New geoid model for Egypt using the shallow-layer method and its validation

Abstract

Geoid determination for Egypt faces several challenges. There are large gravity data gaps, gravity data come from different sources, were measured during different decades, and some lack information about their precision. The aim of the paper is to find out a new geoid model for Egypt using the advantages of the shallow-layer method. The shallow-layer method is different from the traditional Stokes and Molodenesky geoid determination methods as it comes from the definition of the geoid itself. In our work, we apply the shallow-layer method to determine a 5′ × 5′ geoid model for Egypt which ranges from (21∘ ≤ ϕ ≤ 32∘; 24∘ ≤ λ ≤ 37∘). The global digital topographic model DTM2006.0, the Danish National Space Center DNSC08 model, the EGM2008 gravity field model, and the CRUST2.0 crust model were used to distinguish boundaries for the shallow-layer and determine its interior structure. To validate and test the computed geoid, four gravimetric geoid models for Egypt were computed using different techniques, the Meissl-geoid using modified Stokes kernel after Meissel, the Wong-geoid using modified Stokes kernel after Wong and Gore, the Heck-geoid using modified Stokes kernel after Heck and Grüninger, and the window-geoid using window technique. All the computed geoid models were validated using the available GPS benchmark stations in Egypt. The difference between the computed geoid models and GPS benchmark station shows that the shallow-layer method gives the most precise geoid with the narrowest range and smallest standard deviation. The standard deviation of geoid differences for the geoid computed using the shallow-layer method (Shen-geoid) drops by about 0.37 m compared to the Stokes/Meissl geoid, which means that the calculated Shen-geoid fits GPS benchmarks better than the computed gravimetric geoid (using the available gravity data) for Egypt. The results demonstrate that the shallow-layer method can estimate the geoid at higher accuracy.