Abstract:It should be taken seriously to monitor powdery mildew (PM) of wheat quickly and accurately in continuous space since the outbreak of this disease may cause substantially reduction of output and lower the quality of wheat. A new method was presented to monitor the external environment of the PM pathogen based on the remote sensing information including greenness wetness and land surface temperature (LST) and the meteorological information including temperature and the number of rainy days. Three kinds of models were established using algorithms Fisher liner discriminant analysis (FLDA) and support vector machine (SVM), respectively. And these models were individually used to supervise the incidence of PM in the western regions of Guanzhong Plain, Shaanxi Province. The results showed that the integrating models (integrating the remote sensing information and the meteorological information) reached at a satisfactory accuracy (FLDA: 74%, SVM: 77%). The integrating models performed better than the meteorological models (FLDA: 60%, SVM: 65%) and the remote sensing models (FLDA: 65%, SVM: 70%). The SVM method outperformed the FLDA method. The results indicated that the integration of remote sensing information and meteorological information can promote the monitoring accuracy of plant diseases.