5. Conclusions
We have developed a new systematic method for detecting the phenological stages of paddy rice from time-series MODIS data. The method consists of three procedures: (i) prescription of multi-temporal satellite data; (ii) filtering the time-series EVI data; and (iii) specifying the phenological stages from the smoothed EVI time profile. Our method can determine the planting, heading, and harvesting dates in paddy fields. The estimation was validated against statistical data from 30 test sites. In filtering to reduce noise, the wavelet transform using Coiflet 4 gave the best time profile of the EVI for detecting the phenological stages. Finally we proposed this new method using the wavelet transform: Wavelet based Filter for determining Crop Phenology (WFCP).
As crop calendars are not available in many regions, WFCP is particularly useful for determination of regional characteristics of rice phenology. We believe that WFCP has potential applicability to other crops and some natural vegetation, whose EVI profile has an obvious peak with a bell-shape (e.g., wheat, deciduous forest, grassland). However the mother wavelet used and the threshold for determination of phenological stages will depend on the species.