LAKE TAHOE ENVIRONMENTAL IMPROVEMENT PROGRAM
PROJECT PROPOSAL
Project Name: Monitoring Past, Present, and Future Water Quality Using Remote Sensing
Principal Investigator:
Dr. S. Geoffrey Schladow
Institution:
University of California, Davis
Co-Principal Investigator:
Dr. Simon J. Hook
Institution:
Jet Propulsion Laboratory (NASA/JPL)
Address:
Tahoe Environmental Research Center
University of California, Davis
One Shields Avenue
Davis, California 95616
Address:
Jet Propulsion Laboratory
M/S 183-501
4800 Oak Grove Drive
Pasadena CA 91109-8099
Phone: (818) 354-0974
Phone: (530) 752-6932
Fax: (530) 752-7872
Fax: (818) 354-0966
Email Address: simon.j.hook@jpl.nasa.gov
Email Address: gschladow@ucdavis.edu
Theme: Cross Cutting All Science Areas
Sub-theme: B. Evaluation of Basin Monitoring and Networks.
Justification Statement
Remote sensing (RS) technologies are being used worldwide to monitor aquatic,
atmospheric, and terrestrial systems. RS provides an instantaneous synoptic overview of these
systems which when coupled with in situ measurements, can provide unmatched, cost-effective,
quantitative measures of change with spatial resolutions down to the order of meters and temporal
frequency better than daily. With a vast repository of RS data extending back almost 30 years, and the
guarantee of future data availability, RS data provide the means to understand past changes in the
Tahoe Basin, to evaluate the impacts of current activities, and to measure the effectiveness of future
management strategies.
Since 1999 NASA has operated a validation site at Lake Tahoe, based on data collected from
a network of 4 buoys in concert with U.S., Japanese, and European satellites. The statistical accuracy
of the data collected through this system has been tested and found to be extremely high (Hook et al.
2003; Hook et al. 2004; Hook et al. 2005). While the data have been used for monitoring change in
other systems, such as the Great Lakes, there has only been limited use of RS in the Tahoe basin
restricting our understanding of the complex interplays within Lake Tahoe and associated watershed.
The purpose of this project is to establish a basin-wide RS monitoring network for the Lake
Tahoe basin. The establishment of this system will have immediate impacts for many of the Themes
addressed by this RFP, and with further development could address the entire range of
environmental monitoring needs. This first stage will focus on the lake itself (rather than terrestrial or
atmospheric monitoring), and will establish a semi-automated system to evaluate and compare
discrete sources of clarity-reducing substances to the lake (streams, storm drains, intervening zones),
the effectiveness of best management practices (BMPs) at a sub-watershed scale over time, and
changes in lake clarity in both the near-shore and off-shore environments. The system will capitalize
on the local infrastructure developed by NASA together with the numerous and available satellite
datasets
1
Background/Problem Statement
The current knowledge of the decline in water clarity at Lake Tahoe CA/NV is based on the
interpretation of data from the UC Davis Tahoe Environmental Research Center’s (TERC’s) two
deep-water measurement sites, referred to as the Mid-lake and Index stations (Figure 1), the USGS
and TERC monitoring of stream conditions from approximately 20% of the streams that flow into
Lake Tahoe (e.g. Hackley et al. 2005 and Figure 2) and more recently, urban runoff measurements
(e.g. Heyvaert and Parra 2005). The in-lake measurements, made at intervals from 10-30 days, include
primary productivity, nutrient concentration (various forms of nitrogen and phosphorus), chlorophyll
concentration, light penetration, temperature distribution, and Secchi disk transparency. While these
data provide point measurements of what is happening on the lake, they are lacking in both the
temporal and spatial detail needed to understand the changes taking place at different parts of the
lake (such as the near-shore-zone), and the linkage between the lake observations and the input
sources. For example, measurements made at the two sites provide little information about how
nutrients and sediment from the streams and intervening zones are transported throughout the lake
by surface currents. Such information can be crucial in understanding the processes behind the
decline in clarity necessary to predict future changes and can be obtained from remotely sensing data
(Figure 3). Further the 10-30 day temporal sampling is insufficient to capture changes that take place
on the scale of a few days, such as wind-driven upwelling (Schladow et al., 2004; Steissberg et al.,
2005b, Figure 4), which can have a profound effect on the lake clarity measured at these stations. RS
data have the potential to provide a synoptic overview of what is happening over the lake at many
instances in time, which when coupled with in situ data, can provide lake-wide assessments of
changes in both near-shore and offshore water quality.
