A Fast High-Precision Geospatial Grid Interpolation Algorithm

Addressing the persistent trade-off between computational accuracy and execution efficiency in generating high-resolution geospatial grids, this paper proposes a novel high-precision interpolation algorithm specifically designed for rapid, large-scale geospatial grid generation. The algorithm first employs a Gaussian weighted quadratic surface as its interpolation kernel—a method that balances the smoothness of results with adaptability to non-linear terrain variations to ensure high accuracy. Subsequently, to tackle the model's high computational cost, a targeted hybrid optimization framework was designed, leveraging a KD-Tree to accelerate neighbor searches and multithreading to parallelize the independent grid computations, thereby substantially enhancing execution efficiency. Experimental results demonstrate that: 1) The proposed algorithm outperforms three traditional methods (Moving Surface Fitting, Natural Neighbor Interpolation, and Spline Interpolation), achieving reductions in Mean Absolute Error (MAE) and Root Mean Square Error (RMSE) of up to 63.6% and 61.2%, respectively, while attaining a coefficient of determination (R²) as high as 0.998611. 2) The optimization framework delivers a 24-fold speedup in overall runtime without any loss of accuracy. This study offers a high-precision, high-efficiency solution for large-scale geospatial data processing.