i
(0
(;)
(.4
-.4
-2
-3
-4
F
F
H
F
;
-2
CC
-3
G
G
H
D
E
'-1
I'.)
('7,
-4
(0
(0
C,)
A
4
Mean amount of peak, cooling: -1.95°C F*
B
3
1
0)
C.)
C
__-1
D
U)
0)
A
Mean amount of peak, cooling:
-1.35°C (*
C
E
F
Wo
-.4
(0
F
CC
-3
3
H
-4
period (time of day)
G
F
E
1
(7,
.2
(.)
a)
B
02
- a)Wi
D
E
4
C
D.C)Q
,
Mean amount of peak, heating: 2.37°C D*
0)
C
.
C)
F
Barney Creek 1830m, August 2000
6
Mean amount of peak, heating: 3.37°C ('*
CC
C.)
0
H
period (time of day)
Barney Creek 1675m, August 2000
5
(7,
Ni
-4
-4
6
C)
(0
J
period (time of day)
C)
Mean amount of peak, cooling: -2.87°C E*
C
wo
C
2
U)
C)
B
4
CC
.-.
Mean amount of peak, heating: 4.60°C B*
6
a)
0)
C
(0
-2
-3
-4
('7,
(.4
D
(4)
(7,
(0
G
H
I
period (time of day)
H
G
F
Figure 25. 2-hour period temperature changes for water, Stevens Creek August 2000. Temperature changes without a
common letter (black) are significantly different (p<0.05). Mean amount of peak heating/cooling without a common letter
(starred, blue) are significantly different (p<O.05) across elevations.
Stevens Creek 1370m, August 2000
Stevens Creek 1525m, August 2000
6
A
6
Mean amount of peak, heating: 5.23°C A*
5
a,
C)
Mean amount of peak, cooling: -1.64°C E*
4
'5
B
O2
4
C)
3
A
(0
-j
-4
(0
N)
N)
F
G
H
H
G
D
F
C-)
2
E
(0
-
()
G
F
G
F
G
-4
period (time of day)
Stevens Creek 1675m, August 2000
Stevens Creek 1830m, August 2000
6
6
Mean amount of peak, heating: 1.22°C ('*
0,
Mean amount of peak, cooling: -0.52°C G*
4
Mean amount of peak. heating: 1.25°C C*
5
a,
4
Mean amount of peak, cooling: -0.37°C H*
3
0
O2
A
cn
F
G'l
G
a,
C
O2
B
C
D
.
C,
-4
(0
-2
-3
1
E
F
D
B
A
C
-
E
0.w0
(0
C',
a'
a,
4
D
period (time of day)
(5
(31
i3
G
-4
0)
C
E
(0
C-)
F
-3
B
G
D
-2
Mean amount of peak, cooling: -0.64°C F*
F
U)
D
.___-1
c
2O
C
PD,
Mean amount of peak, heating: 1.57 °C I3
a,
0)
Cfl
-
(0
(P
N)
-
N)
()
F
E
G
G
-
G
0
(C"
G
E
F
-2
-3
01
(0
(31
-4
-4
(0
G
N)
N)
N)
(C-'
c)
G
F
H
H
H
G
-
H
-4
-4
period (time of day)
-
(
N)
period (time of day)
C,)
-
C,)
',
E
F
Figure 26. 2-hour period temperature changes for water, Barney Creek July 2001. Temperature changes without a common
letter (black) are significantly different (p<O.O5). Mean amount of peak heating/cooling without a common letter (starred,
blue) are significantly different (p<O.05) across elevations.
Barney Creek 1370m, July 2001
Barney Creek 1525m, July 2001
6
C)
C)
C.)
6
Mean amount of peak, heating: 4.53°C A*
5
C)
A
4
o
C
B
B
3
C
G)i
C)
C
B
a-C)
(0
-1
N)
C')
C.)
cb
.&
(7'
C.)
-2
--4
N)
C')
01
0
D
E
F
F
G
F
F
F
F
E-ø
C)
CD
-3
LII
(0
-.1
0
=
4
C)
G
C.)
a)
C)
C
'V
B
Mean amount of peak, heating: 2.26°C D*
5
Mean amount of peak, cooling: -1.22°C G*
4
A
3
C.)
B
C
C
DCl)
Hi
E0
C)
C)
C5
-3
C
C)
.- Wi
D
El
-4
-2
C
(P
CD
C.)
01
-
G
C.)
a)
C.)
