Abstract Author: Chipanshi A.C. , Rahman M.M. and Fitzmaurice J.
Abstract Title: Estimating Soil Moisture for Drought Monitoring from MODIS Data Using a Thermal Inertia Approach
Abstract: Level 2 MODIS data for computing NDVI and level 3 MODIS data for computing day- and night-time land surface temperature were used to compute cloud and shadow free surface reflectance and an index of apparent thermal inertia respectively, across a 400-km2 agricultural land, south of Winnipeg Manitoba, during the 2007 growing season. The study area is comprised of agricultural fields specializing in small grains on clay soils. The NDVI values were used as a surrogate for albedo in the Thermal Inertia equation. Thermal inertia is a physical property of material that describes the impedance to variations of temperature and it is highly correlated with the soil moisture content (Committee of Soil Physical Properties Measurement of Japan 1979). Thermal inertia indices as derived from MODIS data were then correlated with measured soil moisture at 5-cm and 15-cm depths. Soil moisture measurements were made with a theta probe from pre-selected points within a grid that coincided with the MODIS image pixels. Surface soil moisture was strongly correlated with the thermal inertia index (R= 0.83). At deeper depths (15cm), the correlation coefficient dropped to about 0.40, most likely due to the lag in energy penetration below the soil. Because of the moderate resolution of MODIS data (1km X 1km) and its wide availability, there is potential to use MODIS data for coarse scale scanning of agricultural fields for problems related lack of water on a regular basis. When such areas are identified, field-scale probing or program targeting can then be made without significant omissions.