Abstract:

Mean forest tree height is one important allometric parametric for the measurement of forest stand volume and above ground biomass. The development of quantitative tree height reversion technique of remote sensing can help to improve the estimation accuracy of forest volume and biomass in large scale, and provide technique supports for the scientific management of forest and globe carbon circle study. As one active microwave remote sensing technique, polarimetric synthetic aperture radar interferometry (POLinSAR), is an advanced and effective tree height inversion technique developed by scientists of the microwave remote sensing field in recent years. Two passes of SIR-C/X SAR L band full polarization single look complex (SLC) data were acquired for the test site established in Hetian of Xinjiang Uygur Autonomous Region. In addition to the SAR data, one scene of Landsat-5 TM data and one scene of Landsat-7 ETM+ data and some sheets of 1∶50 000 topographic maps covering the test site were collected. The POLinSAR data processing and tree height reversion algorithm were studied and programmed for the tree height inversion case study using the SIR-C/X POLinSAR dataset acquired. The primary analysis and evaluation of the achieved tree height inversion result have proved the truthfulness of the POLinSAR data processing and tree height inversion algorithms adopted in this paper. It has been observed that there maybe lots of non-volumetric decorrelation scatters on the poloarimetric coherence images acquired by repeat pass INSAR mode, identifying and kicking out these scatters was an important data processing step for getting true tree height extraction result.

Key words: polarimetric synthetic aperture radar, interferometry, tree height inversion, SIR-C/X SAR