Comparison of estimated transplanting dates derived from MODIS data and COSMO-SkyMed data

For prediction of rice yield at theregional and global scales, it is necessary to estimate the geographical distribution of transplanting date. An optical satellite sensor is helpful in estimating transplanting dates. In particular, MODIS data are often used. However, obtainingclearimages from MODIS is difficult because clouds often obscure areas of interest. On the other hand, the COSMO-SkyMeddata,which uses SARsensors, can supply data withhigh spatial and temporal resolution regardless of atmospheric conditions. However, continuous monitoring of transplanting dates has not previously been done using SAR data.In this paper, the geographical distribution of transplanting date was estimated from both MODIS images and COSMO-SkyMedimages, and the results were compared. The transplanting dates could be estimated accurately from COSMO-SkyMedimages, but there were several estimation failures from MODIS images; these failures wereprimarily due tolowspatial resolution.