remote sensing of water quality parameters and cyanobacterial algal ...
REMOTE SENSING OF WATER QUALITY PARAMETERS AND CYANOBACTERIAL ALGAL BLOOMS USING MULTI-SPECTRAL MERIS AND LANDSAT, AND IN SITU HYPERSPECTRAL RADIOMETRIC DATA, IN ZEEKOEVLEI LAKE, CAPE TOWN, SOUTH AFRICA Matthews Mark, Winter Kevin, Bernard Stewart University of Cape Town, South Africa ([email protected])
Introduction: Problem statement • Eutrophication is a major problem in freshwater resources in South Africa and the rest of the world • Associated cyanobacterial algal blooms exhibiting cyanotoxins are a threat to animal and human health • Microcystin producing Microcystis widespread in southern Africa and responsible for numerous animal kills (Oberholster et al. 2005; Scott, 1991)
Introduction: Need • It is crucial, from a management point of view, that sufficient information is available in order to monitor the occurrence and distribution of eutrophication and HABs (Glasgow et al., 2004) • Remote sensing affords the opportunity to monitor for these events with large spatial coverage and high frequencies not afforded by traditional monitoring methods so as to minimize the risks associated with these events
Introduction: Aims & objectives
1. 2. 3. 4.
Enhance understanding and capacity for monitoring duel threats from eutrophication and cyanobacterial algal blooms in Zeekoevlei Lake using remote sensing Determine utility of various multi-spectral sensors Assess water quality and the water-leaving reflectance Derive empirical algorithms and test MERIS Case 2 and Eutrophic Lakes products Produce maps for management
Introduction: Tools • Hyperspectral radiometer used in situ • Medium Resolution Imaging Spectrometer (MERIS) well suited to water-related applications in inland waters (Giardino et al., 2005) • Landsat 7 ETM+ also been used to map water quality parameters phycocyanin) in inland lakes (Vincent et al., 2004)
Site Description • • • • •
Zeekoevlei – literally “Hippo Lake” Small lake (2.56Ha) south of Cape Town Largest vlei on Cape Flats – valuable resource Hypertrophic (mean Chl a = 240mg.m-3) Near-permanent blooms dominated by cyanobacterium Microcystis Aeruginosa (Harding, 1997) • Dog died after drinking water containing toxin nodularin in 1995 (Harding et al., 1995)
GREAT LOTUS RIVER INLET
N LITTLE LOTUS RIVER INLET
W
E S ZEV 2
$
NORTHERN BASIN ZEV 1
$
HOME BAY
260m 290m
#
YACHT CLUB PIER
SOUTHERN BASIN
ZEV 4
$
ZEV 3
$
WEIR OUTLET
500
0
500
1000 Meters
Methodology: Fieldwork Made simultaneous to MERIS overpasses: 1. Upwelling radiance at 0.66m, downwelling irradiance (0+) using Satlantic hyperspectral radiometer buoy 2. Chl a ; Totals Suspended Solids (TSS) (Organic and Inorganic) ; Secchi Disk depth (SD) ; absorption by CDOM (aCDOM) 3. AOT at five wavelengths (Microtops) 4. Wind speed; wave heights; sky conditions
Methodology: Radiometric processing • Correction for instrument-self shading (Leathers et al., 2001) • Derivation of Ku (Albert, Mobley 2003) with simple bio-optical model (Stramski et al., 2001) • Calculation of ρw
Methodology: Atmospheric correction • Image based corrections – Dark Object Subtraction (Moran et al., 1992) – COST method (Chavez, 1996)
• 6S Radiative Transfer Code V1.1 (Vermote et al., 1997) – Input parameters including AOT at 550nm – Urban aerosol model; mid-latitude summer atmospheric model; Lambertian assumption
Methodology: Correlation analysis • Correlation between water quality parameters and reflectance from radiometer, MERIS (TOA, BOA) and simulated Landsat 7 • Linear and non-linear regression analysis to produce empirical predictor algorithms • Variety of algorithms tested following extensive review of the literature
Methodology: Empirical Algorithms • Single bands (e.g. 700nm) • Band ratios (e.g. 700nm/670nm) • Band difference ratios – (560-520)/(560+520) Gitelson et al., 1993 – 700/(560+670) Koponen et al., 2007
• Other algorithms: – Reflectance Line Height (RLH) – 740((1/670)-(1/710)) (Zimba and Gitelson, 2006) – (453-(700/670)-520)/(453+(700/670)+520) (Gitelson et al., 1993)