G
F
G
i
.2
E
0
D
F
E
F
CC
C
0
B
El
o-Wo
E
m
N)
N)
CD
CD
'P
(I.)
C.)
(11
CD
C7fl
-I
N)
CD
period (time of day)
N)
-3
(;)
N)
F
H
H
0
H
period (time of day)
(,)
C,)
-
-4
-4
F
0
Barney Creek 1830m, July 2001
6
Mean amount of peak, cooling: -1.69°C F*
3
F
3
H
I
period (time of day)
Mean amount of peak. heating: 2.96°C C*
A
(0
H
-4
Barney Creek 1675m, July 2001
5
--4
N)
-3
period (time of day)
C)
N.)
CD
CD
-2
-4
6
Mean amount of peak, cooling: -2.46°C E*
0
B
0
A
4
Mean amount of peak, cooling: -2.44 °C E*
3
Mean amount of peak, heating: 4.01 °C B*
5
0)
A
0
F
E
F
Figure 27. 2-hour period temperature changes for water, Stevens Creek July 2001. Temperature changes without a common
letter (black) are significantly different (p<O.O5). Mean amount of peak heating/cooling without a common letter (starred,
blue) are significantly different (p<O.O5) across elevations.
Stevens Creek 1370m, July 2001
6
A
0)
0)
4
Mean amount of peak, heating: 1.35°C B*
Mean amount of peak, cooling: -1.45°C E *
Mean amount of peak, cooling: -0.55 °C F*
B
3
C.,
C
U)
.
a,
E
ID
1
a0
Stevens Creek 1525m, July 2001
Mean amount of peak, heating: 4.28°C A*
A
0
ID
E
E
(P
.2
C
CD
CD
2
-J
i3
(.)
CD
G
G
G
F
E
E
G
F
F
G
-4
period (time of day)
O2
Cl)
E
C
<,,
-
B
-3
a,
C
4
o
3
a,
B
A
D
E
B
4-
cp
;-J
-2
Mean amount of peak, heating: 1.91 °C B*
Mean amount of peak, cooling: -0.45°C F*
4
CD
-4
-
N)
CD
-
ID
0
E
-4
)
()
E
Mean amount of peak, cooling: -0.32°C 6*
O2
wi
C,
A
B
ID
C
C
0
=
EJ
((fl
(i.)
N)
E
C
Stevens Creek 1830m, July 2001
6
Mean amount of peak, heating: 1.14 °C B*
5
C,
C
period (time of day)
Stevens Creek 1675m, July 2001
6
a,
B
-.1
Ci.)
CD
B
C
E
CD
I-
-1
0
-2
E
-3
-J
(n
CD
C.)
CD
.)
.7.)
CD
E
F
F'.)
F')
-
N)
-
F
F
-4
period (time of day)
EJ
F')
(p
C.)
-
0
0
E
E
C.)
D
F
period (time of day)
o
Figure 28. 2-hour period temperature changes for water, Barney Creek August 2001. Temperature changes without a
comnion letter (black) are significantly different (p<O.O5). Mean amount of peak heating/cooling without a common letter
(starred, blue) are significantly different (p<O.05) across elevations.
Barney Creek 1370m, August 2001
6
U)
6
Mean amount of peak, heating: 580°C A*
B
5
a,
Barney Creek 1525m, August 2001
A
Mean amount of peak, cooling: -289°C E*
°_...
C
3
w
C
(
4
U_
3
A
B
Mean amount of peak, cooling: -2.91 °C E*
C
2
D
4-
F
E
a)
1
C
E
D
&a)0)0
Q.0 0
Ea
CP
(i.)
CD
C.)
-2
-
(11
C.)
C.)
CO
CD
r')
1'.)
CO
a
0
-4
F
H
period (time of day)
Barney Creek 1675m, August 2001
Mean amount of peak, heating: 2.22°C I)*
U)
4
A
CO
C)
A
C)
C
Ca
-C
Mean amount of peak, cooling: -1.99°C F*
3
E
U
a)
- C)U)0)1
a)
Li
E
CO
i3
C)
C.)
H
B
B
C
D
CD
CO
U)
C.)
H
A
0)0
<0
CO
2
A
3
C)
D
1
Mean amount of peak. cooling: -1.30°C G*
4
B
C
.
Barney Creek 1830m, August 2001
6
Mean amount of peak, heating: 3.19°C C*
5
E
F
F
G
H
period (time of day)
6
C)
F-..)
-4
J
I
C.)
C'.)
