The Application of Detrended Fluctuation Analysis in Rice Blast in the Early Warning

摘要

In this paper, the 5-9 months of 2000-2011 temperature and humidity data, used the detrended fluctuation analysis, obtained how the two data series' non-uniform scaling index changes with time. In order to comprehensive influence of temperature and relative humidity of the two meteorological factors, the temperature and humidity coefficient is introduced. We also proposed a new non-uniform scaling index taking into account the information of temperature and relative humidity, and discusses the possible correlation between temperature and humidity and rice blast. The preliminary results show, A long-range power-law correlation can be found in the time series of temperature and humidity. About 5-15 days before the occurrence of rice blast will appear anomalies of non-uniform scaling index. It reflects the rice blast made a difference of statistical significance to the characteristic of nonlinear system internal of temperature and humidity coefficient. It can predict the occurrence and prevalence of rice blast according to the abnormal changes of temperature and humidity coefficient scaling exponent.