lake monitoring in wisconsin using satellite remote sensing · lake monitoring in wisconsin using...
TRANSCRIPT
Lake Monitoring in Wisconsin using Satellite Remote Sensing
D. Gurlin and S. GrebWisconsin Department of Natural Resources2015 Wisconsin Lakes Partnership ConventionApril 23‐25, 2105Holiday Inn Convention Center, Stevens Point
LDCM artist’s rendering: NASA/Goddard Space Flight Center Conceptual Image Lab
Remote sensing applications for environmental monitoring
Advantages• Water quality data with a high spatial
and temporal resolution for thousands of lakes at a time
• Evaluation of environmental problems and potential health risks
• Historical data for studies of trends in water quality
• Real time data for integration into early warning systems to protect the public from harmful algal blooms
Disadvantages• Optically complex conditions found in
lakes• Potential interference from the lake
bottom in shallow lakes• Dynamic changes in water quality• Limited number of water quality
parameters• Calibration and validation of models
typically requires the collection of ground truth data
Advantages and disadvantages of remote sensing for lake monitoring
• Systematic processing of Landsat 7 ETM+ and Landsat 8 OLI data for the retrieval of water clarity
Landsat 8 OLI image courtesy of the U.S. Geological Survey
• Increase in Earth observation monitoring capabilities through the optical and biogeochemical characterization of lakes in support of algorithm calibration, refinement, and validation
• Studies of the major drivers of lake water clarity, their interactions, and the potential impacts of land use and climate on water clarity
Remote sensing activities at the Wisconsin DNR
Landsat 8 OLI and TIRS (02/11/2013)
• Scene size 170 x 180 km• Repeat cycle 16 days
OLI• Eight multispectral
bands and one panchromatic band
• Pixel size 30 m for multispectral bands and 15 m for panchromatic band
TIRS• Two thermal bands• Pixel size 100 m
Table from Dekker, A.G. & Hestir, E. L. (2012) Evaluating the Feasibility of Systematic Inland Water Quality Monitoring withSatellite Remote Sensing. CSIRO: Water for a Healthy Country National Research Flagship
Remote sensing activities at the Wisconsin DNR
Remote sensing of water quality
Remote sensing reflectance
Trout Lake
Downwelling irradiance, Ed(λ)
Water leaving radiance, Lw(θ,φ,λ)
Water leaving radiance
Downwelling irradiance
Rrs(θ, φ, λ) =Lw (θ, φ, λ)
Ed (λ)
Remote sensing of water quality
Absorption and scattering coefficients
Sensitivity of the reflectance to variations in the solar zenith angle
Bidirectional properties of the reflectance
Absorption coefficient
Backscattering coefficient
f(λ) bb(λ)
Q(λ) a(λ) + bb(λ)Rrs(θ, φ, λ) =
a(λ) = aφ(λ) + aNAP(λ) + aCDOM(λ) + aw(λ)
Remote sensing of water quality
0.000
0.001
0.002
0.003
0.004
0.005
400 450 500 550 600 650 700 750
R rs, sr‐1
Wavelength, nm
ChlorophyllsCarotenoidsNAPCDOM
Chlorophyll‐aPhycocyanin
0.000
0.001
0.002
0.003
0.004
0.005
400 450 500 550 600 650 700 750
R rs, sr‐1
Wavelength, nm
ChlorophyllsCarotenoidsNAPCDOM
Chlorophyll‐a
Trout Lake 06/17/2014
Rainbow Flowage 06/17/2014
Landsat 8 spectral bands graph from http://landsat.gsfc.nasa.gov
Landsat 8 spectral bands
Systematic processing of satellite data for water clarity
2013 water clarity estimation• 54 satellite images• 3992 ground truth measurements• 32 data processing steps• 9 image mosaics for algorithm
development• 475 ground truth measurements for
algorithm development• 8561 water clarity estimates• 3788 files• 0.94 TB of data
Systematic processing of satellite data for water clarity
