Crop area estimation is one of main application fields of remote sensing The paper focuses on the operational methods for rice area estimation

in which TM data are used to extract the basic rice planting area in a given year 1992

and NOAA AVHRR data are used to predict the change tendency of the rice planting area The basic area data need to be updated for every rice growth period upon the availability of TM data Three methods are presented to extract the base rice area They are:（1） visual interpretation with interactive adjusting on the screen;（2） interactive automatic classification with manual eliminating of the non-rice pixels on the screen;（3） automatic classification with GIS overlay and spatial analysis They can be combined to increase the reliability and the accuracy of rice planting area extraction The SAVI data profile from NOAA AVHRR are merged into a multi-bands image which is classified for predicting the tendency of the change The methods are used to estimate the early rice area in Hubei province of China in 1994 The results are acceptable