Characterizing Slope Correction Methods Applied to Satellite Radar Altimetry Data: A Case Study Around Dome Argus in East Antarctica
Gang Hai,Huan Xie,Wenjia Du,Menglian Xia,Xiaohua Tong,Rongxing Li
DOI: https://doi.org/10.1016/j.asr.2021.01.016
IF: 2.611
2021-01-01
Advances in Space Research
Abstract:Slope correction is important to improve the accuracy of satellite radar elevation measurements by mitigating the slope-induced error (SE), especially over uneven ground surfaces. Although several slope correction methods have been proposed, guidance in the form of stepwise algorithm on how to implement these methods in processing radar altimetric data at the coding level, and the differences among these methods need to be presented and discussed systematically. In this paper, three existing types of slope correction methods-the direct method (DM), intermediate method (IM), and relocation method (RM, further divided into RM1 and RM2)-are described in detail. In addition, their main differences and features for various scientific applications are analyzed. We conduct a systematic experiment with CryoSat-2 Low Resolution Mode (LRM) data in a physically stable area around Dome Argus in East Antarctica, where in-situ measurements were available for comparison. The slope correction is implemented separately using the three methods, with the latest high-accuracy Reference Elevation Model of Antarctica (REMA) as the a-priori topography model. The bias and precision of the slope-corrected CryoSat-2 data results from the RM2 is -0.18 +/- 0.86 m based on the comparison with the field Global Navigation Satellite System (GNSS) data. The results from the RM2 indicate higher precision compared to those from the RM1. According to the correlation analysis of the slope-corrected CryoSat-2 data results (RM1 and RM2), the bias enlarges and the precision becomes worse when the surface slope increases from 0 to 0.85 degrees. After a comprehensively comparative analysis, we find that the results from the RM1 and RM2 are superior in precision (0.93 m and 0.86 m) with respect to the GNSS data. The relatively low precision (1.22 m) from the IM is due to the potential error from the a-priori digital elevation model (DEM). The DM has the lowest precision (2.66 m). Another experiment over rough topography in West Antarctica is carried out for comparison, especially between the RM1 (precision of 15.27 m) and RM2 (precision of 16.25 m). In general, the RM is recommended for the SE elimination among the three methods. Moreover, the RM2 is firstly considered over smooth topography due to the superior performance in bias and precision, while the RM1 is more suggested over the rough topography because of the slightly smaller bias and better precision. The IM relies much on the accuracy of the a-prior DEM and is not usually recommended, because of the strict requirement in the sampling time between the radar altimetry data and the a-priori DEM to avoid any surface change over time. (C) 2021 COSPAR. Published by Elsevier B.V. All rights reserved.