Thermal and vegetation indices of geothermal steam spots derived by high resolution images and field verifications

Abstract

Retrieving land surface temperature originating from subsurface thermal data using satellite images has some challenges, especially in tropical areas. The vegetation, cloud cover, and thick soil layers affect the detected ground temperatures. The low-to-medium spatial resolution of thermal infrared images leads to low accuracy compared with ground measurements. Therefore, proper image correction, calibration, and spatial resolution are required for comparison with kinetic temperature measured from the ground. The objective of this study is to detect thermal and vegetation anomalies related to steam spots in subsurface geothermal systems using multivariable thermal infrared corrections and the red band angle of a high spatial resolution optical image, respectively. In this study, the Kamojang–Guntur Volcanic Complex, West Java, Indonesia was selected as the study area. The exploitation and exploration of steam fields were used to assess the accuracy of the proposed method. Thermal infrared images were obtained using an advanced spaceborne thermal emission and reflection radiometer (ASTER). Principal and multivariable corrections were applied to obtain the surface temperature related to steam spots using ground thermal measurements and land cover classification to recognize surface emissivity originating from vegetation, urban areas, bare land, and water bodies. To improve multivariable analyses and interpretations, we used the high spatial resolution image of PlanetScope to obtain vegetation indices from steam spots. The gradient redness index was calculated from the atmospherically corrected PlanetScope image and used as an indicator of ground steam spot signatures. A field measurement campaign was performed to verify and analyze the thermal and vegetation indices at the ground level. Accordingly, we found that the high anomalies of the corrected surface temperature and physiological leaves were concordant with the opened and closed steam spots in the Kamojang–Guntur Volcanic Complex. Thermal and vegetation indices have the potential to estimate hidden geothermal systems and can be used in other similar areas.