Methodology: MERIS Neural Network algorithms • Inverse Radiative Transfer Model Neural Network (IRTM-NN) parameterisation algorithms – MERIS Case 2 Waters Algorithm (Schiller & Doerffer, 2005) – MERIS Eutrophic Lakes Processor Algorithm (Doerffer & Schiller, 2008)
• Atmospheric products • Water constituent products
Results: Fieldwork • Four cloud-free MERIS images acquired simultaneous to in situ measurements during April 2008 giving nine data points • One Landsat image acquired May 2nd 2008 • Sixteen radiometric spectra with simultaneous in situ measurements
Results: Water Quality Chl a
TSS
OSS
SD
aCDOM (440nm)
mg.m-3
mg.l-1
%
cm
m-1
148.6 ± 8.1
49.1 ± 2.0
76.9 ± 1.4
27.9 ± 0.8
2.69 ± 0.08
Results: Hyperspectral data
• Absorption by chlorophyll pigments, phycocyanin • Scattering by suspended matter • Absorption by CDOM
Results: Correlation analysis
Results: Correlation analysis
Results: Correlation analysis
• Chl a corresponds very well with reflectance peak near 700nm
Results: MERIS TOA
• Atmospheric scattering contributes greatly in blue • Reflectance peaks/ absorption minima clearly visible
Results: Atmospheric corrections: Image based corrections
Results: Atmospheric correction: 6S RTC
Results: MERIS L2 and Eutrophic Lakes Products
• Produces negative, erroneous reflectance • Failure of atmospheric correction • Underestimates water constituents
Results: Empirical algorithms for MERIS • Single band algorithms performed well with BOA data • Band ratio algorithms performed well with TOA data • Strong co-variance meant that many algorithms appeared to perform equally well with a number of the parameters • Algorithms have a limited ability to distinguish between the parameters
Results: Empirical algorithms for MERIS • Strong non-linear correlation with Chl a and the 708/664 ratio algorithm • More than 75% of variability can be explained from TOA MERIS (excluding ISS) • RMSE of less than 15% (except for ISS)
Results: MERIS Maps • Chl a concentration from MERIS TOA reflectance • 15 – 20 pixels shapes vary as some duplicate pixels • Area +- 1.3 Ha • Max observed error = 30% • Spatial and temporal variability corresponds strongly with wind and wave conditions Strong southerly wind causes mixing of suspended matter into the northern basin from the southern basin
Results: MERIS Maps Suspended solids
Organic solids
• Spatial variability identical/ strongly correlated to Chl a and wind/wave conditions
Results: MERIS Maps Secchi disk depth (water clarity)
Absorption by CDOM (440nm)
• Secchi disk depth shows overall effect of gross suspended particulate matter • Northern basin has ‘improved’ water quality over southern basin
Results: Empirical algorithms for simulated Landsat reflectance
Testing band difference and band ratio algorithms
Results: Empirical algorithms for simulated Landsat reflectance Parameter
Algorithm
r2
Chl a
b1 – b2
0.83
TSS
b2 – b3
0.84
ISS
b2 – b3
0.7
OSS
b2 – b3
0.91
SD
b3/b1
0.73
aCDOM
b3/b2
0.57
Results: Landsat Maps
• 2662 pixels gives an area of 2.39 Ha • Standard error of the mean for Chl a = 1.1% • Distribution is patchy, adjacency effect is evident
Results: Landsat Maps
• Secchi Disk depth shows northern basin water is clearer
Results: Landsat Maps
• Algal blooms occur in irregularly distributed patches across the lake surface
Conclusions • Empirical procedures suitable for monitoring water quality parameters, eutrophication and cyanobacteria dominated algal blooms in a small inland lake • Allows observation of the spatial and temporal variability unrivalled by traditional monitoring systems, with a capacity to provide vital information to lake managers • MERIS: – – – – –
spatial resolution suitable even for small inland lakes free, frequent availability shows variability and range sufficiently, if not better allows application of variety of algorithms allows TOA estimations
Conclusions: Algorithms • • • •
MERIS NN algorithms perform poorly Empirical algorithms are robust, although limited MERIS 708nm band crucial Atmospheric correction not a pre-condition to deriving products using empirical procedures • Combining remote sensing with traditional monitoring methods allows creation of products useful for management