Through this study we will put in place a system to utilize remotely sensed and field
measurement data to quantify changes in water clarity measurements over the entire lake. Moderateresolution (1 km pixel size) satellite data are available several times per day and high-resolution (15 m
pixel size) satellite data are available every 16 days. These data can be used to create maps of water
clarity to within 15 m of the shoreline and to assess the impacts and fate of individual sources such as
a highway drain. The changes in near-shore clarity at particular areas of interest, such as near stream
inflows, can be observed in time series derived from these clarity maps.
Objective, Goals, and Hypotheses
Objective
The objective of this project is to utilize remotely sensed (satellite) data to provide a
quantitative management tool for lake-wide assessments of water quality and to link changes in water
quality to discrete sources at the sub-watershed (e.g. the Incline Creek watershed) scale.
Goals
1. Create a web-accessible repository of existing Tahoe basin RS imagery and integrate with
COMET cyber-infrastructure system being developed with NSF funding at UC Davis.
2. Develop tools for semi-automatically acquiring and storing future RS imagery
3. Refine existing algorithms for characterizing the spatial variability of nearshore water clarity
4. Refine existing algorithms for characterizing the spatial variability of mid-lake water clarity
5. Work in collaboration with TERC, DRI, and UNR researchers measuring changes in nearshore clarity to ground-truth near-shore clarity algorithms.
6. Relate near-shore spatial and temporal clarity variability to changes in stream, storm drain,
and intervening zone inputs to the lake, following storms
7. Relate mid-lake spatial and temporal clarity variability to lake mixing, following wind-driven
upwelling and surface current patterns
8. Develop a reporting system where RS-derived measures of water quality are made available
on a near-real-time basis.
2
Hypotheses
The central hypothesis of this project is that lake clarity, both near-shore and deep-water,
can be determined using remote sensing in conjunction with existing in situ measurements. These
data can be used to derive historical changes in clarity as well are provide a near-real-time measure of
clarity throughout the lake.
Other hypotheses include:
1. Nearshore clarity varies around the lake, dependent on land use, stream inflows, non-pointsource runoff, and nearshore currents
2. Offshore clarity is affected by mixing events and surface currents, such as during and
following wind-driven upwelling
Approach, Methodology, Geographic Location
Remote Sensing Imagery Repository
Remotely sensed satellite data will be acquired and made available to the Tahoe community
in a web-accessible repository, to be integrated with the COMET cyber-infrastructure system being
developed with NSF funding at UC Davis. Access to the RS data and other Tahoe data streams will
be made possible for a wide range of users and applications through the COMET web portal. As well
as research applications, the data will be available to resource managers, stakeholders, educational
institutions and the general public as part of TERC’s outreach program through its new Incline
Village lab.
The RS data sets include visible-band, shortwave infrared, and thermal infrared
(temperature) images acquired by the Advanced Spaceborne Thermal Emission and Reflectance
Radiometer (ASTER), the Landsat Enhanced Thematic Mapper Plus (ETM+), the Moderate
Resolution Imaging Spectrometer (MODIS) and the Advanced Along Track Scanning Radiometer
(AATSR) over Lake Tahoe. MODIS data are available multiple times per day, AATSR data are
available every 4 days and ASTER and ETM+ data are available every 16-days over Lake Tahoe.
AATSR and MODIS data are available at no charge for scientific research, however, ETM+ and
ASTER data cost between $400 and $600 per scene. Dr. Hook is a member of the NASA ETM+
and ASTER science teams. ETM+ and ASTER data over Lake Tahoe are used by Hook for his work
as a science team member and products derived from these data will be made available to this project
at no charge. The ETM+ sensor was recently affected by an instrument problem which limits the
usefulness of the data at the extremities of the scene. Fortunately the image acquisition is always
centered on Lake Tahoe so this has no effect on our work.
