فهرست نشریات علمی دسترسی آزاد ایران

Recent advancements in remote sensing technology have optimized spatial data by enhancing the resolution of field measurement data. Regional geological maps are still used to validate geoscience models because of their high reliability, which is based on field measurements, but they have relatively low spatial resolution. Combining remote sensing models with machine learning algorithms offers a promising method to reconstruct regional geological maps into high-resolution geological maps, especially in volcanic regions with active geothermal systems such as the Patuha Geothermal Field in Indonesia. Models derived from pre-processed satellite gravity data and normalized satellite images, with standardized pixel raster sizes, form the quantitative basis for reconstructing regional geological map models. The Convolutional Neural Network (CNN) algorithm serves as the computational basis for reconstructing spatial models. The results of spatial reconstruction modeling using remote sensing data provide detailed insights into the distribution of geological features, achieving an accuracy rate of 81.26%. These varying geological feature zones are likely related to the dynamics of active volcanic regions. Since the active volcanic activity, geological fault structures have been formed and could be identified by combination of derivative analysis and remote sensing approach. Second Vertical Derivative (SVD) provides physical characteristics of active fault planes, integrating it with remote sensing analysis to indicate fault planes appeared in the surface. There is a clear correlation between the distribution of reconstructed lithological features and geological fault planes. Areas with a high concentration of fault planes often have a more diverse distribution of geological features, likely due to the influence of active faults, volcanic activity, and material erosion. Adding hyperparameters and geological feature constraints to the machine learning algorithm is a promising option for further research in this area of geological map reconstruction.