Monitoring Maize Lodging Disaster Via Multi-Temporal Remote Sensing Images

The study aimed to monitor maize lodging in a large scale by using multi-temporal HJ-1B CCD images. The variation of vegetation indexes before and after lodging was analyzed. The sensitive vegetation index of maize lodging was selected by correlation analysis method. The remote sensing monitoring model of maize lodging disaster was constructed, which used to map maize lodging distribution and disaster grade at large scale. The model was validated by field measured samples at last. Results showed that correlation between ΔRVI and lodging ratio was highest. The ΔRVI can be used as the best vegetation index for quantitative inversion of maize lodging by remote sensing. The overall accuracy of disaster grade classification was 87.5%, and Kappa coefficient was 0.817. It indicated that the model developed in the study could be used to map maize lodging coverage and spatial distribution of disaster grade.