Analysis of the Impact of Topographic Features and Baselines on the Accuracy of Digital Elevation Models Extracted Using InSAR Technique

Abstract

Digital Elevation Models (DEMs) represent a fundamental dataset in a wide range of geospatial, engineering, and environmental applications, as they provide a precise quantitative description of surface morphology. This study investigates the influence of key acquisition and terrain-related parameters on the accuracy of DEMs derived through Interferometric Synthetic Aperture Radar (InSAR) processing. The examined factors include satellite orbit direction (ascending/descending), temporal baseline length, perpendicular baseline length, terrain morphology, and vegetation density. The analysis is based on multiple interferometric pairs acquired in IW-SLC mode from the Sentinel-1 mission, allowing for a systematic evaluation of DEM quality and reliability under varying geometric configurations.

Two topographically contrasting environments were considered: a flat area characterized by gentle slopes and sparse vegetation cover, and a rugged area characterized by steep relief and dense vegetation. The results reveal a strong dependency between temporal baseline length and interferometric coherence, with shorter temporal baselines significantly enhancing phase stability and improving elevation accuracy in both terrain settings. Moreover, descending orbit acquisitions consistently outperformed ascending configurations in terms of reducing elevation errors. The lowest Root Mean Square Error (RMSE) values were obtained over the flat terrain, whereas higher errors were observed in the rugged environment due to pronounced temporal decorrelation and increased topographic complexity.

The findings highlight that baseline optimization is a critical determinant of InSAR-derived DEM accuracy and that terrain morphology plays a decisive role in controlling coherence preservation and elevation precision. These results provide practical guidance for optimizing Sentinel-1 acquisition strategies for reliable topographic mapping across diverse landscape conditions.