Image processing steps• Conversion of data to TOA spectral radiance• Reprojection of images to WTM• Removal of cirrus clouds• Removal of land• Removal of shallow waters and aquatic
vegetation • Mosaicking• Extraction of radiance values for CLMN
stations with data collected within one week from image acquisition date
Systematic processing of satellite data for water clarity
Algorithm calibration
0.00.51.01.52.02.53.03.5
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
LN_S
D
B2_RAD/B4_RAD
Predicted LN_SDLN_SD
0.00.51.01.52.02.53.03.5
40 42 44 46 48 50 52 54 56 58 60
LN_S
D
B2_RAD
Predicted LN_SDLN_SD
ln(SD) = a + b + c OLIB2OLIB2OLIB4
2013 preliminary water clarity composite
Systematic processing of satellite data for water clarity
29185453 5150
1010
74406318 5598 6147
22387
29957
1169312953
7898
0
5000
10000
15000
20000
25000
30000
1999 2000 2001 2003 2004 2005 2007 2008 2009 2010 2011 2012 2013
Num
ber o
f satellite retrieve
d water
clarity m
easu
remen
ts
Year
Lakes and Aquatic Invasive Species (AIS) Mapping Tool
http://dnr.wi.gov/lakes/viewer/
Systematic processing of satellite data for water color
Algorithm calibration
‐2.0
0.0
2.0
4.0
6.0
1.50 1.60 1.70 1.80
LN_C
_VALU
E
B3_RAD/B4_RAD
Predicted LN_C_VALUE
LN_C_VALUE
‐2.0
0.0
2.0
4.0
6.0
55 56 57 58 59 60 61 62 63 64 65
LN_C
_VALU
E
B2_RAD
B2_RAD Line Fit Plot
Predicted LN_C_VALUE
LN_C_VALUE
ln(C) = a + b + c OLIB2OLIB3OLIB4
2013 preliminary water color product
AverageWater Color
Big Saint Germain Lake• 5.5 PCURainbow Flowage• 33.0 PCUPickerel Lake• 13.3 PCU
Major drivers of lake water clarity
National Land Cover DatabaseUSDA National Soil Survey
Gridded climate
Water chemistry
Digital elevation models
Predictor categories• Climate• Land use/land cover• Surficial geology• Water chemistry• Lake morphology & position• Runoff potential
Focus on• Explained variance• Response distributions
Data courtesy of Kevin Rose, University of Wisconsin‐Madison
Major drivers of lake water clarity
What are the implications of long term trends in temperature and precipitation?
Data courtesy of Kevin Rose, University of Wisconsin‐Madison
Major drivers of lake water clarity
Predictor categoryPredictors
(#)
2005 variance explained
2010 variance explained
Climate 13 30.0 23.1Land use/land cover 26 28.5 21.3Lake morphometry 4 27.5 23.7Run‐off potential 5 18.8 11.7Catchment morphometry 11 17.9 12.0Water chemistry 3 12.8 2.1Geology 18 4.3 3.6
Total 80 64.4 52.4
Water clarity is regulated by many different drivers.
Dry year, 2005Wet year, 2010
Data courtesy of Kevin Rose, University of Wisconsin‐Madison
Major drivers of lake water clarity
Wet year, 2010
Dry year, 2005
Data courtesy of Kevin Rose, University of Wisconsin‐Madison
Major drivers of lake water clarity
Data courtesy of Kevin Rose, University of Wisconsin‐Madison
Regulators of lake water clarity• Deep lakes high in the landscape tend
to be the clearest• Agricultural land use is the best land
use predictor of water clarity• High precipitation is associated with
lower water clarity
Increase in Earth observation monitoring capabilities
Field and laboratory measurements• Water temperature, dissolved oxygen,
conductivity, and Secchi depth• Reflectance• Water color and turbidity• TSS, ISS, and OSS• Absorption and backscattering
coefficients
Optical and biogeochemical characterization of lakes • Field data collection in summer and fall
2014 for algorithm development • 24 lakes in Wisconsin
Return from field data collection at Lake Geneva
Thank you!
LDCM artist’s rendering: NASA/Goddard Space Flight Center Conceptual Image Lab
D. Gurlin and S. GrebWisconsin Department of Natural Resources2015 Wisconsin Lakes Partnership ConventionApril 23‐25, 2105Holiday Inn Convention Center, Stevens Point