Recommendations • Remote sensing should be integrated into conventional water quality monitoring programmes in southern Africa, in order to determine status and trends of eutrophication • Development of semi—analytical methods should be prioritised because of their cross-temporal/spatial applicability • Multi-site/scale projects should be initiated for development and validation • Operational water quality monitoring programme based on remote sensing would be of immense benefit to southern Africa
References ALBERT, A. and MOBLEY, C., 2003. An analytical model for subsurface irradiance and remote sensing reflectance in deep and shallow case-2 waters. Optics Express, 11(22), pp. 2873-2890. OBERHOLSTER, P.J., BOTHA, A.M. and CLOETE, T.E., 2005. An overview of toxic freshwater cyanobacteria in South Africa with special reference to risk, impact and detection by molecular marker tools. Biokemistri, 17(2), pp. 57-71. SCOTT, W.E., 1991. Occurrence and Significance of Toxic Cyanobacteria in Southern Africa. Water Science and Technology, 23(1-3), pp. 175-180 GLASGOW, H.B., BURKHOLDER, J.M., REED, R.E., LEWITUS, A.J. and KLEINMAN, J.E., 2004. Real-time remote monitoring of water quality: a review of current applications, and advancements in sensor, telemetry, and computing technologies. Journal of experimental marine biology and ecology, 300(1-2), pp. 409-448. GIARDINO, C., CANDIANI, G. and ZILIOLI, E., 2005. Detecting chlorophyll-a in Lake Guarda using TOA MERIS radiances. Photogrammetric Engineering & Remote Sensing, 71(9), pp. 1045-1051. VINCENT, R.K., QIN, X.M., MCKAY, R.M.L., MINER, J., CZAJKOWSKI, K., SAVINO, J. and BRIDGEMAN, T., 2004. Phycocyanin detection from LANDSAT TM data for mapping cyanobacterial blooms in Lake Erie. Remote Sensing of Environment, 89(3), pp. 381-392. HARDING, W.R., ROWE, N., WESSELS, J.C., BEATTIE, K.A. and CODD, G.A., 1995. Suspected toxicosis of a dog attributed to the cyanobacterial (blue-green algal) hepatotoxin nodularin in South Africa. Journal of the South African Vertinary Association, 66(4), pp. 256-259. HARDING, W.R., 1997. Phytoplankton primary production in a shallow, well-mixed, hypertrophic South African lake. Hydrobiologia, 344, pp. 87-102. LEATHERS, R., DOWNES, T.V. and MOBLEY, C., 2001. Self-shading correction for upwelling sea-surface radiance measurements made with buoyed instruments. Optics Express, (10), pp. 561-570. STRAMSKI, D., BRICAUD, A. and MOREL, A., 2001. Modeling the inherent optical properties of the ocean based on the detailed composition of the planktonic community. Applied Optics,40(18), pp. 2929–2945. MORAN, M.S., JACKSON, R.D., SLATER, P.N. and TEILLET, P.M., 1992. Evaluation of simplified procedures for retrieval of land surface reflectance factors from satellite sensor output.Remote Sensing of Environment, 41(2), pp. 169-184. VERMOTE, E.F., TANRE, D., DEUZE, J.L., HERMAN, M. and MORCETTE, J.J., 1997. Second Simulation of the Satellite Signal in the Solar Spectrum, 6S: an overview. Geoscience and Remote Sensing, IEEE Transactions on, 35(3), pp. 675-686. ZIMBA, P.V. and GITELSON, A., 2006. Remote estimation of chlorophyll concentration in hyper-eutrophic aquatic systems: Model tuning and accuracy optimization. Aquaculture, 256(1-4), pp. 272-286. GITELSON, A., GARBUZOV, G., SZILAGYI, F., MITTENZWEY, K., KARNIELI, A. and KAISER, A., 1993. Quantitative remote sensing methods for real-time monitoring of inland waters quality. International Journal of Remote Sensing, 14(7), pp. 1269-1295. KOPONEN, S., ATTILA, J., PULLIAINEN, J., KALLIO, K., PYHÄLAHTI, T., LINDFORS, A., RASMUS, K. and HALLIKAINEN, M., 2007. A case study of airborne and satellite remote sensing of a spring bloom event in the Gulf of Finland. Continental Shelf Research, 27(2), pp. 228-244. SCHILLER, H. and DOERFFER, R., 2005. Improved determination of coastal water constituent concentrations from MERIS data. IEEE Transactions onGeoscience and Remote Sensing, 43(7), pp. 1585-1591 DOERFFER R. and SCHILLER, H., 2008b. MERIS regional coastal and lake case 2 water project - atmospheric correction Algorithm Theoretical Basis Document (ATBD). 1.0. Geesthacht: GKSS Research Center.
