Vegetation Temperature Condition Index (VTCI) and Its Application for - PowerPoint PPT Presentation

Vegetation Temperature Condition Index (VTCI) and Its Application for Low Streamflow Regional Regression Model Satoshi Hirabayashi

Outline ESPM271 Project Outline � Introduction � Objectives � Methods � Data & Processing � Results � Conclusions 11/15/2005 SUNY-ESF SUNY-ESF

Outline ESPM271 Project Outline � Introduction � Objectives � Methods � Data & Processing � Results � Conclusions 11/15/2005 SUNY-ESF SUNY-ESF

Introduction ESPM271 Project Research Theme � Low streamflow prediction in ungauged watersheds - Regional regression model β γ = α ⋅ ⋅ ⋅ Q X X 7 , 10 1 2 Q 7,10 : 7-day, 10-year low streamflow statistics X i : Watershed characteristics α , β , γ : model parameter to be estimated � Groundwater discharge (Base flow) Flood - Major source of the streamflow in low flow periods Discharge Surface Low flow flow � Remotely sensed data Base flow - To derive a good indicator of the soil dryness Time 11/15/2005 SUNY-ESF SUNY-ESF

Introduction ESPM271 Project Soil Dryness Indicator � Vegetation Temperature Condition Index (VTCI) � Temperature-Vegetation Dryness Index (TVDI) - Calculated from Normalized Difference Vegetation Index (NDVI) and Land Surface Temperature (LST) - NOAA-AVHRR, MODIS images - Good correlation with rainfall events and soil moisture - Applicable to a various geographical scales, from regional (~10,000 km 2 ) to semi-continental (whole China divided into three parts) 11/15/2005 SUNY-ESF SUNY-ESF

Outline ESPM271 Project Outline � Introduction � Objectives � Methods � Data & Processing � Results � Conclusions 11/15/2005 SUNY-ESF SUNY-ESF

Objectives ESPM271 Project Objectives 1. Explore and get familiar with MODIS data & VTCI 11/15/2005 SUNY-ESF SUNY-ESF

Objectives ESPM271 Project Objectives 1. Explore and get familiar with MODIS data & VTCI 2. Develop an integrated VTCI calculation procedure 11/15/2005 SUNY-ESF SUNY-ESF

Objectives ESPM271 Project Objectives 1. Explore and get familiar with MODIS data & VTCI 2. Develop an integrated VTCI calculation procedure 3. Apply VTCI in low streamflow modeling 11/15/2005 SUNY-ESF SUNY-ESF

Outline ESPM271 Project Outline � Introduction � Objectives � Methods � Data & Processing � Results � Conclusions 11/15/2005 SUNY-ESF SUNY-ESF

Methods ESPM271 Project NDVI-LST Space LST No Evaporation bare soil Dry Edge No Transpiration partial cover full cover Max Transpiration Max Evaporation Wet Edge NDVI 11/15/2005 SUNY-ESF SUNY-ESF

Methods ESPM271 Project VTCI Calculation LST LST max (NDVIi) LST max LST(NDVIi) LST min (NDVIi) LST min NDVI NDVIi = + − LST ( NDVIi ) a bNDVIi LST ( NDVIi ) LST ( NDVIi ) = max max VTCI − = + LST ( NDVIi ) LST ( NDVIi ) ( ) ' ' LST NDVIi a b NDVIi max min min 11/15/2005 SUNY-ESF SUNY-ESF

Outline ESPM271 Project Outline � Introduction � Objectives � Methods � Data & Processing � Results � Conclusions 11/15/2005 SUNY-ESF SUNY-ESF

Data & Manipulation ESPM271 Project Study Area � TN, KY, NC � 31 watersheds for USGS gauging sites 11/15/2005 SUNY-ESF SUNY-ESF

Data & Manipulation ESPM271 Project Drought Monitor Low flow condition in Oct, Nov, Dec of 2005 11/15/2005 SUNY-ESF SUNY-ESF

Data & Manipulation ESPM271 Project MODIS/Terra Vegetation Indices 16-Day L3 Global 1km SIN Grid (MOD13A2) � NDVI band � 5 periods in 2005 � Oct.16 – Oct.31 � Nov.1 – Nov.16 � Nov.17 - Dec.2 � Dec. 3 – Dec.18 � Dec.19 - Jan.3 NDVI (Oct.16 – Oct.31) � Quality band - 16-bit field indicating quality of each NDVI pixel � View angle band - Average view zenith angle for each NDVI pixel 11/15/2005 SUNY-ESF SUNY-ESF

Data & Manipulation ESPM271 Project MODIS/Terra Land Surface Temperature 8-Day L3 Global 1km SIN Grid (MOD11A2) � LST band � 10 periods in 2005 � Oct.16 – Oct.23 � Oct.24 – Oct.31 � Nov.1 – Nov.8 � Nov.9 – Nov.16 � Nov.17 - Nov.24 � Nov.25 - Dec.2 � Dec. 3 – Dec.10 � Dec.11 – Dec.18 � Dec.19 – Dec.26 � Dec.27 - Jan.3 LST (Oct.16 – Oct.23) � Quality band - 16-bit field indicating quality of each LST pixel � View angle band - Average view zenith angle for each LST pixel 11/15/2005 SUNY-ESF SUNY-ESF

