Robert Sylvander’s research while affiliated with Swedish University of Agricultural Sciences and other places

MAESTRO I data from the Flevoland lest site in the Netherlands have been used for this study. From a complementary ground data collection, the bole volumes of a large number of stands of mainly poplar and ash have been estimated. The relationships between radar backscattering and bole volumes have been examined experimentally and theoretically. In the case studied, the radar backscattering sensitivity to bole volumes increases as the wavelengths increase. and is highest at P band. The sensitivity of the radar backscattering to variations of forest canopy components and moisture contents has been investigated theoretically at P band. It is important to obtain information on such variations before the inversion problem can be solved. The present study indicates a potential for bole volume determination by P-band SAR.

Discusses the information content in satellite SAR images of forested areas. As methodology, the authors study temporal changes in ERS-1 (C-band) and JERS-I (L-band) images. Three ERS-1 scenes and five JERS-1 scenes are used. Experience from the MAESTRO-1 data set demonstrates a higher sensitivity of L-band to biomass attributes, while C-band images are not that sensitive to such parameters. A multi-temporal C-band analysis over a year helps however to discriminate between deciduous and coniferous stands and to detect changes in moisture conditions. The test area is located 20 km north-west of Umea, a town in northern Sweden by the Gulf of Bothnia. The forest consists mainly of naturally generated Norway spruce and Scots pine with ages up to 150 years. Birch is the only prevalent broadleaved species. Maximum values of tree heights and volumes are 25 m and 300 m³/ha respectively. Most of the area is located in flat terrain. Stand sizes vary between 3 and 25 ha. Eleven stands within the scenes have been chosen to be test sites. Four stands of Norway spruce, five stands of Scots pine, one birch stand and one clear-cut were selected. Also, one bog has been examined

Two experiments with differential GPS were organized in northern Sweden. The first study was organized as helicopter flights along lines with known positions. Obtained GPS coordinates corresponded to a mean deviation of a few metres and a standard deviation of about the same size. This was judged to be satisfactory for navigation purposes in connection with forest fertilization or liming. In the second study the accuracy of repeated measurements for estimation of coordinates of sample plots in forested terrain resulted in standard deviations of a few metres. Stand boundaries were mapped with accuracy sufficient for maps on the scale of 1:20000, also under closed canopy

The analysis of polarimetric SAR observations acquired by the JPL polarimetric multi frequency (P-, L- and C-band) SAR during the MAESTRO-1 campaign has continued. We study the Flevoland test site, which extends over two scenes of mainly forested areas and one scene of agricultural areas. The data type studied is the high resolution complex one look product. Natural extended targets have been used to correct for cross-talk and phase errors, and co-channel imbalance is removed by using trihedral reflector responses. Additional field measurements have provided us with a larger amount of ground data. 'we have measured the basal area and the mean stand height of over one hundred stands of mainly poplar, ash, pine and spruce. The correlations between radar backscattering and the measured parameters generally increase with increasing wavelength. The correlation is however strongly species dependent.