POTENTIALS FOR DETECTING CANOPY WATER STRESS USING GEOSTATIONARY MSG-SEVIRI SWIR DATA

Transcription

1 POTENTIALS FOR DETECTING CANOPY WATER STRESS USING GEOSTATIONARY MSG-SEVIRI SWIR DATA Rasmus Fensholt, *Department of Geography and Geology, University of Copenhagen, Denmark Co-workers; Silvia Huber*, Simon R. Proud*, Mads O. Rasmussen*, Inge Sandholt, Simon Stisen**, Cheikh Mbow, UCAD Senegal 1

2 Water; primary potential climatic constraint to plant growth (4% of Earth's Terrestrial Surface) Outline: EO-based Canopy Water Stress detection In situ measured Canopy Water Stress Potential climatic constraints to plant growth. Nemani et al. (23) Results & validation - from point observations - validation in the spatial domain 2

3 EO-based Canopy Water Stress detection Absorption by leaf water occurs in SWIR - Shortwave infrared reflectance is negatively related to leaf water content - Increased reflectance in SWIR is the most consistent leaf reflectance response to plant stress in general, including water stress. SWIR reflectance influenced by; Leaf water content Leaf internal structure Leaf dry matter content Prospect+Sail models; Zarco-Tejada and Ustin 21 Cw (water thickness)=.1 NIR reflectance influenced by; Leaf internal structure Leaf dry matter content Cw =.3 CW =.1 MSG Band 1 Band 2 Band 3 CW =.3 3

4 EO-based Canopy Water Stress detection - Water Stress index development Physically based studies; Tucker, 198; Fourty and Baret, 1997 Laboratory measurements; Hunt, Rock, & Nobel, 1987 Carter, 1994 Physically Emperically applied to sat sensors Hunt and Rock, 1989 Landsat TM Gao, AVIRIS Moisture Stress Index Normalized Difference Water Index Serrano, Ustin, Roberts, et al., 2 - AVIRIS Zarco-Tejada and Ustin, 23 - MODIS Simple Ratio Water Index Ceccato et al., 21; 22 - SPOT VGT. Fensholt and Sandholt, 23 - MODIS Rubio et al., 26 - MODIS Trombetti et al., 28 MODIS Fensholt et al. 21 SEVIRI MSG Shortwave Infrared Water Stress Index Normalized Difference Water Index 7 Shortwave Infrared Ratio Shortwave Infrared Water Stress Index 4

5 Variable performance validation From the Dahra test site in semi-arid Senegal Senegal Mauritania Mali Burkina Faso Nigeria Niger Chad Sudan Ethiopia Northern Sahel Southern Sahel 5 1, 2, Kilometers Annual Rainfall Dahra Senegal

6 Dahra test site setup Since 24 - Air temperature, - Relative humidity - Wind speed - Net radiation Global radiation - Ground heat flux Full surface energy balance Flux profile estimates of latent and sensible heat Since 22 - Precipitation & surface temperature - Soil moisture & soil temperature profiles - Sensor specific reflectances, matching various sensors for estimation of spectral vegetation indices & fapar. Ancillary sampling: biomass, vegetation height, root depth etc. Measurements every 15 minutes Windspeed Net radiation IR radiometer Air temperature & humidity Global radiation Spectral Reflectance 6

7 Dahra test site setup Since 28 - LST (collaboration with Institute of Technology (KIT) Institute for Meteorology and Climate Resarch (IMK) Atmospheric Trace Gases and Remote Sensing (ASF) - Poster: Rasmussen, Mads, O. et al. Intercomparison between SEVIRI LST-products And comparison with in situ LST measurements 7

8 East West Since 21 - ASD spectroradiometers (35-18 nm) Eddy covariance fluxes (water and carbon) Collaboration with University of Lund Eddy Flux Tower 8

9 Prospect+Sailh models; Zarco_Tejada and Ustin 21 NDVI Can SWIR based canopy water status be detected 2 In the field?,8,7,6 ASD data 21 preliminary results...,5,4 Poster: Huber, S. et al.,3 Detecting,2 drought related stress with field spectroradiometric,1 measurements of natural grass savanna NDVI ASD NADIR MSG config Soil moisture 5cm Soil Moisture Cw =.3 Cw (water thickness)=.1 -, Reflectance (%) DOY 238 DOY 24 DOY 242 DOY

10 Mauritania Mali Niger Chad Sudan Senegal Burkina Faso Nigeria Ethiopia Northern Sahel Southern Sahel 5 1, 2, Kilometers,8 NDVI ASD NADIR MSG config SIWSI ASD NADIR MSG config. 2,7 Soil moisture 5cm Soil moisture 1cm,6 16,5 12,4,3,2,1 -, NDVI and SIWSI 8 Soil Moisture 4 1

11 4-day BRDF corrected reflectances Febr. 28 Annual BRDF corrected reflectances avg. 28 MSG SEVIRI data University of Copenhagen SMAC corrected (MOD8 input), reflectances BRDF (NBAR) Proud, S. R. 21 Poster: Evaluating the effectiveness of producing BRDF models from SEVIRI surface reflectance data. 11

