Quantitative remote sensing for detection and monitoring of cyanobacterial blooms

Size: px

Start display at page:

Download "Quantitative remote sensing for detection and monitoring of cyanobacterial blooms"

Clifton Cobb
7 years ago
Views:

1 Quantitative remote sensing for detection and monitoring of cyanobacterial blooms Tim Malthus, Arnold Dekker, Nagur Cherukuru, Vittorio Brando Research Group Leader, CSIRO Land and Water, Environmental Earth Observation group 12 th August 2009

2 Structure Introduction why remote sensing? Underwater optics a physical understanding of how we can remotely sense cyanobacteria Algorithms Case studies and applications Future outlook for monitoring cyanobacteria

3 Introduction Significant benefits to adopt remote sensing of inland and coastal waters: Quantitative, provides spatial and temporal coverage At effectively lower cost than traditional field-based methods Additional means to map cyanobacterial distributions A complement to traditional in situ methods From 2010 onwards, our ability to map cyanobacterial concentrations from space will take a significant step forward

ALOS ALOS ENVISAT Air Photo TERRA, AQUA Daedallus Semi-Operational And Experimental (Weather, Oceans Land-mapping, Environment, Atmosphere) Operational, Opportunistic and Experimental (Coastal,

4 Current Multiplicity of Satellite & Airborne Measurements FY-3A MTSAT GOES Operational (Weather and Comms.) Operational (GPS Navigation) Ikonos QuickBird Landsat Radarsat SeaWIFS NOAA SPOT Operational (Weather, Oceans Land-mapping) ERS Proba-2 BIRD DMC CBERS EO-1 CASi IRS Calipso CloudSat Lidar Jason ALOS ALOS ENVISAT Air Photo TERRA, AQUA Daedallus Semi-Operational And Experimental (Weather, Oceans Land-mapping, Environment, Atmosphere) Operational, Opportunistic and Experimental (Coastal, Resources, Topo Land-mapping, Precision Ag Etc.) Hymap Australia is one of the largest users of foreign satellite information; roughly close to TB of this satellite data per month, across various agencies CSIRO. and Quantitative the commercial remote sensing sector for cyanobacterial blooms, National Cyanobacterial workshop, Sydney August 2009

5 Optical complexity of inland and coastal waters Optically shallow water Optically deep water

6 Relationships There is a need to understand the relationship between water-leaving radiances and other biological or optical parameters Optically Active Components (OACs) - Phytoplankton - Suspended matter - Coloured dissolved organic matter - Water itself Conclusion: optical complexity of inland and coastal waters requires a physics based approach (as opposed to an empirical one) Empirical approaches (Semi) analytical approaches Inherent optical properties - Beam attenuation (c) - Absorption (a) - Scattering (b) - Volume Scattering [ (0)] c = a + b Apparent optical properties - Reflectance [R(0-)] - Attenuation (K d ) - Transparency (Z SD ) - Colour

7 Colour of cyanobacterial dominated waters

8 R(0-) Subsurface reflectance from lakes Chl a: 7 to 130 mg m Wavelength (nm) UK lakes

9 Algorithms Dekker (1993) Casi sensor PC R R R, R Jupp et al. (1994) Casi sensor Vincent et al. (2004) Landsat using multivariate regression Simis et al based on field reflectance data

10 Implications Requires high spectral resolution or sensors designed with specifically targeted band sets Requires high spatial resolution to map, often small, water bodies Requires high temporal resolution to match bloom dynamics The challenge is robustness of the algorithms under varying: Inorganic turbidities CDOM concentrations In the presence of other phytoplankton functional types Provides only a surface view, but overcomes the limitations in ship-board sampling of near surface concentrations

11 Examples Airborne imaging spectrometry derived cyanophycocyanin image over Dutch lakes indicative of the concentration of potentially toxic cyanobacteria, 1997 Dekker et al. (1997)

12 Biomass micro?m2 x 103 Hawkesbury river ( ) Phytoplankton composition in the Hawkesbury River Other Diatom 400 Green 300 Blue- Green Jupp et al. (1994) 0 Windsor Bridge Wilberforce Cattai Sackville Ferry Lower Portland Ferry Leets Vale

13 Pigment concentrations Hawkesbury ( ) Chlorophyll Measurement Phycocyanin Jupp et al. (1994)

14 Hyperspectral sensors Sensor Spectral resolution Spatial resolution Repeat cycle Launch (no. bands) MERIS m 3 days 2002 CHRIS m na 2001 HYPERION m 16 days 2000 HJ-1A/HSI HYSI na HICO (ISS) na Sept 2009 MIMSat na 30 na 2010? PRISMA Daily 2011 Hyper-X ? 2013 EnMAP HYSPIRI days 2014?

15 Future outlook Quantitative estimation of TSM, CDOM and chlorophyll is now routine Currently, we are limited in space-borne resolution (spectral, spatial, temporal) But: on threshold of major developments in space sensor technology High spectral, high spatial, higher temporal resolution Routine quantitative, cyanobacterial monitoring from space could be a reality Requires physics-based information extraction techniques (analytical approach, multi-temporal applicability) Spatially comprehensive data will be a considerable benefit will allow the integration of the rs-derived products to growth and environmental models (next talk) Combination of RS and in situ fluorescence measurement of pigments could be come quite powerful Future research: Optical properties of cyanobacteria / tank experiments Robust algorithms, adapted to Australian conditions (inland and coastal) Tested over a range of concentrations and conditions (even where cyanobacteria don t dominate) Detection of early stage bloom development?

16 Environmental Earth Observation Group CSIRO Land and Water Dr Tim J Malthus Research Group Leader Phone: tim.malthus@csiro.au Web: Thank you Contact Us Phone: or enquiries@csiro.au Web:

Lake Monitoring in Wisconsin using Satellite Remote Sensing

Lake Monitoring in Wisconsin using Satellite Remote Sensing D. Gurlin and S. Greb Wisconsin Department of Natural Resources 2015 Wisconsin Lakes Partnership Convention April 23 25, 2105 Holiday Inn Convention