Original paper

Using MLR to model the vertical error distribution of ASTER GDEM V2 data based on ICESat/GLA14 data in the Loess Plateau of China

Zhao, Shangmin; Cheng, Weiming; Zhou, Chenghu; Liu, Haijiang; Su, Qiaomei; Zhang, Shifang; He, Weican; Wang, Li; Wu, Wenjiao

Zeitschrift für Geomorphologie, Supplementary Issues Volume 61 Supplementary Issue 2 (2017), p. 9 - 26

published: Nov 1, 2017

DOI: 10.1127/zfg_suppl/2016/0325

BibTeX file

ArtNo. ESP023106102001, Price: 29.00 €

Download preview PDF Buy as PDF

Abstract

Using the multinomial logistic regression (MLR) model, this study quantitatively simulate the vertical error distribution of ASTER GDEM V2 data based on the ICESat/GLA14 data and land surface factors (including topographic, NDVI and land use factors) in the Loess Plateau of China. Research results show: (1) there is a positive correlation between the vertical error and the topographic factors includ- ing elevation, relief and slope factors. With regard to the aspect factor, a symmetrical aspect direction for the distribution of the negative and positive error values is found. In general, the vertical error decreases with increasing NDVI values. With regard to land use factor, the highest vertical error distributes in forestland and grassland. (2) The vertical error distribution probability shows a near normal distribution with marginal negative skewness. (3)The accuracy of the model results is estimated to be higher than 70% based on the different checked datasets including the simulated and checked ICESat/GLA14 data and ground control points in topographic maps.

Keywords

land surface factors • vertical error distribution • aster gdem v2 • icesat/gla14 • mlr model