Determination of GPS Coordinate Transformation Parameters of Geodetic Data

The issues of datum transformation of GPS coordinates from the War Office to the WGS84 coordinate systems are a real, persistent and serious problem in Ghana that requires pragmatic measures for solutions. The Ghana national mapping coordinates data are based on a no of 1926, which like many others around the world was realized through conventional methods of surveying. Therefore, GPS based WGS84 coordinates data are not compatible with the War Office. This study a investigating and comparing the similarity (abridged Molodensky, Bursa position dependent (3D projective and 12 determining their associated transformation parameters so that users from all related disciplines can utilize them without any problem. To achieve the aims of this research, nineteen common points between the War Office and WGS84 system were used to estimate the transformation paramet derived transformation parameters, the reference standard deviation, reference adjustment variance and the individual standard deviations of the parameters determined were used. Residual analysis using maximu maximum positive, mean error and standard error was then conducted to evaluate the performance of the transformation models. The test of normality of residuals, deterministic model, correlation coefficient r, t derived transformation parameters, t-test on the correlation coefficient, Root Mean Square error and Root Mean square error distribution were used to evaluate and check the adequacy of the transformation models. Tests were carried out on nineteen points with coordinates known indicated that the Bursa-Wolf, Molodensky War Office coordinates accurate to ±1 m and the abridged Molodensky parameters produce War Office accurate to ±1.9 m. In the case of the position dependent methods, the 3D projective model is in the order of approximately ±1.7 m and ±0.8 m for the 12 parameter linear affine model. Based on the results it was found that both similarity and position dependent methods generate results consistent with their accuracies that have been quoted in scientific literature