For many surveying applications in civil engineering, Global Navigation Satellite Systems (GNSS) are proving to be an attractive data collection and measurement technology. Real-Time kinematics (RTK) and kinematics are the most well-known observation techniques. However, the RTK and kinematics accuracy still falls short of the accuracy needed for some engineering projects, particularly those depending on the heights. This research aims to improve the accuracy of heights derived using RTK and kinematic methods data and to evaluate the using of improved heights in estimating earthworks. In the improvement process, the idea of eliminating systematic errors was used. The amount of error was first computed at a control point at the start of the task, and then the amount of this error was subtracted from the heights of the remaining points. A variety of statistical criteria and measures of quality of fit were employed to evaluate the performance of the models under study, and the outputs of the various models were then compared. The study's conclusions showed that the accuracy of heights from RTK and kinematic data had increased by 71% and 52%, respectively. Researchers or professionals that need to successfully adapt RTK and kinematic techniques to their applications might benefit from the experiences with the increased RTK and kinematic data.