Forest Parameters Derivation from SAR Data in the Czech Republic
Slacikova, Jana; Potuckova, Marketa
Charles University in Prague, CZECH REPUBLIC

The study focuses on the possibilities of the use of SAR data in the forest management in the Czech Republic. The aim is to derive forests parameters such as tree height or growing stock volume from various SAR datasets.
Integration of the SAR data complements existing methods of the National Forest Inventory and eliminates some current problems. Synthetic Aperture Radar allows regular repetition of the measurements and provides data of large areas. In contrast to optical satellite data could penetrate fog and clouds and compared to very accurate measurements from airborne laser scanning can provide regular data collection of extensive areas. The National Forest Inventory uses nowadays data originated from the terrestrial measurements, aerial images and space-born optical images. The terrestrial data are collected every 10 years and the period of the acquisition of the aerial data is 3 years. The space-borne data are used for the estimation of annual change and the new clear-cuts detection.
The motivation of the study is to provide a solution for forest parameters acquisition in between the field and aerial data collection period or after events like the windthrows caused by strong winds, hence, the area of interest is particularly forested mountainous regions. Therefore, the study area of approximately 15x30 km is situated in the Ore Mountains on the borders between the Germany and the Czech Republic. The digital surface model from SRTM-X with 30 m spatial resolution is available for this area. The surface heights in a grid with 10 m spatial resolution were provided by Forest Management Institute of Czech Republic for validation of the results.
Four SLC scenes from TerraSAR-X with 11 days time baseline were examined with the interferometric approach. For the areas with dense vegetation the results turned out completely decorrelated which indicates that even the 11 days baseline is too long for X-band wavelengths for these conditions. Further, the tandem TSX-TDX interferometric acquisitions will be explored. Correlation between coherence and growing stock volume is expected but the effect of mountainous topography has to be examined. Additionally, the phase information will be used to derive the forest height and the X-band backscatter intensity is going to be investigated. Recently the intensity and interferometric phase of nine scenes of L-band ALOS PALSAR data are examined.

Acknowledgment:
TerraSAR-X and SRTM-X data were obtained from the German Aerospace Center (DLR) under the student grant project of Charles University in Prague (project ID HYD1239).
The ALOS PALSAR data were used during the ERASMUS internship at the Department for Earth Observation of Friedrich-Schiller University in Jena and were provided by Japan Aerospace Exploration Agency (JAXA) within the Kyoto & Carbon Initiative.
Validation data were provided by the Forest Management Institute of the Czech Republic.