Determination of 3D Transformation Parameters for the Sri Lankan Geodetic Reference Network Using Ordinary and Total Least Squares

Since the Global Navigation Satellite System (GNSS) has been a common and convenient practice for many local geospatial applications, it is required to transform the geodetic coordinates from global to local datums. In the Sri Lankan context, datum transformation is performed from the WGS84 to the Everest (1830) geodetic datums. Usually, transformation parameters are determined using the least squares adjustment technique. This research compares the ordinary and total least squares adjustment techniques in three coordinate transformations: Bursa-Wolf; Molodensky-Badekas; Veis, and determines new transformation parameters for coordinate transformation in Sri Lankan local geodetic reference network. Cross-Validations of hold out and leave-one-out were applied for the validation of the results, in which the latter performed best. The research concludes that though the Bursa-Wolf model is simple, Molodensky-Badekas and Veis models are more suitable for coordinate transformation in Sri Lanka.