Inversion of a Radiative Transfer Model for Estimation of Rice Canopy Chlorophyll Content and ALOS Imagery

Inversion of physical models which are based on the concepts of radiation propagation within a canopy structure, has demonstrate to be among the successful approaches in retrieving canopy parameters such as leaf area index (LAI) and chlorophyll, using remotely sensed data. In this study rice chlorophyll content as an indicator of crops health status, was estimated using inversion techniques. To do so, the PROSAILH radiative transfer model has been utilized taken by with images from AVNIR-2 sensor on board of ALOS satellite. An extensive field campaign was conducted during June and July 2009 in Amol city, northern part of Iran. Sixty plots of 20 by 20 m2 were randomly selected and in each of them four to seven subplots, depending of homogeneity of plots were adopted. In each subplot chlorophyll was measured using an SPAD-502 chlorophyll - meter instrument. The MODTRAN model was used for radiometric corrections of the images, and the PROSAIL model was modified in order to be used for AVNIR-2 bands. After sensitivity analysis, the PROSAIL model was inverted by an iterative optimization method and the amount of the estimated chlorophyll was retrieved at the leaf and canopy level. To evaluate the results of inversion, R2 and RMSE criterion between the measured and estimated chlorophyll were calculated (R2=0.57, RMSE= 0.47). The results approved the ability of PROSAILH model to retrieve rice plant chlorophyll content with an acceptable accuracy required for ecological modeling or crop land monitoring.