(7,
-3
0
CD
31
CO
-2
F
H
H
L:J
-1
.2
I-
0
-3
0)
Mean amount of peak, heating: 4.75°C B*
0
F
F
F
2
C.)
CY
--4
CO
N)
E
G
G
-4
-4
period (time of day)
C.)
period (time of day)
F
F
E
E
Figure 29. 2-hour period temperature changes for water, Stevens Creek August 2001. Temperature changes without a
common letter (black) are significantly different (p<O.O5). Mean amount of peak heating/cooling without a common letter
(starred, blue) are significantly different (p<O.O5) across elevations.
Stevens Creek 1525m, August 2001
Stevens Creek 1370m, August 2001
6
A
5
C)
Co
o
B
3
0
a)
C)
E
0
D
a)
Wo
E
(P
a,
a)
(0
I.-
0
-2
Co
a)
4(5
F
-3
0
3
A
B
(I)
a)
E
-1
Mean amount of peak, cooling: -0.62 °C F*
4
(5
C
U)2
Wi
0)
C
B
Mean amount of peak, heating: 1.77°C B*
5
a)
Mean amount of peak, cooling: -170°C E*
4
C
6
Mean amount of peak. heating: 4.89°C A*
G
G
H
H
G
F
F
Wi
F
D
0)0
Li
7'
=
-1
-)
(0
C)
J
C.)
H
-2
(p
Ni
-
Ni
C.)
F
F
G
G
-
()
F
F
G
G
(f,,
G
Stevens Creek 1830m, August 2001
Mean amount of peak, heating: 1.40°C C*
0)
0)
Mean amount of peak, cooling: -0.46°C G*
4
3
C)
02
C
B
Mean amount of peak, cooling: -0.35 °C 11*
4
3
A
C
E
F
B
D
E
Mean amount of peak, heating: 1.27 °C (:
5
A
a)
aQ
-(0
E
6
0)
! wi
-;J
period (time of day)
Stevens Creek 1675m, August 2001
4-
--
-4
6
0
(
(P
period (time of day)
C
C
-3
-4
0)
C
B
0
LI
C,,
CO
CO
-
C
E
Li
0)
-
-;i
(0
D
(P
Ni
-
E
Ni
0)
(7'
-
(.)
E
E
N.)
(0
-(0
(P
Ni
G
H
H
E
-4
period (time of day)
0)
F
period (time of day)
G
H
G
F
C
H
G
H
109
APPENDIX 3
Tables of variables strongly associated with maximum water temperature for both
creeks in 2000 and 2001.
110
Table 20: Table of variables strongly associated with maximum water temperature,
for Barney Creek, year 2000.
Barney 1370-rn, July 2000
Variable
max air temp, current day
water cooling, prey, day
mm. water temp, prey, day
Barney 1370-rn, August 2000
Variable
max air temp. current day
mm. water temp, prey, day
Barney 1525-rn, July 2000
Variable
max air temp. current day
air cooling, prey, day
Barney 1525-rn, August 2000
Variable
max air temp, current day
mm water temp, prey, day
Barney 1675-rn, July 2000
Variable
max air temp, current day
air cooling, prey, day
peak air heating change, current day
Barney 1675-rn, August 2000
Variable
max air temp, current day
mm water temp, prey. day
Barney 1830-rn, July 2000
Variable
max air temp, current day
peak air heating change, current day
air cooling, prey. day
Barney 1830-rn, August 2000
Variable
max air temp, current day
mm water temp. prey, day
Partial R2
Prob F
.84 14
.0446
.0243
<.0001
.0077
.0264
Partial R2
Prob F
.7803
.0796
<.000 1
.0011
Partial R2
Prob F
.8472
.0296
<.0001
Partial R2
Prob F
.8200
.0723
<.000 1
Partial R2
Prob F
.8648
<.000 1
.04 12
.0169
.0052
.0438
Partial R2
Prob F
.8248
.0926
<.000 1
<.000 1
Partial R2
Prob F
.8954
.0351
<.000 1
.003 0
.03 13
.0005
.03 04
.0006
Partial R2
Prob F
.8264
.1149
<.0001
<.000 1
111
Table 21: Table of variables strongly associated with maximum water temperature,
for Stevens Creek, year 2000.
Stevens 1370-rn, July 2000
Variable
max air temp, current day
mm water temp, prey, day
peak air heating change, current day
Stevens 1370-rn, August
Variable
max air temp, current day
mm water temp, prey, day
Stevens 1525-rn, July 2000
Variable
max air temp, current day
mm water temp. prey, day
water cooling, prey, day
Stevens 1525-rn, August
2000
.
