Estimasi Produktivitas Tebu di Wilayah Perkebunan Glenmore Menggunakan Indeks Vegetasi Berdasarkan Analisis Truncated Spline

Abstract

Accurate estimation of sugarcane productivity are essential for planning and managing national sugar production. However, the methods commonly used in the field to determine sugarcane productivity still rely on conventional measurements that require significant time, cost, and labor, making them inefficient for monitoring large-scale plantations. Advances in remote sensing technology provide opportunities to continuously monitor crop conditions through the use of satellite imagery. The Normalized Difference Vegetation Index (NDVI) is one of the parameters often used to describe vegetation conditions, although the relationship between NDVI and crop productivity is not always linear. This study aims to estimate sugarcane productivity based on maximum NDVI values using a truncated spline regression approach. The study analysis was conducted to the data collected from the Glenmore sugarcane plantation area, Banyuwangi Regency, using eight plots of land planted with the PS862 sugarcane variety. NDVI data were obtained from multitemporal Sentinel-2 satellite imageries processed using Google Earth Engine. To improve data quality, the NDVI time series data was smoothed using the Savitzky–Golay method to reduce the effects of cloud cover and temporal noise. The maximum NDVI values from each plot were then used as predictor variables in sugarcane productivity modeling. The results show a non-linear relationship between maximum NDVI and sugarcane productivity. The truncated spline regression model produced a coefficient of determination (R²) value of 0.664, indicating that most of the variation in sugarcane productivity can be explained by variations in maximum NDVI values. Based on these results, it can be concluded that the truncated spline regression approach has the potential to be used as an alternative method in estimating sugarcane productivity based on Sentinel-2 satellite imagery, with specific application to the varieties and conditions of the study area.