Verification Of Land Moisture Estimation Model Based On MODIS Reflectances In Agricultural Land

From this research, it is found that reflectances in the 1, 2, and 6" channels (R1.R2.86) of MODIS have high correlations with surface soil moisture (% weight) at 0-20 cm depth. An index called Land Moisture Index (LMI) was created from the linier combination of R1 (%), P2 (%), and R6 (%). The MODIS reflectances and field soil moisture in paddy field taken from the Central and East Java during July September 2005 are applied into the previous model which have been generated from data during July September 2004. The result showed that there was a high correlation between Land/Soil Moisture (SM) which was measured from field survey, and LMI. which was generated from the MODIS refectances. The best model equation between SM and LMI is the power regression model, which has the coefficient of determination of 88%. It is implied that soil moisture condition can be obtained from the MODIS data using Land Moisture Index. Therefore, the spatial information of drought condition analysed throught the soil moisture in the agricultural land can be provided from the MODIS data.