Variable
air temp, prey, day
max air temp, current day
peak air heating change, current day
mm water temp, prey, day
REMOVED mm air temp. prey. day
water cooling, prey, day
mm
Stevens 1675-rn, July 2000
Variable
max air temp. current day
mm water temp, prey, day
Stevens 1675-rn, August
2000
.
Variable
mm air temp, prey, day
max air temp, current day
Stevens 1830-rn, July 2000
Variable
max air temp. current day
mm water temp, prey. day
Stevens 1830-rn, August
Variable
mm air temp. prey, day
max air temp. current day
mm water temp, prey, day
REMOVED mm air temp, prey, day
water cooling, prey, day
Partial R2
Prob F
.87 12
.03 87
<.000 1
.0260
.0056
.0082
Partial R2
Prob F
.7670
.1372
<.000 1
Partial R2
Prob F
.9048
.0796
.0031
<.0001
<.0001
<.000 1
.0349
2
Partial R
Prob F
.8272
.1092
.0166
.0098
.0056
.0216
<.000 1
Partial R2
Prob F
.9054
.0723
<.0001
<.0001
Partial R2
Prob F
.8535
.0824
<.0001
<.0001
Partial R2
Prob F
.9006
.0779
<.0001
Partial R2
Prob F
.8 147
<.0001
<.0001
.0040
.0948
.0039
.1088
.0235
.0070
.0 186
<.0001
.0087
.0244
.0799
<.000 1
<.000 1
112
Table 22: Table of variables strongly associated with maximum water temperature,
for Barney Creek, year 2001.
Barney 1370-rn, July 2001
Variable
Partial R2
Prob F
.8240
<.0001
Partial R2
Prob F
max air temp, current day
.6090
<.000 1
Variable
max air temp, current day
Variable
Partial R2
Prob F
.8285
<.0001
Partial R2
.4670
Partial R2
Prob F
.8456
<.000 1
Partial R2
Prob F
.6402
.0771
<.0001
Partial R2
Prob F
.8695
.0223
<.000 1
Partial R2
Prob F
.6683
.1482
<.000 1
max air temp, current day
Barney 1370-rn, August 2001
Barney 1525-rn, July 2001
Barney 1525-rn, August 2001
Variable
max air temp, current day
Barney 1675-rn, July 2001
Variable
max air temp, current day
Barney 1675-rn, August 2001
Variable
max air temp, current day
air cooling, prey, day
Barney 1830-rn, July 2001
Variable
max air temp, current day
peak air heating change, current day
Barney 1830-rn, August 2001
Variable
max air temp, current day
mm water temp, prey, day
<.000 1
Prob F
.0 172
.0359
.0002
113
Table 23: Table of variables strongly associated with maximum water temperature,
for Stevens Creek, year 2001.
Stevens 1370-rn, July 2001
Variable
max air temp. current day
Stevens 1370-rn, August
Variable
max air temp. current day
water cooling, prey, day
mm air temp. prey, day
Stevens 1525-rn, July 2001
Variable
max air temp, current day
peak air heating change, current day
mm water temp, prey. day
Stevens 1525-rn, August
Variable
mm air temp. prey, day
max air temp, current day
mm water temp, prey, day
peak air heating change, current day
REMOVED mm air temp, prey, day
Stevens 1675-rn, July 2001
Variable
max air temp, current day
mm air temp, prey, day
Stevens 1675-rn, August
Variable
mm air temp, prey, day
max air temp, current day
mm water temp, prey, day
Stevens 1830-rn, July 2001
Variable
max air temp, current day
peak air heating change, current day
mm water temp. prey, day
Stevens 1830-rn, August
Variable
max air temp, current day
mm water temp, prey, day
air cooling, prey. day
peak air heating change, current day
Partial R2
Prob F
.7906
<.0001
Partial R2
Prob F
.465 8
.0003
.0179
.0319
.1334
.0882
Partial R2
Prob F
.754 1
.042 7
<.0001
<.0001
.0001
Partial R2
Prob F
.7 114
.1841
<.000 1
.0366
.00 18
.0101
.008 0
.0633
Partial R2
Prob F
.7899
.1042
<.0001
<.0001
Partial R2
Prob F
.7068
.1720
<.000 1
.1573
<.0001
.095 5
<.0001
.026 1
.0 194
Partial R2
Prob F
.7034
<.000 1
.12 10
.0004
.0296
.04 15
Partial R2
Prob F
.5474
<.000 1
.3531
.0 157
.0 180
<.0001
.0493
.0226
114
APPENDIX 4
Scatterplots showing the fit of predicted versus observed values for both creeks in
2000 and 2001. Dashed lines indicate fit within 0.5°C and solid outer lines indicate
fit within 1°C.