Acknowledgements Stewart Bernard (Supervisor) Kevin Winter (Supervisor) Zeekoevlei Nature Reserve – Assief Kahn MCM – laboratory analysis The Lord Jesus Christ
Thank You
Introduction: Problem statement • Eutrophication is a major problem in freshwater resources in South Africa and the rest of the world • Associated cyanobacterial algal blooms exhibiting cyanotoxins are a threat to animal and human health • Microcystin producing Microcystis widespread in southern Africa and responsible for numerous animal kills (Oberholster et al. 2005; Scott, 1991)
Introduction: Need • It is crucial, from a management point of view, that sufficient information is available in order to monitor the occurrence and distribution of eutrophication and HABs (Glasgow et al., 2004) • Remote sensing affords the opportunity to monitor for these events with large spatial coverage and high frequencies not afforded by traditional monitoring methods so as to minimize the risks associated with these events
Introduction: Aims & objectives
1. 2. 3. 4.
Enhance understanding and capacity for monitoring duel threats from eutrophication and cyanobacterial algal blooms in Zeekoevlei Lake using remote sensing Determine utility of various multi-spectral sensors Assess water quality and the water-leaving reflectance Derive empirical algorithms and test MERIS Case 2 and Eutrophic Lakes products Produce maps for management
Introduction: Tools • Hyperspectral radiometer used in situ • Medium Resolution Imaging Spectrometer (MERIS) well suited to water-related applications in inland waters (Giardino et al., 2005) • Landsat 7 ETM+ also been used to map water quality parameters phycocyanin) in inland lakes (Vincent et al., 2004)
Site Description • • • • •
Zeekoevlei – literally “Hippo Lake” Small lake (2.56Ha) south of Cape Town Largest vlei on Cape Flats – valuable resource Hypertrophic (mean Chl a = 240mg.m-3) Near-permanent blooms dominated by cyanobacterium Microcystis Aeruginosa (Harding, 1997) • Dog died after drinking water containing toxin nodularin in 1995 (Harding et al., 1995)
GREAT LOTUS RIVER INLET
N LITTLE LOTUS RIVER INLET
W
E S ZEV 2
$
NORTHERN BASIN ZEV 1
$
HOME BAY
260m 290m
#
YACHT CLUB PIER
SOUTHERN BASIN
ZEV 4
$
ZEV 3
$
WEIR OUTLET
500
0
500
1000 Meters
Methodology: Fieldwork Made simultaneous to MERIS overpasses: 1. Upwelling radiance at 0.66m, downwelling irradiance (0+) using Satlantic hyperspectral radiometer buoy 2. Chl a ; Totals Suspended Solids (TSS) (Organic and Inorganic) ; Secchi Disk depth (SD) ; absorption by CDOM (aCDOM) 3. AOT at five wavelengths (Microtops) 4. Wind speed; wave heights; sky conditions
Methodology: Radiometric processing • Correction for instrument-self shading (Leathers et al., 2001) • Derivation of Ku (Albert, Mobley 2003) with simple bio-optical model (Stramski et al., 2001) • Calculation of ρw
Methodology: Atmospheric correction • Image based corrections – Dark Object Subtraction (Moran et al., 1992) – COST method (Chavez, 1996)
• 6S Radiative Transfer Code V1.1 (Vermote et al., 1997) – Input parameters including AOT at 550nm – Urban aerosol model; mid-latitude summer atmospheric model; Lambertian assumption
Methodology: Correlation analysis • Correlation between water quality parameters and reflectance from radiometer, MERIS (TOA, BOA) and simulated Landsat 7 • Linear and non-linear regression analysis to produce empirical predictor algorithms • Variety of algorithms tested following extensive review of the literature
Methodology: Empirical Algorithms • Single bands (e.g. 700nm) • Band ratios (e.g. 700nm/670nm) • Band difference ratios – (560-520)/(560+520) Gitelson et al., 1993 – 700/(560+670) Koponen et al., 2007
• Other algorithms: – Reflectance Line Height (RLH) – 740((1/670)-(1/710)) (Zimba and Gitelson, 2006) – (453-(700/670)-520)/(453+(700/670)+520) (Gitelson et al., 1993)
Methodology: MERIS Neural Network algorithms • Inverse Radiative Transfer Model Neural Network (IRTM-NN) parameterisation algorithms – MERIS Case 2 Waters Algorithm (Schiller & Doerffer, 2005) – MERIS Eutrophic Lakes Processor Algorithm (Doerffer & Schiller, 2008)
