Integration of Principal Component Analysis and Very Low Frequency Geophysical Method for Modeling Recharge Area Distribution The Case Study of Krawak Spring, Tuban

Recharge areas play an important role in the subsurface hydrological cycle, especially in karst regions such as Tuban Regency, which has unique hydrological and topographical conditions with steep morphology. This research was conducted around the Krawak spring, as this spring is very important for the needs of the surrounding community, so it must be maintained to ensure that the land function does not change, especially the recharge area, thus preserving the sustainability of the Krawak spring. Considering the difficult terrain for direct field surveys, this research utilizes a machine learning method based on Principal Component Analysis (PCA) with Very Low Frequency (VLF) data and Landsat-8 satellite imagery as supporting data. VLF data are used as training data for the machine learning model, whereas Landsat-8 imagery serves as the main data source, which is processed to produce five classification parameters, namely Normalized Difference Vegetation Index (NDVI), land cover, elevation, slope, and soil type. The results of this classification are then weighted and scored to produce a map of the potential recharge area distribution. The results of the PCA analysis show consistency with conventional scoring methods as well as the results from VLF data, making this method effective for mapping recharge areas in regions with challenging topography. This research provides an efficient and accurate alternative for modeling recharge areas in karst environments, where with direct field surveys are limited.