Prediction on spatial elevation using improved kriging algorithms: An application in environmental management

Abstract

A reliable landscape visualization is important in environmental management. Environmental managers need accurate spatially continuous data across an area to make competent and confident decisions. Spatial interpolation methods can be used to predict the height values at unsampled sites using data from point observations. In this paper, we propose three novel algorithms for spatial interpolation methods using kriging models. Since there are not many findings of how kriging parameters in the semivariance model affect the performance of the spatial interpolators, we explore different parameter selection methods of the kriging algorithm and propose three optimization algorithms (EPO, EKO, and ETL) to improve the performance of the spatial interpolation technique. Maximum likelihood estimation (MLE)-based parameter selection produces substantially less error than traditional semivariogram fitting. Our three models are compared against four MLE-based models. The performance is evaluated by error reduction that the seven models can perform. The strengths of each model are analyzed based on a different set of input sizes coming from two zones of study areas. The resulting errors of our proposed methods are relatively small. Comparing methods adopting the same exponential model, we observe that our recommended EKO performs marginally better than the exponential method based on MLE. In general, the result implies non-significant differences among models. Our work offers additional alternative parameter selection methods with equally high performance, and these methods can be practically used to improve the 3D surface plot in environmental management.