Skip to content
SDGs Center UDINUS

Evidence record · UDINUS · SDGs

Innovation in Crop Nutrition Planning Based on Rainfall Prediction Using Singular Spectrum Analysis and Boosting to Optimize Agricultural Management

Rektorat UDINUS 2026 Published 2026-10-30

Evidence record · UDINUS · SDGs

Unit
Rektorat UDINUS
Period
2026
Goals
Record
3e736074-00c6-4b82-b335-6bd6e070ed6f
Published
2026-10-30

Universitas Dian Nuswantoro conducted research to develop a rainfall prediction framework for flexible and precise crop nutrient scheduling. The activity addressed the high variability of rainfall in tropical climates, where weather uncertainty can lead to inefficient fertilization practices because of nutrient loss. The research focused on using rainfall predictions to support agricultural management and responsive decisions about fertilizer application.

The activity used an hourly rainfall dataset obtained from IoT sensors. Spline interpolation was applied to handle missing data. Singular Spectrum Analysis was used for signal decomposition and noise reduction, including the separation of deterministic trends from random noise. The refined dataset was then processed with Gradient Boosting algorithms, specifically LightGBM and XGBoost. This approach was intended to enable the models to perform more effectively when predicting rainfall.

Empirical evaluation found that the SSA-XGBoost hybrid model achieved an RMSE of 0.0057 and an R² of 0.8278. It outperformed the SSA-LightGBM model, which achieved an R² of 0.2879 and struggled to capture non-linear patterns within the dataset. The reported predictive accuracy facilitates responsive nutrient management. Fertilizer application can be deferred during forecasted periods of high rainfall to prevent runoff and environmental pollution.

The research contributes to hydroinformatics by demonstrating the effectiveness of combining Singular Spectrum Analysis and XGBoost as a cost-efficient, high-performance solution for mitigating climate-related risks in tropical wetland agriculture. The activity connects rainfall variability, rainfall prediction, crop nutrient scheduling, and agricultural management through an integrated analytical framework.

Reference https://www.scopus.com/pages/publications/105042164780?origin=resultslist