Improving the accuracy of local gravimetric geoid modelling using simulated terrestrial gravity data

• Main Authors: Jalal, Shazad Jamal, Musa, Tajul Ariffin, Md. Din, Ami Hassan, Wan Aris, Wan Anom, Pa’suya, Muhammad Faiz
• Format: Article
• Published: Springer 2022
• Subjects: G Geography (General); NA Architecture
• Online Access: http://eprints.utm.my/103894/; http://dx.doi.org/10.1007/s00024-022-03092-y

Geodetic observations in any country or region require a precise local geoid model. Hence, this study has improved the geoid modelling using simulated terrestrial gravity data. However, the sparse and limited number of terrestrial gravity data is the primary reason for the inability to develop an accurate gravimetric geoid model in Iraq, including within the Sulaymaniyah Province selected as the case study in this research work. The ability to use the global navigation satellite system (GNSS) to determine orthometric height has been restricted due to the lack of precise geoid models within the region. Hence, 3327 gravity points from several international and local datasets were applied, 160 of which were collected via gravity survey, to simulate and model the gravity in the Sulaymaniyah Province. A stepwise multiple linear regression with a correlation coefficient (r) of 0.997 and a determination coefficient of (R2) of 0.993 (both very close to 1) was deployed to extract the geographical coordinates and the orthometric height of the points to formulate a cutting-edge gravity model. Next, 120 local gravimetric models were generated using software from KTH (a university in Sweden) with two conditions: (1) the simulated gravity data were composed of a variety of grid and cap sizes, and (2) both the interpolated gravity data and the terrestrial data were combined with the downloaded World Gravity Map 2012 (WGM2012) data. Next, ITU_GRACE16 and IGGT_R1 global geoid models (GGMs) were used to support the cap size area. As the quadratic model fit the 11 available global positioning system (GPS)-levelling points, the simulated gravity data revealed the lowest root mean square error (RMSE) result of ± 17 cm when using IGGT_R1 GGM, in comparison to the other two datasets. Meanwhile, EGM2008 scored an RMSE of ± 31 cm. In conclusion, this new data entry method improves the accuracy of local geoid models by mathematically simulating the gravity data instead of interpolating them.