STANDWISE CHANGE DETECTION FOR GROWING STOCK USING REPEAT-PASS ALOS PALSAR / PALSAR-2 DATA

Abstract

This study demonstrates the possibility of detecting the changes of growing stocks in mountainous forest stands derived from ALOS PALSAR and PALSAR-2 images. The ALOS PALSAR were obtained over the Kwangneung Experiment Forest (KEF, Korea) during the period of nineteen and a half months from the April 26, 2009 to December 12, 2010, whereas the PALSAR-2 data were acquired on the April 7, 2015. The KEF test site comprises 58 stands, which cover approximately 1,000ha and have steep slope topography. Owing to topographic effects of SAR data in mountainous areas, the DEM-assisted topographic normalized backscattering coefficient γ<sup>0</sup> was applied to the evaluation of the relationships between the ALOS PALSAR / PALSAR-2 HV backscatter and the field inventory–based stand stock volume. The results indicate that: 1) the γ<sup>0</sup> values for the volume obtained from ALOS PALSAR data on December 12, 2010 show a gradual increase higher than those computed from the data on April 26, 2009, here the γ<sup>0</sup> value increases in accordance with an increase in the volume: 2) the γ<sup>0</sup> values determined from the PALSAR-2 data increase with the same inventory-based volume, when compared with those computed from both ALOS PALSAR data. They also increase substantially as the values of the volume rise, with the exception of the volume interval from 130 m<sup>3</sup> ha<sup>−1</sup> to 160 m<sup>3</sup> ha<sup>−1</sup>. This is understandable because the volume of the aforementioned interval has been reduced through clearing. Consequently, the γ<sup>0</sup>–based relationship between PALSAR-2 HV backscatter and growing stock can lead to detecting the stand growth changes in the KEF of Korea.