• Atmospheric products • Water constituent products
Results: Fieldwork • Four cloud-free MERIS images acquired simultaneous to in situ measurements during April 2008 giving nine data points • One Landsat image acquired May 2nd 2008 • Sixteen radiometric spectra with simultaneous in situ measurements
Results: Water Quality Chl a
TSS
OSS
SD
aCDOM (440nm)
mg.m-3
mg.l-1
%
cm
m-1
148.6 ± 8.1
49.1 ± 2.0
76.9 ± 1.4
27.9 ± 0.8
2.69 ± 0.08
Results: Hyperspectral data
• Absorption by chlorophyll pigments, phycocyanin • Scattering by suspended matter • Absorption by CDOM
Results: Correlation analysis
Results: Correlation analysis
Results: Correlation analysis
• Chl a corresponds very well with reflectance peak near 700nm
Results: MERIS TOA
• Atmospheric scattering contributes greatly in blue • Reflectance peaks/ absorption minima clearly visible
Results: Atmospheric corrections: Image based corrections
Results: Atmospheric correction: 6S RTC
Results: MERIS L2 and Eutrophic Lakes Products
• Produces negative, erroneous reflectance • Failure of atmospheric correction • Underestimates water constituents
Results: Empirical algorithms for MERIS • Single band algorithms performed well with BOA data • Band ratio algorithms performed well with TOA data • Strong co-variance meant that many algorithms appeared to perform equally well with a number of the parameters • Algorithms have a limited ability to distinguish between the parameters
Results: Empirical algorithms for MERIS • Strong non-linear correlation with Chl a and the 708/664 ratio algorithm • More than 75% of variability can be explained from TOA MERIS (excluding ISS) • RMSE of less than 15% (except for ISS)
Results: MERIS Maps • Chl a concentration from MERIS TOA reflectance • 15 – 20 pixels shapes vary as some duplicate pixels • Area +- 1.3 Ha • Max observed error = 30% • Spatial and temporal variability corresponds strongly with wind and wave conditions Strong southerly wind causes mixing of suspended matter into the northern basin from the southern basin
Results: MERIS Maps Suspended solids
Organic solids
• Spatial variability identical/ strongly correlated to Chl a and wind/wave conditions
Results: MERIS Maps Secchi disk depth (water clarity)
Absorption by CDOM (440nm)
• Secchi disk depth shows overall effect of gross suspended particulate matter • Northern basin has ‘improved’ water quality over southern basin
Results: Empirical algorithms for simulated Landsat reflectance
Testing band difference and band ratio algorithms
Results: Empirical algorithms for simulated Landsat reflectance Parameter
Algorithm
r2
Chl a
b1 – b2
0.83
TSS
b2 – b3
0.84
ISS
b2 – b3
0.7
OSS
b2 – b3
0.91
SD
b3/b1
0.73
aCDOM
b3/b2
0.57
Results: Landsat Maps
• 2662 pixels gives an area of 2.39 Ha • Standard error of the mean for Chl a = 1.1% • Distribution is patchy, adjacency effect is evident
Results: Landsat Maps
• Secchi Disk depth shows northern basin water is clearer
Results: Landsat Maps
• Algal blooms occur in irregularly distributed patches across the lake surface
Conclusions • Empirical procedures suitable for monitoring water quality parameters, eutrophication and cyanobacteria dominated algal blooms in a small inland lake • Allows observation of the spatial and temporal variability unrivalled by traditional monitoring systems, with a capacity to provide vital information to lake managers • MERIS: – – – – –
spatial resolution suitable even for small inland lakes free, frequent availability shows variability and range sufficiently, if not better allows application of variety of algorithms allows TOA estimations
Conclusions: Algorithms • • • •
MERIS NN algorithms perform poorly Empirical algorithms are robust, although limited MERIS 708nm band crucial Atmospheric correction not a pre-condition to deriving products using empirical procedures • Combining remote sensing with traditional monitoring methods allows creation of products useful for management
