Information content of ERS SAR interferometric products for forest classification in SIBERIA: a case study over the Bolshemurtinskii forest enterprise

Due to their all-weather and day and night capabilities, satellite based Synthetic Aperture Radar (SAR) systems may become an important and relatively cost-efficient source of up-to-date information about Siberian forests. It is the goal of the SIBERIA (SAR Imaging for Boreal Ecology and Radar Interferometry Applications) project to demonstrate the value of these data for Siberian foresters. SIBERIA uses the advantages of dual-frequency, interferometric, and multi-temporal SAR products from the ERS and JERS missions. For each scene of the SIBERLA project area, one ERS tandem pair from fall 1997, one ERS summer image from 1998, and additionally, one JERS image from 1998 are available. To explore the use of the derived SAR products in forest applications and to define a classification methodology a detailed forest inventory data base containing data from more than 36 testsites, with each having a size between 20,000 and 100,000 ha, has been collected. the information content of ERS SAR products (backscattering coefficient, coherence) is investigated over four testsites located in the Bolshemurtinskii forest enterprise. The enterprise is situated in the central part of the Krasnoyarsk Oblast. The forest inventory of these four sites was updated in 1998, so that a good correspondence between the radar and field data is given