TO THE QUESTION ON ACCURACY OF FOREST HEIGHTS MEASUREMENTS BY THE TanDEM-X RADAR INTERFEROMETRY DATA
T. N. Chimitdorzhiev1, M. E. Bykov1, Yu. I. Kantemirov2, I. I. Kirbizhekova1, B. B. Labarov3, A. K. Baltukhaev1
1Institute of Physical Materials Science, Russian Academy of Sciences, Siberian Branch, Sakhyanova str., 6, Ulan-Ude, Republic of Buryatia, 670047 Russian Federation
2Sovzond Company, Shipilovskaya str., 28a, Milan Business Center, Moscow, 115563 Russian Federation
3Institute of Land Management, Cadasters and Melioration, Pushkin str., 8, Ulan-Ude, Republic of Buryatia, 670024 Russian Federation
Keywords: спутниковая радиолокация, дифференциальная интерферометрия, высота леса, валидация, satellite radiolocation, differential interferometry, forest height, validation
Abstract
The paper presents the validation results of the InSAR method for determining the forest canopy height, based on TanDEM-X and ALOS PALSAR data. The research conducted on the territory of the Baikal-Kudara forest area of the Republic of Buryatia (52°10'N, 106°48'E). Forest vegetation is represented mainly by conifers - pine, and spruce, with a small admixture of deciduous trees - aspen, birch, etc. The forest vegetation height was determined by subtracting the digital elevation model (DEM) of the digital terrain model (DTM). DEM is built according to the L-band (wavelength of 23.5 cm) ALOS PALSAR satellite with horizontal co-polarization mode. In the investigation it was assumed that a radar signal of ALOS PALSAR passes all forest thickness and reflected from the underlying surface, made it possible to recover terrain under forest canopy. DTM has been built using the TanDEM-X data (wavelength 3 cm). In this case, it was assumed that the radar echoes scattered from a some virtual phase centers of scattering surface, which characterizes the upper limit of the continuous forest canopy. To check the accuracy of satellite definitions of forest height in study area were made high-precision geodetic measurement of trees heights using electronic total station and the coordinates of geographic control points using differential GPS receivers. The discrepancy between the satellite and ground-based measurements at 11 test sites did not exceed 2 m, which is mainly due to the difference in measurement techniques: height of individual trees by ground methods and continuous forest canopy height using radar interferometry.
|
