ABSTRACT
The current study integrates remote sensing, machine learning, and physicochemical parameters to detect hydrodynamic conditions and groundwater quality deterioration in non-rechargeable aquifer systems. Fifty-two water samples were collected from all water resources in Siwa Oasis and analyzed for physical (pH, T°C, EC, and TDS) chemical (SO4 2-, HCO3 -, NO3 -, Cl-, CO3 2-, SiO2, Mg2+, Na+, Ca2+, and K+), and trace metals (AL, Fe, Sr, Ba, B, and Mn). A digital elevation model supported by machine learning was used to predict the change in the land cover (surface lake area, soil salinity, and water logging) and its effect on water quality deterioration. The groundwater circulation and interaction between the deep aquifer (NSSA) and shallow aquifer (TCA) were detected from the pressure-depth profile of 27 production wells penetrating NSSA. The chemical facies evolution in the aquifer systems were (Ca-Mg-HCO3) in the first stage (freshwater of NSSA) and changed to (Na-Cl) type in the last stage (brackish water of TCA and springs). Support vector machine successfully predicted the rapid increase of the hypersaline lake area from 22.6 km2 to 60.6 km2 within 30 years, which deteriorated a large part of the cultivated land, reflecting the environmental risk of over-extraction of water for irrigation of agricultural land by flooding technique and lack of suitable drainage network. The waterlogging in the study was due to a reduction in the infiltration rate (low permeability) of the soil and quaternary aquifer. The cause of this issue could be a complete saturation of agricultural water with chrysotile, calcite, talc, dolomite, gibbsite, chlorite, Ca-montmorillonite, illite, hematite, kaolinite and K-mica (saturation index >1), giving the chance of these minerals to precipitate in the pore spaces of the soil and decrease the infiltration rate. The NSSA is appropriate for irrigation, whereas TCA is inappropriate due to potential salinity and magnesium risks. The best way to manage water resources in Siwa Oasis could be to use underground drip irrigation and combine water with TCA and NSSA.
ABSTRACT
Integrating various tools in targeting mineral deposits increases the chance of adequate detection and characterization of mineralization zones. Selecting a convenient dataset is a key for a precise geological and hydrothermal alteration mapping. Remote sensing and airborne geophysical data have proven their efficiency as tools for reliable mineral exploration. Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER), Advanced land imager (ALI), Landsat 8 (L8), and Sentinel 2 data are widely-used data among various types of remote sensing images in resolving lithological and hydrothermal alteration mapping over the last two decades. ASTER is a well-established satellite in geological remote sensing with detailed Short-wave infrared (SWIR) range compared to visible and near-infrared region (VNIR) that controls iron-associated alteration detection. On contrary, ALI has excellent coverage of the VNIR area (6 bands), but does not possess the potentiality of ASTER for the SWIR and thermal regions. Landsat 8 is widely used and highly recommended for lithological and hydrothermal alteration mapping. The higher spatial (up to 10 m) resolution of Sentinel 2 MSI has preserved its role in producing accurate geological mapping. Notwithstanding the foregoing, implementing the four datasets in a single study is time-consuming. Thus, an important question when commencing an exploration project for hydrothermal alterations-related mineralization (orogenic mineral deposits in the current research) is: which dataset should be adopted to fulfill proper and adequate outputs? Here the four widely recommended datasets (ASTER, ALI, L8, and sentinel 2) have been tested by applying the widely-accepted techniques (false color combinations, band ratios, directed principal component analysis, and constrained energy minimization) for geological and hydrothermal alteration mapping of Gabal El Rukham-Gabal Mueilha district, Egypt. The study area is covered mainly by Neoproterozoic heterogeneous collection of ophiolitic components, island arc assemblage, intruded by enormous granitic rocks. Additionally, airborne magnetic and radiometric data were applied and compared with the remote sensing investigations for deciphering the structural and hydrothermal alteration patterns within the study area. The results demonstrated a different extent from one sensor to another, highlighting their varied efficacy in detecting hydrothermal alterations (mainly hydroxyl-bearing alterations and iron oxides). Moreover, the analysis of airborne magnetic and radiometric data showed hydrothermal alteration zones that are consistent with the detected alteration pattern. The coincidence between high magnetic anomalies, high values of the K/eTh ratio, and the resultant alterations confirm the real alteration anomalies. Over and above that, the remote sensing results and airborne geophysical indications were verified with fieldwork and petrographic investigations, and strongly recommend combining ASTER and Sentinel 2 results in further investigations. Based on the outputs of the current research, we expect better hydrothermal alteration delineation by adopting the current findings as they sharply narrow the zones to be further investigated via costly geophysical and geochemical methods in mineral exploration projects.
