Impact of Various Global Digital Elevation and Crustal Density Models on Accurate Topographic Correction for Geoid Modelling Over East Malaysia
摘要
High-accuracy geoid modelling in complex terrain such as East Malaysia requires robust DEMs, high-quality gravity observations, and realistic crustal density information. Although global models of topographic and crustal density such as UNB_TopoDens and CRUST1.0, respectively have shown promising results internationally, their applicability and impact on geoid computation in the Malaysian region remain largely unexplored. Thus, this study investigates the influence of six global DEMs (ALOS, ASTER, SRTM, FABDEM, MERIT, and NASADEM), together with density information derived from two global models namely UNB_TopoDens and CRUST1.0 on gravimetric geoid determination using the Least Squares Modification of Stokes’ Formula with Additive Corrections (LSMSA). A comprehensive gravity dataset comprising terrestrial, airborne, and marine observations was combined with the WHU-SWPU-GOGR2022S global geopotential model for long-wavelength recovery. The geoid models based on different DEMs were evaluated using 43 GNSS-levelling benchmarks, revealing that MERIT, ALOS, and ASTER show superior vertical consistency. The impact of laterally varying crustal density was then investigated through its effect on Bouguer reductions, free-air anomalies, and final geoid heights. Although the resulting geoid surfaces differ only slightly, typically within ± 0.10–0.20 m, significant enhancements occur in mountainous regions such as the interior of Sabah. GNSS validation demonstrates that variable-density geoids (UNB and CRUST1.0) achieve lower mean differences and RMSE (~ 0.520 m) compared with the constant-density solution (~ 0.548 m). The findings confirm that global crustal density models, while not drastically altering the geoid, contribute measurable accuracy improvements and should be incorporated in future regional vertical datum modernisation efforts for East Malaysia.