A New Strategy to Fuse Remote Sensing Data and Geochemical Data with Different Machine Learning Methods

Geochemical data can reflect geological features, making it one of the basic types geodata that have been widely used in mineral exploration, environmental assessment, resource potential analysis and other research. However, final decisions regarding activities are often limited by spatial accuracy geochemical data. sampling is sometimes difficult to conduct because harsh natural geographic conditions (e.g., mountainous areas with high altitude complex terrain), meaning only medium/low-precision survey could be obtained, which may not adequate for regional mapping exploration. Modern techniques such as remote sensing address this issue. In recent decades, development technology has provided a huge amount earth observation spatial, temporal spectral resolutions. The advantage rapid acquisition information large promoted broad use geoscientific Remote help differentiate various ground features recording electromagnetic response surface solar radiation. Many problems occur during process fusing reported, feasibility existing fusion methods low accuracies less useful practice. paper, new strategy integrating (referred ASTER data) proposed; achieved through linear regression well random forest support vector algorithms. results show obtain better available sets prove currently proposed effectively high-spatial-resolution (15 m) low-spatial-resolution (2000 wide-range accurate applications lithological identification exploration).