Tools for Acquiring and Storing RS Imagery
The MODIS and AATSR instruments acquire data continuously and therefore data are
available whenever Lake Tahoe is in view by the satellite sensor. Conversely the ASTER and
ETM+ instruments do not acquire data continuously and have to be commanded to acquire
data over a particular target. Due to satellite validation work being performed by the authors
which uses Lake Tahoe as a target, ASTER and ETM+ data have been regularly acquired
over the lake, and future acquisitions are planned. As part of our ongoing validation work all
the datasets acquired thus far have been ordered, retrieved, and stored on a 9 Terabyte data
server at JPL. Both the MODIS and AATSR instrument teams automatically “ftp push” any
new data acquired over Lake Tahoe to the data server. The data server provides a central
repository of remotely sensed data for the Lake, and will be linked to the COMET server.
The imagery covers both the lake and the surrounding watershed. The data are ready to be
3
mined to extract information on historical changes in clarity as well as the current clarity to
better understand the processes behind the change in clarity at Lake Tahoe.
Using Remote Sensing to Measure Water Clarity
Preliminary research by Dr. Hook has demonstrated (Figure 5) that the point of lake bottom
extinction of light can be derived from satellite imagery data (ASTER and ETM+), and when
combined with a high resolution digital elevation model (available from USGS) can be used to
determine the maximum penetration depth (MPD) around the lake. The MPD varies around the lake
depending on changes in the amount and size distribution of fine particles and chlorophyll in the
water and therefore provides a measure of clarity. Since the satellite instruments measure the
reflected energy at different wavelengths, the data also provide information on particle type and size
distributions. The MPD is affected by changes in the bottom reflectance, so care must be taken to
correctly account for any changes in the reflectance of the bottom substrate, such as the change from
areas dominated by sand to areas dominated by silt.
For the purposes of calibration of satellite data with in situ measurements and interpretation
of results, it is desirable to relate satellite measurements of radiance to either the diffuse beam
attenuation coefficient of light, Kd, or Secchi disk transparency (Secchi depth). Kd provides an
objective measure of water clarity that can be measured in the field, e.g., with a Seabird SBE-25
profiler as is presently used in the routine lake monitoring program. Kd can be converted to an
estimate of Secchi depth using an empirical relationship (Swift et al. 2006).
Two methodologies allow conversion of satellite radiance directly or indirectly to Secchi
depth. Both algorithms will be used independently to produce maps of nearshore water clarity, using
the visible and near-infrared (NIR) bands of the two high-resolution satellite sensors, the Advanced
Spaceborne Thermal Emission and Reflection Radiometer (ASTER) and Landsat Enhanced
Thematic Mapper Plus (ETM+).
The first methodology (Maritorena et al., 1994; Maritorena, 1996; Palandro et al., 2004) uses
a simplified model of the Beer-Lambert light attenuation law, requiring satellite measurements at two
different depths, the depth of interest and a reference depth, in order to solve for the beam
attenuation coefficient, Kd. To minimize any effects of changes in bottom reflectance on these
parameters, a computer algorithm will be implemented to search a small region (~250 m2) around
each pixel, selecting a reference depth for each that:
1. Differs in depth by >5%. All depths will be computed using a 10 m resolution USGS
bathymetry map (http://tahoe.usgs.gov/products/dem/tahoebathy_e00.zip).
2. Most closely matches the bottom reflectance. A 10 m resolution USGS backscatter map
(http://tahoe.usgs.gov/products/dem/tahoebackscat_e00.zip) will be used as a proxy
for visible-light reflectance.
The resulting maps of Kd will be studied directly, and also converted to Secchi depth, using the
relationship derived by Swift et al. (2006) for comparison to current and historical in situ Secchi depth
measurements.
