Interpolation of Geochemical Data with Aster Images Based on AlexNet Convolution Neural Network

Being an important geological information source, geochemical data is widely used in mineral exploration, environmental protection, pollution monitoring, etc. However, geochemical data with extensive coverage and fine resolution has become inaccessible, especially in some unreachable and remote areas. Remote sensing data with fast and efficient ability to collect geology related geoinformation has long been employed in many of geological studies. Joint utilization of geochemical and remote sensing data, as well as other sources of geo-data can consequently assist in geological applications such as mineral exploration In recent decades, methodology to integrate remote sensing and geochemical data have significantly improved. During the integration, geochemical data are often interpolated or resampled to finer resolution for match that of remote sensing images but without notable improvement in geo-information quality containedwith. This study proposeda new integration method that uses the AlexNet convolution neural network to interpolate geochemical data with ASTER images. The interpolated geochemical data presents not only with a higher spatial resolution, but also with geological information from remote sensing images.