12 ,9 Dahra,8 test site in Senegal,7 Terra MODIS NDVI Aqua MODIS NDVI IN situ MODIS NDVI GoogleEarth MODIS NDVI 5m resolution,6,5,4,3,2 MODIS NDVI (4.5 km aggregated),1,8,7,6,5,4,3,2,1 MODIS 5m y = 1,1x -,2, R2 =,99 MODIS 25m y =,99x -,1, R2 =,99 Dahra Village Test site , ,1,2,3,4,5,6,7,8 MODIS NDVI 25 & 5 m MODIS 5m MODIS 25m MODIS NDVI 25m,8,7,6 37,4,3,2,1 Terra MODIS y =,92x +,9, R2 =,7 Aqua MODIS y =,95x +,1, R2 =, km ,1,2,3,4,5,6,7,8 In situ NDVI Terra MODIS Aqua MODIS

13 Comparing MSG SEVIRI vegetation indices with in situ measuremets,7,6,5,4,3,2,1 Mauritania Mali Niger Senegal Burkina Faso Nigeria Northern Sahel Southern Sahel 5 1, 2, Kilometers Chad Sudan Ethiopia 13 NDVI MSG DGG NDVI MSG In situ MSG LSA SAF NDVI

14 Mauritania Mali Niger Chad Sudan Senegal Burkina Faso Nigeria Northern Sahel Southern Sahel 5 1, 2, Kilometers Ethiopia Julian day 28,8,7,6,5,4,3,2,1 Above ground Biomass (g/m 2 ) In situ NDVI and Root depth (m) Evaluation of EO-based SIWSI from in situ measuremets - 28 In situ NDVI (MODIS config. Daily avg) Root depth Aboveground Biomass 66.5% % 66.7% 67.7% 6.4% 51.2%,7,2,6,15,5,1,4,5,3,2 -,5 -,1 -,15 -,1 -,2 -,25 -,3 14 soil moisture (Vol %) NDVI and SIWSI indices NDVI and SIWSI indices

15 Evaluation of EO-based SIWSI from in situ measuremets Mauritania Mali Niger Chad Sudan Senegal Burkina Faso Nigeria Northern Sahel Southern Sahel 5 1, 2, Kilometers,7,2,6,15,5,1,4,5,3,2 -,5 -,1 -,15 -,1 -,2 -,25 -,3 Ethiopia 15 soil moisture (Vol %) NDVI and SIWSI indices NDVI and SIWSI indices

16 Evaluation of EO-based SIWSI from in situ measuremets Mauritania Mali Niger Chad Sudan Senegal Burkina Faso Nigeria Northern Sahel Southern Sahel 5 1, 2, Kilometers,2,7,15,6,1,5,5,4,3 -,5,2 -,1,1 -,15 -,1 -,2 -,2 -,25 -,3 -,3 Ethiopia 16 soil moisture (Vol%) %) NDVI and SIWSI indices NDVI and SIWSI indices

17 Evaluation of EO-based SIWSI from in situ measuremets Mauritania Mali Niger Chad Sudan Senegal Burkina Faso Nigeria Northern Sahel Southern Sahel 5 1, 2, Kilometers,7,6,5,4,3,2,1 -,1 -,2 -,3 Ethiopia 17 soil moisture (Vol %) NDVI and SIWSI indices

18 Mauritania Mali Niger Chad Sudan Senegal Burkina Faso Nigeria Northern Sahel Southern Sahel 5 1, 2, Kilometers Ethiopia Julian day 28,8,7,6,5,4,3,2, Above ground Biomass (g/m 2 ) In situ NDVI and Root depth (m) Evaluation of EO-based SIWSI from in situ measuremets - 28 In situ NDVI (MODIS config. Daily avg) Root depth Aboveground Biomass 66.5% % 66.7% 67.7% 6.4% 51.2%,2,15,1,5 -,5 -,1 -,15 -,2 -,25 -,3 18 soil moisture (Vol %) NDVI and SIWSI indices

19 14,25 Soil moisture (vol%) Soil moisture 5 cm Terra MODIS SIWSI MSG SEVIRI daily BRDF SIWSI MSG SEVIRI SIWSI Aqua MODIS SIWSI,2,15,1,5 -,5 -,1 -,15 -,2 MSG SEVIRI SIWSI (9 am. 4 pm. local solar time) Net radiation Sensible heat flux Latent heat flux Soil flux 7 -, Watt/m

20 - Spatio-temporal Evaluation of SIWSI (2 pixels, 32 km 2 ) using NOAA RFE rainfall as surface water status indicator 2

21 - Spatio-temporal evaluation of MSG SIWSI using a hydrological model (Mike-She distributed model) Is the model able to simulate water status at the Dahra test site? Are model inputs (RFE rainfall) reliable? Soil moisture (Vol %) EvapoTranspiration (mm/day) MSH model soil moist 25 cm in situ soil moist 3 cm MSH ETa

22 - Spatio-temporal evaluation of MSG SIWSI using a hydrological model (Mike-She distributed model) Preliminary data analysis 22

23 23

24 SIWSI NDVI ETa 24

25 Conclusions and perspectives - SWIR sensitivity to Canopy water content (semi-arid grass land) - MSG sensitivity on a daily scale - Biomass dependancy - SWIR based indices complementary to VIS/NIR approaches - SWIR based indices more robust to atm correction than VIS/NIR 9 Rainfall In situ MODIS NDVI Soil moisture 1 cm MODIS NDVI Rainfall (mm) adn Soil moisture (vol%)