The second methodology (Kloiber et al., 2001; Chipman et al., 2004) consists of regressing
atmospherically corrected satellite-measured radiance against field-measured Secchi depth and/or
beam attenuation coefficient, Kd. Separate calibration curves will be calculated for ASTER and
ETM+. This method does not depend on reflectance of light from the lake bottom. Therefore, this
algorithm is suitable for measuring clarity across the entire lake surface. This algorithm will be tested
and refined using the instruments and field measurements detailed below.
Sun glint has, in the past, presented the greatest obstacle for automated use of remotely
sensed visible-band data, as it can erroneously create the appearance of variations in lake clarity, and
simultaneously obscure the true variations. However, a recently developed methodology allows this
4
effect to be removed (Hochberg et al., 2003; Hedley et al., 2005). In addition, this algorithm performs
a first-order atmospheric correction of the visible-band satellite radiance data.
Characterizing Spatial Variability of Nearshore Water Clarity
The lake clarity maps derived above will be used to assess the spatial and temporal variability
in near-shore lake clarity. Several regions will be delineated around the shore of Lake Tahoe, near
stream inflows (LTIMP streams, Figure 2) and regions of upwelling and mixing (Figure 3). Time
series of Secchi Depth will be plotted for each region and compared with field measurements. Storm
and runoff events, determined from stream gage data combined with precipitation and wind data
measured at the meteorological stations (Figure 1), will be correlated with spatial and temporal
variability in near-shore clarity
Characterizing Spatial Variability of Mid-Lake Water Clarity
A real-time network of 12 meteorological stations (Figure 1), onshore and offshore, will
provide validation data for storm and mixing events. Four buoys operated by NASA/JPL collect a
combination of water temperature, wind speed and direction, net radiation, relative humidity, and air
temperature. Satellite temperature maps will be produced, applying the methodology of Steissberg et
al. (2005b) to the thermal infrared bands of ASTER and Landsat ETM+, and will be paired with the
in situ measurements to detect wind-driven upwelling. When available, ASTER and ETM+ images
acquired within 40 min. of one another will be used to compute surface current vector fields of Lake
Tahoe (Steissberg et al, 2005a). Offshore and lake-wide clarity will be correlated with wind-driven
upwelling and surface circulation patterns.
Starting in 2005, measurements at the Index and Mid-lake stations were augmented to
include Seabird SBE-25 profiles of conductivity, temperature, depth, dissolved oxygen, chlorophyll,
and diffuse beam attenuation coefficient, Kd. Combined, these measurements will provide excellent
calibration and validation of the satellite data. A continuous, real-time system of measuring light
scattering at two depths in the water column is presently being installed on the buoy at site TB3
(Figure 1), to improve ground-truthing of water clarity algorithms. These instruments, a Hydrorad
(optical) sensor and a Hydroscat (backscattering) sensor, will provide quantitative estimates of water
properties including absorption, volume scattering, total backscatter function, backscattering
coefficient and beam attenuation coefficient, plane irradiance, scalar irradiance, radiance. These data
can then be used to obtain quantitative estimates of turbidity, suspended sediment concentrations,
photosynthetically active radiation (PAR) chlorophyll concentrations and colored dissolved organic
matter.
Ground-Truthing of Nearshore Clarity Algorithms
Numerous in situ measurements will be collected across the lake, in collaboration with
TERC, DRI, and UNR researchers measuring changes in near-shore clarity as part of ongoing and
proposed research to provide further independent ground-truthing of nearshore clarity algorithms.
Near-Real-Time RS-Derived Water Quality
Near-real-time data from the buoy-mounted optical sensors described above will be
incorporated with the other data in the NASA-UCD near-real-time data network. This network
already provides information on lake temperature and meteorological conditions every hour from the
NASA and UCD buoys over the World Wide Web and is used by the boating community and US
Coast Guard station. A reporting system will be developed where RS-derived measures of water
quality are made available on a near-real-time basis.
