The inversion of average vegetation height using ICESat GLAS and MODIS data: a case study of three provinces in Northeastern China

Abstract

The average vegetation height can be accurately extracted from ICESat GLAS data, however, a certain spatial interval exist in laser strips and dots reduces the mapping accuracy of average canopy height after the interpolation of the GLAS data. The MODIS-BRDF/albedo data consist of canopy structural data, such as LAI, canopy height etc. So the combination of ICESat GLAS and MODIS data can be obtained more accurate distribution of average canopy height and achieve the distribution of continuous canopy height. In this paper, the GLAS / MODIS data were collected in forest-rich three provinces in northeastern China. We firstly filtered GLAS waveform data and get the average vegetation height, and then selected the optional MODIS-BRDF / albedo bands to retrieve the average vegetation height. An artificial neural networks model was esTablelished by training the MODIS BRDF data, and finally obtained the average vegetation height over the whole three provinces. The fusion method between GLAS data and optical remote sensing image was proposed to make up for their shortages and obtained a continuous distribution of average vegetation height. It increases the analysis dimensions of forest ecosystem and produces more accurate data for forest biomass and carbon storage estimates.