Assessment and Update of Forest Classification Using Multi-parameter Satellite Sar Data for Ne China

The objective of this paper is to improve as well as to assess the performance of mapping forest by ENVISAT ASAR data and update forest map from ERS SAR images. The preliminary results of there test sites show it is very potential to use HH/HV ASAR AP data for forest map update. Tuqiang site have been undergoing dramatic changes during the last several decades due to the regeneration after forest fire. The Multi-temporal Stand Deviation feature extraction can improve the discrimination ability among some land cover types, especially for summer-winter multi-temporal compositions of ASAR data. Totally, about 900 scenes, covering North-East China, have been acquired since August 2004, and most of them are HH/HV polarization composition. Based on this processing method, the forest/non-forest map of whole North-East China have been generating. Assessed on basis of the land use map from 2001 ETM+ image, the accuracy of the maps from multi-temporal ASAR AP data can meet the request of mapping the boreal forests at large scale.