Detecting Lithium Brine in Afghanistan Using Spaceborne Hyperspectral Data

Sammanfattning: The purpose of this thesis is to utilize hyperspectral remote sensing technology for the identification and evaluation of lithium mineral deposits in Afghanistan. Afghanistan boasts unique geological formations with substantial potential for lithium resources beneath its land crust. The focal point of this project revolves around exploring and developing lithium brine-rich deposits in the Dashti-e- Nawar field, an exceptional brine accumulation that also harbors essential boron (B) deposits, crucial for decarbonization efforts. These efforts aim to expand the understanding of lithium mineralization beyond a specific location and explore the possibility of transferring spectral signatures to identify lithium deposits in areas beyond Dashti-e-Nawar. To identify the potential of lithium deposits in Dashti-e-Nawar, Afghanistan, advanced feature extraction algorithms such as Principal Component Analysis (PCA), supervised and unsupervised classification techniques, spectral matching, and spectral profile analysis were applied to PRISMA hyperspectral spaceborne data. Through this analysis, various types of lithium minerals including Gypsum, Hectorite, Kaolinite, and Lepidolite were identified across the dry lake of Dashti-e-Nawar. Furthermore, the analysis revealed the presence of Boron and Borax deposits in the study area (Dashti-e-Nawar) as well as in the mountainous regions on both the eastern and western sides, indicating significant potential for Boron and Borax deposits in Dashti-e-Nawar, Ghazni, Afghanistan. The spectral profiles of lithium minerals in Dashti-e-Nawar were analyzed and compared to reference spectra of lithium minerals from the United States Geological Survey (USGS) spectral library and the JPL spectral library using ERDAS Imagine software. The results demonstrate that Dashti-e-Nawar spans an approximate area of 361.6 km² of lithium mineral resources over the dry lake. To identify the lithium (Li) mineral spectral signature, PRISMA data was used to compare with the reference spectra from Bolivia, a known lithium area. The analysis of lithium (Li) spectral signatures revealed a distinctive absorption pattern, with a narrow absorption feature in the near-infrared (NIR) region and a broad absorption feature in the short-wave infrared (SWIR). The most useful bands for identifying the lithium spectral signature using PRISMA hyperspectral data were found at 1372.72 nm (NIR) and 1904.06 nm (SWIR). The natural environmental conditions in Afghanistan facilitate relatively easy in-field exploration in certain areas, and hyperspectral imaging (HSI) proves helpful for regional geological mapping and mineral identification. With this advantageous combination of lithium sources, Afghanistan is well-positioned to emerge as a major global producer of lithium in the future.