Accuracy assessment of the global TanDEM-X Digital Elevation Model with GPS data

作者：Wessel, Birgit^*; Huber, Martin; Wohlfart, Christian; Marschalk, Ursula; Kosmann, Detlev; Roth, Achim

来源：ISPRS Journal of Photogrammetry and Remote Sensing, 2018, 139: 171-182.

DOI：10.1016/j.isprsjprs.2018.02.017

• TanDEM-X
• Digital Elevation Model
• Kinematic GPS
• Vertical accuracy
• Validation

The primary goal of the German TanDEM-X mission is the generation of a highly accurate and global Digital Elevation Model (DEM) with global accuracies of at least 10 m absolute height error (linear 90% error). The global TanDEM-X DEM acquired with single-pass SAR interferometry was finished in September 2016. This paper provides a unique accuracy assessment of the final TanDEM-X global DEM using two different GPS point reference data sets, which are distributed across all continents, to fully characterize the absolute height error. Firstly, the absolute vertical accuracy is examined by about three million globally distributed kinematic GPS (KGPS) points derived from 19 KGPS tracks covering a total length of about 66,000 km. Secondly, a comparison is performed with more than 23,000 "GPS on Bench Marks" (GPS-on-BM) points provided by the US National Geodetic Survey (NGS) scattered across 14 different land cover types of the US National Land Cover Data base (NLCD). Both GPS comparisons prove an absolute vertical mean error of TanDEM-X DEM smaller than +/- 0.20 m, a Root Means Square Error (RMSE) smaller than 1.4 m and an excellent absolute 90% linear height error below 2 m. The RMSE values are sensitive to land cover types. For low vegetation the RMSE is +/- 1.1 m, whereas it is slightly higher for developed areas (+/- 1.4 m) and for forests (+/- 1.8 m). This validation confirms an outstanding absolute height error at 90% confidence level of the global TanDEM-X DEM outperforming the requirement by a factor of five. Due to its extensive and globally distributed reference data sets, this study is of considerable interests for scientific and commercial applications.

• 出版日期2018-5