Figure 30. Scatterplots demonstrating fit of prediction model for Barney Creek 2000; dashed lines indicate agreement within
0.5° C, solid outer lines indicate agreement within 1 ° C.
Scatterplot of predicted vs observed values testing
fit of prediction model for Barney 1370m, yr 2000.
--
predicted
Scatterplot of predicted vs observed values testing
fit of prediction model for Barney 1675m, yr 2000.
__-
Scatterplot of predicted vs observed values testing
fit of prediction model for Barney 1525m, yr 2000.
predicted
__-
Scatterplot of predicted vs observed values testing
fit of prediction model for Barney 1830m, yr 2000.
3
predicted
predicted
Figure 31. Scatterplots demonstrating fit of prediction model for Stevens Creek 2000; dashed lines indicate agreement within
0.5° C, solid outer lines indicate agreement within 10 C.
Scatterplot of predicted vs observed values testing
fit of prediction model for Stevens 1370m, yr 2000.
__-
Scatterplot of predicted vs observed values testing
fit of prediction model for Stevens 1525m, yr 2000.
.
picted
-
Scatterplot of predicted vs observed values testing
fit of prediction model for Stevens 1675m, yr 2000.
predicted
predicted
__-
Scatterplot of predicted vs observed values testing
fit of prediction model for Stevens 1830m, yr 2000.
predicted
Figure 32: Scatterplots demonstrating fit of prediction model for Barney Creek 2001; dashed lines indicate agreement within
0.5° C, solid outer lines indicate agreement within jO C.
Scatterplot of predicted vs observed values testing
fit of prediction model for Barney 1370m, yr 2001.
predicted
__-
Scatterplot of predicted vs observed values testing
fit of prediction model for Barney 1675m, yr 2001.
predicted
__-
Scatterplot of predicted vs observed values testing
fit of prediction model for Barney 1525m, yr 2001.
predicted
__-
Scatterplot of predicted vs observed values testing
fit of prediction model for Barney 1830m, yr 2001.
predicted
Figure 33: Scatterplots demonstrating fit of prediction model for Stevens Creek 2000; dashed lines indicate agreement within
0.5° C, solid outer lines indicate agreement within 10 C.
Scatterplot of predicted vs observed values testing
fit of prediction model for Stevens 1370m, yr 2000.
predicted
Scatterplot of predicted vs observed values testing
fit of prediction model for Stevens 1675m, yr 2000.
predicted
Scatterplot of predicted vs observed values testing
fit of prediction model for Stevens 1525m, yr 2000.
--predicted
Scatterplot of predicted vs observed values testing
fit of prediction model for Stevens 1830m, yr 2000.
predicted
119
APPENDIX 5
Tables showing fit of the model to predict maximum water temperature change, by
elevation, for both creeks in 2000 and 2001.
120
Table 24: Barney Creek 2000, showing fit of model, mean absolute deviation in °C
(with SDs in brackets), and variable coefficients (with SEs in brackets).
Elevation (m)
R2 (%)
Mean absolute deviation (°C)
Variable
Coefficient
Barney 1370
87.58
0.3977 (0.2777)
Intercept
airday
maxair
minair
0.5667 (0.672 1)
0.3388 (0.0340)
0.0493 (0.0238)
0.0962 (0.0720)
-0.03 14 (0.1182)
-0.2567 (0.096 1)
0.9665 (0.7063)
0.1981 (0.0330)
0.0879 (0.0309)
0.1535 (0.0497)
-0.2725 (0.1142)
-0.4680 (0.0823)
1.5077 (0.6896)
0.1434 (0.0337)
0.0511 (0.0335)
0.1936 (0.0532)
-0.3605 (0.1326)
-0.4951 (0.0905)
1.4141 (0.5667)
0.1244 (0.0324)
0.0601 (0.0301)
0.1557 (0.0480)
-0.3571 (0.1337)
-0.4670 (0.0878)
watdayl
Barney 1525
79.31
0.4146 (0.3350)
minwat 1
intercept
airday
maxair
minair
watday_1
Barney 1675
74.51
0.4096 (0.3274)
Barney 1830
75.37
0.3810 (0.2856)
minwat 1
Intercept
airday
maxair
minair
watday 1
minwati
Intercept
airday
maxair
minair
watdayl
minwatl
121
Table 25: Barney Creek 2001, showing fit of model, mean absolute deviation in °C
(with SDs in brackets), and variable coefficients (with SEs in brackets).
