Indonesia is rich in plant biodiversity, with various species used for medicinal purposes, spices, and other applications. Curcuma xanthorrhiza (temulawak) is one of nine priority medicinal plants designated by the Indonesian Food and Drug Authority (BPOM) since 2004. The successful cultivation of temulawak is strongly influenced by several factors, particularly soil quality, including key chemical properties such as pH, soil organic carbon, total nitrogen, total phosphorus (P), and total potassium (K), which serve as essential nutrients for plant growth. Soil nutrient testing is commonly performed in laboratories; however, conventional laboratory analysis can be time-consuming and relatively expensive. An effective alternative for rapidly and non-destructively assessing soil chemical properties is the use of Visible Near-Infrared Spectroscopy (Vis-NIRS). Vis-NIRS is a fast and promising technique for detecting soil characteristics without the need for complex and costly laboratory extraction procedures. This study aims to evaluate the predictive accuracy of Vis-NIRS in estimating key soil chemical parameter namely pH, total phosphorus (P), and total potassium (K), under two different sample conditions: fresh soil and air-dried soil. The research was conducted from February to June 2025. A total of 70 soil samples were collected from various regions in Central Java using a purposive sampling method. Reflectance spectra were recorded across a wavelength range of 350–2500 nm and analyzed using regression-based statistical approaches. The results indicate that Vis-NIRS can predict soil pH and total phosphorus with relatively high accuracy, particularly in air-dried samples (R² > 0.75). However, predictions for total potassium showed low accuracy under both conditions (R² < 0.2). These findings suggest that Vis-NIRS is more suitable for estimating soil pH and phosphorus, while potassium prediction requires further calibration model development. This research contributes to the advancement of rapid and efficient soil characterization methods.
