It is aimed to enable analyzing vegetation growth conditions at multiple times of a year accurately using radar images obtainable from a flying body such as artificial satellites, etc. A plurality of radar images of a ground surface of a same target area, which have been taken at multiple times of a year by a radar device mounted on a flying body, are acquired. The acquired plurality of radar images is stored in a map database. While using as a criterion image a radar image of the plurality of radar images stored in the map database, taken at a predetermined time in the multiple times of a year, other radar images than the criterion image of the plurality of radar images, taken at other times than the predetermined time in the multiple times of a year, are aligned with the criterion image, respectively. Then, backscatter coefficients of specified areas in the plurality of radar images are extracted. Based on a backscatter coefficient of a specified area in the criterion image of the plurality of radar images stored in the map database, backscatter coefficients of other radar images than the criterion image of the plurality of radar images are calibrated. And, based on a correlation between backscatter coefficients of radar images and growth values of vegetation shown in the radar images, growth values of vegetation shown in other radar images than the criterion image of the plurality of radar images, whose backscatter coefficients have been calibrated in the calibrating step, are calculated using the calibrated backscatter coefficients.