Elevation m
R2 %
Mean absolute deviation (°C)
Barney 1370
Variable
Intercept
Airday
maxair
minair
watdayl
Barney 1525
75.55
0.5118 (0.4863)
minwat 1
Intercept
airday
maxair
minair
watdayl
minwatl
Barney 1675
83.67
0.3671 (0.368 1)
Interce.t
Coefficient
1.3562 (0.7637)
0.4053 (0.0330)
0.0 140 (0.0210)
0.2103 (0.0703)
-0.1711 (0.0957)
-0.3222 (0.0883)
1,5416 (1.0033)
0.2046 (0.0366)
0.0620 (0.0285)
0.1821 (0.0603)
-0.3410 (0.1187)
-0.4602 (0.0978)
0.43 10 (0.8775)
maxair
minair
watdayl
minwat I
Barney 1830
81.30
0.3738 (0.3464)
Intercept
airday
maxair
minair
watday 1
minwat I
I s':
1.7623 (0.8013)
0.2063 (0.0354)
0.0419 (0.0233)
0.1540 (0.05 13)
-0.1832 (0.1225)
-0.4223 (0.0930)
122
Table 26: Stevens Creek 2000, showing fit of model, mean absolute deviation in °C
(with SDs in brackets), and variable coefficients (with SEs in brackets).
Elevation (m)
Stevens 1370
R2 (%)
68.66
Mean absolute deviation (°C)
Variable
0.6144 (0.8436)
Stevens 1525
89.43
0.2296 (0.2000)
Intercept
airday
maxair
minair
watday 1
minwat 1
Intercept
airday
maxair
minair
watday 1
minwatl
Stevens 1675
82.70
0.2436 (0.1837)
Intercept
airday
maxair
minair
watdayl
Stevens 1830
83.01
0.1192(0.1179)
minwat 1
Intercept
airday
maxair
minair
watdayl
minwat 1
Coefficient
-0.1274 (0.7465)
0.1926 (0.0374)
0.1633 (0.0387)
0.0783 (0.0437)
-0.2784(0.1168)
-0.4840 (0.1011)
0.95 19 (0.3737)
0.2292 (0.0247)
0.0538 (0.0184)
0.1780 (0.0447)
-0.1884(0.1043)
-0.3823 (0.0633)
1.0889 (0.3978)
0.1599 (0.0243)
0.0281 (0.0189)
0.1443 (0.0387)
-0.1851 (0.1239)
-0.3360 (0.0702)
1.3629 (0.3681)
0.0885 (0.0146)
0.0148 (0.010 1)
0.0709 (0.0218)
-0.1650 (0.1301)
-0.3398 (0.0811)
123
Table 27: Stevens Creek 2001, showing fit of model, mean absolute deviation in °C
(with SDs in brackets), and variable coefficients (with SEs in brackets).
Elevation (m)
Stevens 1370
R2 (%)
91.36
Mean absolute deviation (°C)
0.3396 (0.3171)
Variable
Coefficient
Intercept
airday
maxair
minair
1.4561 (0.6548)
0.4081 (0.0357)
0.0204 (0.0189)
0.2282 (0.0658)
-0.1908 (0.0925)
-0.34 18 (0.0860)
2.0168 (0.5559)
0.2266 (0.0260)
0.0251 (0.0142)
0.2013 (0.0433)
-0.1 878 (0.0972)
-0.4064 (0.0675)
1.6445 (0.4333)
0.1765 (0.0240)
0.0330 (0.0130)
0.1473 (0.03 14)
-0.2038 (0.0957)
-0.3791 (0.0617)
1.5149 (0,3327)
0.1053 (0.0126)
0.0011 (0.0069)
0.0860 (0.0173)
watday_l
minwat 1
Stevens 1525
88.14
0.2781 (0.2536)
Stevens 1675
88.10
0.2345 (0.2006)
Intercept
airday
maxair
minair
watday_1
minwatl
Intercept
airday
maxair
minair
watdayl
minwatl
Stevens 1830
87.40
0.1305 (0.0869)
Intercept
airday
maxair
minair
watday 1
minwatl
-0.1258(0.1017)
-0.3077 (0.0583)