Products
5
1. A web-accessible repository of existing Tahoe basin RS imagery, integrated with the
COMET cyber-infrastructure system
2. Tools for semi-automatically acquiring and storing future RS imagery
3. Algorithms for characterizing the spatial variability of nearshore and offshore water clarity
4. A description of the linkage between spatial and temporal clarity variability and changes in
stream, drain, and intervening zone inputs to the lake, following storms
5. A description of the relation between spatial and temporal variability in offshore clarity and
lake mixing, following wind-driven upwelling, and surface current patterns
6. A reporting system where RS-derived measures of water quality are made available on a
near-real-time basis
7. A methodology to study future changes in near-shore and offshore water clarity for any
region of concern around Lake Tahoe, which can be used in water quality management
decision-making and design
8. A methodology that can either be directly applied or easily adapted to current and previous
measurements acquired by other sensors, including Landsat-5 Thematic Mapper (TM), to
create a long-term record of clarity to help understand the historical patterns of clarity
change, of importance to present and future basin management
9. Publication of findings in peer-reviewed journals
Schedule of Events/Reporting and Deliverables
DATE
TASK
Jul-07 Commence project
Complete web-accessible repository of existing Tahoe basin RS
Sep-07
imagery
Complete tools for semi-automatically acquiring and storing future RS
Oct-07
imagery
Complete validated algorithms for characterizing the spatial variability
Jan-08
of nearshore and offshore water clarity
Complete description of the linkage between spatial and temporal
Mar-08 clarity variability and changes in stream, drain, and intervening zone
inputs to the lake, following storms
Complete description of the relation between spatial and temporal
May-08 variability in offshore clarity and lake mixing, following wind-driven
upwelling, and surface current patterns
Complete reporting system where RS-derived measures of water
Aug-08
quality are made available on a near-real-time basis
Sep-08 Final Report and submission of manuscript(s) to peer-reviewed journal
6
References
Chipman, J. W., Lillesand, T. M., Schmaltz, J. E., Leale, J. E., and Nordheim, M. J. (2004). Mapping
lake water clarity with Landsat images in Wisconsin, USA. Canadian Journal of Remote Sensing,
30(1):1–7.
Hackley, S.H., B.C. Allen, D.A. Hunter and J.E. Reuter. 2005. Lake Tahoe Water Quality
Investigations, Annual Report: July 1, 2004- June 30, 2005. Tahoe Environmental Research
Center, University of California, Davis. 69 p.
Hatch, LK., Reuter, J. E., and Goldman, C. R. (2001). Stream phosphorus transport in the Lake
Tahoe basin, 1989–1996. Environmental Monitoring and Assessment, 69(1):63–83.
Hedley, J. D., Harborne, A. R., and Mumby, P,J. (2005). Simple and robust removal of sun glint for
mapping shallow-water benthos. International Journal of Remote Sensing, 26(10):2107–2112.
Heyvaert, A., and A. Parra. 2005. Performance Assessment of the Coon Street Detention Basin,
Kings Beach, CA. Final report prepared for the Placer County Department of Public Works and
the California Tahoe Conservancy. March 2005.
Hochberg, E.J., Andréfouët, S., and Tyler, M.R. (2003). Sea surface correction of high spatial
resolution IKONOS images to improve bottom mapping in near-shore environments. IEEE
Transactions On Geoscience And Remote Sensing, 41(7):1724–1729.
Hook , S. J., Prata, F. J., Alley, R. E., Abtahi, A., Richards. R. C., Schladow, S. G. and Palmarsson, S.
O. (2003). Retrieval of Lake Bulk-and Skin-Temperatures using Along Track Scanning
Radiometer (ATSR) Data: A Case Study using Lake Tahoe, CA. J. Atmospheric and Oceanic
Technology, vol. 20, No. 4, pp 534-548
Hook, S. J., Chander, G, Barsi, J. A., Alley, R. A., Abtahi, A., Palluconi, F. D., Markham, B. L.,
Richards, R. C., Schladow, S. G. and Helder, D. L. (2004). In-Flight Validation and Recovery of
Water Surface Temperature with Landsat 5 Thermal Infrared Data Using an Automated High
Altitude Lake Validation Site at Lake Tahoe CA/NV, USA. IEEE Transactions Geoscience and
Remote Sensing, 42(12), 2767-2776.
Hook, S. J., Clodius, W. B., Balick, L., Alley, R. E., Abtahi, A., Richards, R. C. and Schladow, S. G.
