Several methods exist for estimating reservoir capacity based on field surveys of reservoir bottom features and the sediment thickness. Recently, the grid method was proposed which reduces the cost by measuring only a few sediment cross sections on selected parts of the reservoir from which the water depths in the remainder can be estimated by interpolation. The objective of the research reported here is to enhance the performance of the grid method by utilizing three additional numerical techniques, namely the Newton and Lagrangian polynomials and the cubic spline. The proposed technique was verified by estimating the known volume of a model dam reservoir in the laboratory. The reservoir volume was successfully evaluated using these numerical methods. It was found that the maximum error using the Newton and Lagrangian methods was 10%, while the maximum error using the cubic spline was as high as 24% and so this method is clearly less applicable. Although the Newton and the Lagrangian methods yield similar results, the latter is preferred as it is easier to use.