Assessment of Nitrate Fluxes in Intensive Aquaculture Region in Godavari Delta Using Spatial Interpolation Kriging

Abstract

In areas with a high concentration of intense aquaculture, nitrate pollution and nutrient enrichment are growing concerns. With predicted future climate changes, these problems are expected to intensify for aquifers and surface waters. The possibility exists to reduce some of these worries through land management and utilization modifications. However, there is much ambiguity surrounding how these alterations will relate. This article uses conventional kriging and empirical Bayesian kriging (EBK) to estimate nitrate levels in India’s intensive aquaculture zone, the Godavari delta. The stable, exponential, rational quadratic, and Gaussian models were used to fit experimental variograms using weighted least squares. The number of neighbors that generated the best cross-validation outcome has been further investigated for the model with the shortest residual sum of the squares. Kriging’s statistical approaches provided the best root mean square error (RMSE) values overall. No additional summary statistics shed any light on the regression method’s selection or settings. After thorough testing, we concluded that many parameters might be better detected using cross-validation. © 2024, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.