(2005). In-flight validation of mid- and thermal infrared data from the multispectral thermal
imager (MTI) using an automated high-altitude validation site at Lake Tahoe CA/NV, USA.
IEEE Transactions Geoscience and Remote Sensing, 43(9), 1991-1999.
Kloiber, S.N., Brezonik, P.L., Olmanson, L.G., and Bauer, M.E. (2002). A procedure for regional
lake water clarity assessment using Landsat multispectral data. Remote Sensing of Environment,
82(1):38–47.
Maritorena, S., Morel, A., and Gentili, B. (1994). Diffuse reflectance of oceanic shallow waters:
influence of water depth and bottom albedo. Limnology and Oceanography, 39(7):1689–1703.
Maritorena, S. (1996). Remote sensing of the water attenuation in coral reefs: a case study in French
Polynesia. International Journal of Remote Sensing, 17(1):155–166.
Schladow, S.G., Pálmarsson, S.Ó., Steissberg, T.E., Hook, S.J., and Prata, F.J. (2004). An
extraordinary upwelling event in a deep thermally stratified lake. Geophysical Research Letters,
31(15):L15504.
Sibson, R. (1981). A brief description of natural neighbor interpolation. In Barnett, V., editor,
Interpreting Multivariate Data, pages 21–36. John Wiley & Sons.
Steissberg, T.E., Hook, S.J., and Schladow, S.G. (2005a). Measuring surface currents in lakes with
high spatial resolution thermal infrared imagery. Geophysical Research Letters, 32(11):L11402.
Steissberg, T.E., Hook, S.J., and Schladow, S.G. (2005b). Characterizing partial upwellings and
surface circulation at Lake Tahoe, California-Nevada, USA with thermal infrared images. Remote
Sensing of Environment, 99(1-2):2–15.
12
Swift, T.J., J. Perez-Losada, S.G. Schladow, J.E. Reuter, A.D. Jassby and C.R. Goldman. 2006.
Water clarity modeling in Lake Tahoe: Linking suspended matter characteristics to Secchi depth.
Aquatic Sciences. 68:1-15.
13
Table 1: Satellite sensor characteristics
Satellite Sensor
Band
Spectral Region
ASTER
Landsat ETM+
1
2
3
1
3
4
Green-Yellow
Red
Near-Infrared (NIR)
Blue
Red
Near-Infrared (NIR)
14
Wavelengths
(nm)
520 – 600
630 – 690
760 – 860
450 – 520
630 – 690
780 – 900
Spatial
Resolution (m)
15
15
15
30
30
30
Figure 1. Sampling station location map. TB1-TB4 are NASA buoys. TDR1 and TDR2 are UCD
buoys. MLTP (Mid-lake) station is red star by TB1, LTP (Index) station is red star south of
Sunnyside. Solid boxes around lake are onshore meteorological stations.
15
Figure 2. Map of the Lake Tahoe basin, delineating the watersheds of the nine Lake Tahoe
Interagency Monitoring Program (LTIMP) streams. Source: Hatch et al. (2001).
16
Figure 3. Maps of surface temperature anomaly (temperatures minus the spatial median) of Lake
Tahoe, collected 38 min. apart, June 3, 2001 (a) Landsat ETM+ (b) ASTER. Strong upwelling of cold
(blue) water from the west is shown traveling eastward in the form of a jet. Upwelling is also visible
along the southern shore. Source: Steissberg et al. (2005a).
17
Figure 4. Satellite images acquired June 3– 4, 2001, showing relative skin temperatures. Stations: L =
LTP, M = MLTP, 1 –4 =TB1–TB4. Strong upwelling is visible along the western shore and
transported eastward in a surface jet. Note that this jet intersects both the LTP (Index) and MLTP
(Mid-lake) clarity sampling stations, indicating the possibility that measurements at these stations may
significantly differ from other sites, depending on lake mixing and surface current patterns. Source:
Steissberg et al. (2005b).
18
Figure 5. Near-shore clarity map derived from ASTER data. Different colors represent different
maximum penetration depths. Greatest depth obtained from west shore in south (33 m).
19