Data & Manipulation ESPM271 Project Data Manipulation Process Flow • NDVI/quality/angle process • Mosaicing • NDVI-LST plot • LST/quality/angle process • Reprojection VTCI calculation • LST compositing • Dry/wet edges • Clipping • NDVI-LST extraction MRT ArcGIS macro R 11/15/2005 SUNY-ESF SUNY-ESF

Outline ESPM271 Project Outline � Introduction � Objectives � Methods � Data & Processing � Results � Conclusions 11/15/2005 SUNY-ESF SUNY-ESF

Results ESPM271 Project NDVI-LST Plot & VTCI Whole area (610 * 260 km 2 ) in Oct.16 - Oct.31 NDVI NDVI-LST plot LST VTCI 11/15/2005 SUNY-ESF SUNY-ESF

Results ESPM271 Project NDVI-LST Plot & VTCI Western part (250 * 260 km 2 ) Eastern part (360 * 260 km 2 ) NDVI-LST plot NDVI-LST plot VTCI VTCI 11/15/2005 SUNY-ESF SUNY-ESF

Results ESPM271 Project NDVI-LST Plot & VTCI Whole area (610 * 260 km 2 ) Oct.16 – Oct.31 Nov.1 – Nov.16 Nov.17 – Dec.2 Dec.3 – Dec.18 Dec.19 – Jan. 3 VTCI 11/15/2005 SUNY-ESF SUNY-ESF

Results ESPM271 Project NDVI-LST Plot & VTCI Eastern part (360 * 260 km 2 ) Oct.16 – Oct.31 Nov.1 – Nov.16 Nov.17 – Dec.2 Dec.3 – Dec.18 Dec.19 – Jan. 3 VTCI 11/15/2005 SUNY-ESF SUNY-ESF

Results ESPM271 Project NDVI-LST Plot & VTCI Western part (250 * 260 km 2 ) Oct.16 – Oct.31 Nov.1 – Nov.16 Nov.17 – Dec.2 Dec.3 – Dec.18 Dec.19 – Jan. 3 VTCI 11/15/2005 SUNY-ESF SUNY-ESF

Results ESPM271 Project Low Streamflow Regional Regression Model � Merged VTCI result with other watershed characteristics database � Stepwise regression - With VTCI = − 4 . 06 0 . 88 3 . 87 2 . 96 Q 43 . 9 BFI DA RDL VTCI 5 7 , 10 − = 2 Adj R 77 . 7 % - Without VTCI − = − 4 . 18 0 . 86 2 . 39 0 . 72 Q 26 . 9 BFI DA RDL OM 7 , 10 = 2 R 76 . 5 % 11/15/2005 SUNY-ESF SUNY-ESF

Outline ESPM271 Project Outline � Introduction � Objectives � Methods � Data & Processing � Results � Conclusions 11/15/2005 SUNY-ESF SUNY-ESF

Conclusions ESPM271 Project Conclusions Objective 1. Explore and get familiar with MODIS data & VTCI Conclusion � MODIS L3 NDVI & LST, quality, view angle data � VTCI indicates soil dryness � NDVI-LST plot not always a triangle In the future… � Topographic influences � Geographical scales 11/15/2005 SUNY-ESF SUNY-ESF

Conclusions ESPM271 Project Conclusions Objective 2. Develop an integrated VTCI calculation procedure 3. Apply VTCI in low streamflow modeling Conclusion � VTCI calculation procedure with MODIS Reprojection Tool (MRT), ArcGIS macro codes and R � One VTCI data entered in the model In the future… � Further study on NDVI may lead to model improvement 11/15/2005 SUNY-ESF SUNY-ESF

ESPM271 Project References Andersen, J., I. Sandholt, K. H. Jensen, J. C. Refsgaard and H. Gupta, (2002), Perspetives in using a remotely sensed dryness index in distributed hydrological models at the river-basin scale, Hydrological Processes , 16 (2002), 2973 - 2987. Sandholt, I., K. Rasmussen and J. Andersen, (2002), A simple interpretation of the surface temperature/vegetation index space for assessment of surface moisture status, Remote Sensing of Environment, 79 (2002), 213 – 224. Wan, Z., P. Wang and X. Li, (2004), Using MODIS land surface temperature and normalized difference vegetation index products for monitoring drought in the southern Great Plains, USA, Journal of Remote Sensing, 25(1), 61 – 72. Wang, P., X., Li, J., Gong and C. Song, (2001), Vegetation temperature condition index and its application for drought monitoring, IEEE , 2001. Wang, C., S., Qi, Z., Niu and J. Wang, (2004), Evaluating soil moisture status in China using the temperature-vegetation dryness index (TVDI), Journal of Remote Sensing, 30(5), 671 – 679. 11/15/2005 SUNY-ESF SUNY-ESF

ESPM271 Project SUNY-ESF SUNY-ESF END Outline 11/15/2005