Characterization of the Jiu Valley Mining Basin, Romania, using Remote Sensing Techniques in Relation to Land Use Categories and Relief Parameters

Abstract

The study aimed to analyze and characterize the mining basin of the Jiu Valley, Romania, through remote sensing techniques. The land use categories and morphological parameters of the area, represented by DEM and SLOPE, were considered. Twenty-one land use categories were identified, with areas ranging from 27.83 ha (Water bodies, 512 Code) to 33543 ha (Broad-leaved forest, 311 Code). Five land use categories were included in the lower quartile (63.055 quartile threshold), with a total area of ​​204.71 ha. Eleven land use categories were included in the middle quartile (63.055 – 4650.855, interquartile range), with a total area of ​​18651.42 ha. In the upper quartile (4650.855, quartile threshold) five categories were included, with a total area of ​​76501.70 ha. In the agricultural domain, five categories (211, 231, 242, 243, and 321 Codes) were found with a total area of ​​23093.18 ha, which represents 24.22% of the total area of ​​the study area. In the case of DEM parameters, 21 classes (DC1 to DC21) were generated, with values ​​ranging from 435.59 ha (DC21) to 7080.98 ha (DC5). The decreasing trend of DEM values ​​was confirmed (Mann-Kendall test), and the linear model described the decreasing trend of DEM values, with slope = -326.8, in conditions of statistical safety (p<0.001). Within the SLOPE parameter, 21 classes (SC1 to SC21) were generated with values ​​ranging from 17.22 ha (SC21) to 9960.56 ha (SC8). The Mann-Kendall test confirmed the decreasing trend of the data series, and a linear model described the data layout, with slope = -405.96, in statistical safety conditions (p<0.001). Cluster analysis grouped the DEM classes (Coph.corr. = 0.833), and the SLOPE classes (Coph.corr. = 0.804) based on similarity into clusters with several subclusters.