In order to retrieve land surface temperature (LST) from satellite remote sensing data accurately, the information of atmospheric condition for atmospheric correction, especially atmospheric water vapor content, is absolutely imperative. When hourly LST is retrieved from Multi-functional Transport Satellite (MTSAT) data, the hourly atmospheric water vapor products and those which covers the area correspond with the MTSAT observation area are needed. In this study, we developed an optimum water vapor product for LST retrieval from MTSAT data. We assessed the accuracy of water vapor products, which are obtained from MTSAT IR data and reanalysis data, and that of LST retrieved from MTSAT data using each water vapor product for atmospheric correction. It was revealed that the reanalysis PW product is better for LST retrieval from MTSAT data, but the accuracy of the PW is reduced in the high altitudes due to the low spatial resolution of the reanalysis products. Therefore, we proposed the method for improving the accuracy of the reanalysis PW product with refinement using both digital elevation model and reanalysis products at each pressure level. The RMSE of LST retrieved from MTSAT data with the refined reanalysis PW product was less than 2.0 K and the spatial distribution of LST can be figured out in more detail using it.
Harvesting at the optimum timing is critical for production of hig-quality rice. Relationship between the optimum harvesting date (i.e., maturing date) and spectral reflectance was investigated using reflectance data from airborne-sensor and hig-resolution satellite sensors. The reflectance at around 670 nm was best correlated with the date of maturity. A new remote sensing method was developed for predicting the maturing date of rice in each field by combining the reflectance data and the data of the spatial and temporal distribution patterns of maturing date in a region.
In reflecting spectra of the earth’s surface, though the magnitude varies with the intensity of sunstroke, the angle of land inclination, the observation angle of the sensors and so on, the shape is less deformed with these effects. From this point of view, we have developed a spectral shape-dependent analysis by normalizing each band-radiance with their mean value. Here, we applied this normalization method to spectral radiometer data for a forest of Larix kaempferi using 3 components of wavelength ranges, Green, Red, and NIR. Each normalized spectral components measured on the ground observation site become almost stable in the sunlight daytime, and their seasonal variations are smooth. Moreover, respective band values of normalized reflectance obtained from ground site and satellite data are almost same. Therefore this method is effective for analysis of vegetation vitality. The normalized reflecting index of NIR is linear to NDVI, and those of Green and Red are sensitive to the variation of leaf color.