Evaluating Urban Sewage Sludge Distribution on Agricultural Land Using Interpolation and Machine Learning Techniques

The European Union promotes the development of a sustainable approach to solid waste management and disposal. Sewage sludge (SWS) is a good example of this economic model because it has fertilizing and soil-conditioning characteristics. This study employed a conventional manure spreader to evaluate the distribution of SWS on agricultural land. Various interpolation methods and machine learning models were employed to analyze the spatial distribution patterns of the sludge. Data were collected from 15 sampling trays across a controlled field during three separate trials. Statistical analysis using ANOVA highlighted significant variations in sludge quantities along the longitudinal axis but not along the latitudinal one. Interpolation methods, such as spline, cubic spline, and inverse distance weighting (IDW) were used to model the distribution, while machine learning models (k-nearest neighbors, random forest, neural networks) classified spatial patterns. Different performance metrics were calculated for each model. Among the interpolation methods, the IDW model combined with neural networks achieved the highest accuracy, with an MCC of 0.9820. The results highlight the potential for integrating advanced techniques into precision agriculture, improving application efficiency and reducing environmental impact. This approach provides a solid basis for optimizing the operation of agricultural machinery and supporting sustainable waste management practices.