Recommendations • Remote sensing should be integrated into conventional water quality monitoring programmes in southern Africa, in order to determine status and trends of eutrophication • Development of semi—analytical methods should be prioritised because of their cross-temporal/spatial applicability • Multi-site/scale projects should be initiated for development and validation • Operational water quality monitoring programme based on remote sensing would be of immense benefit to southern Africa
References ALBERT, A. and MOBLEY, C., 2003. An analytical model for subsurface irradiance and remote sensing reflectance in deep and shallow case-2 waters. Optics Express, 11(22), pp. 2873-2890. OBERHOLSTER, P.J., BOTHA, A.M. and CLOETE, T.E., 2005. An overview of toxic freshwater cyanobacteria in South Africa with special reference to risk, impact and detection by molecular marker tools. Biokemistri, 17(2), pp. 57-71. SCOTT, W.E., 1991. Occurrence and Significance of Toxic Cyanobacteria in Southern Africa. Water Science and Technology, 23(1-3), pp. 175-180 GLASGOW, H.B., BURKHOLDER, J.M., REED, R.E., LEWITUS, A.J. and KLEINMAN, J.E., 2004. Real-time remote monitoring of water quality: a review of current applications, and advancements in sensor, telemetry, and computing technologies. Journal of experimental marine biology and ecology, 300(1-2), pp. 409-448. GIARDINO, C., CANDIANI, G. and ZILIOLI, E., 2005. Detecting chlorophyll-a in Lake Guarda using TOA MERIS radiances. Photogrammetric Engineering & Remote Sensing, 71(9), pp. 1045-1051. VINCENT, R.K., QIN, X.M., MCKAY, R.M.L., MINER, J., CZAJKOWSKI, K., SAVINO, J. and BRIDGEMAN, T., 2004. Phycocyanin detection from LANDSAT TM data for mapping cyanobacterial blooms in Lake Erie. Remote Sensing of Environment, 89(3), pp. 381-392. HARDING, W.R., ROWE, N., WESSELS, J.C., BEATTIE, K.A. and CODD, G.A., 1995. Suspected toxicosis of a dog attributed to the cyanobacterial (blue-green algal) hepatotoxin nodularin in South Africa. Journal of the South African Vertinary Association, 66(4), pp. 256-259. HARDING, W.R., 1997. Phytoplankton primary production in a shallow, well-mixed, hypertrophic South African lake. Hydrobiologia, 344, pp. 87-102. LEATHERS, R., DOWNES, T.V. and MOBLEY, C., 2001. Self-shading correction for upwelling sea-surface radiance measurements made with buoyed instruments. Optics Express, (10), pp. 561-570. STRAMSKI, D., BRICAUD, A. and MOREL, A., 2001. Modeling the inherent optical properties of the ocean based on the detailed composition of the planktonic community. Applied Optics,40(18), pp. 2929–2945. MORAN, M.S., JACKSON, R.D., SLATER, P.N. and TEILLET, P.M., 1992. Evaluation of simplified procedures for retrieval of land surface reflectance factors from satellite sensor output.Remote Sensing of Environment, 41(2), pp. 169-184. VERMOTE, E.F., TANRE, D., DEUZE, J.L., HERMAN, M. and MORCETTE, J.J., 1997. Second Simulation of the Satellite Signal in the Solar Spectrum, 6S: an overview. Geoscience and Remote Sensing, IEEE Transactions on, 35(3), pp. 675-686. ZIMBA, P.V. and GITELSON, A., 2006. Remote estimation of chlorophyll concentration in hyper-eutrophic aquatic systems: Model tuning and accuracy optimization. Aquaculture, 256(1-4), pp. 272-286. GITELSON, A., GARBUZOV, G., SZILAGYI, F., MITTENZWEY, K., KARNIELI, A. and KAISER, A., 1993. Quantitative remote sensing methods for real-time monitoring of inland waters quality. International Journal of Remote Sensing, 14(7), pp. 1269-1295. KOPONEN, S., ATTILA, J., PULLIAINEN, J., KALLIO, K., PYHÄLAHTI, T., LINDFORS, A., RASMUS, K. and HALLIKAINEN, M., 2007. A case study of airborne and satellite remote sensing of a spring bloom event in the Gulf of Finland. Continental Shelf Research, 27(2), pp. 228-244. SCHILLER, H. and DOERFFER, R., 2005. Improved determination of coastal water constituent concentrations from MERIS data. IEEE Transactions onGeoscience and Remote Sensing, 43(7), pp. 1585-1591 DOERFFER R. and SCHILLER, H., 2008b. MERIS regional coastal and lake case 2 water project - atmospheric correction Algorithm Theoretical Basis Document (ATBD). 1.0. Geesthacht: GKSS Research Center.
Acknowledgements Stewart Bernard (Supervisor) Kevin Winter (Supervisor) Zeekoevlei Nature Reserve – Assief Kahn MCM – laboratory analysis The Lord Jesus Christ
Thank You
