Meteorological multi-source fusion lattice point data terrain correction method and system based on deep learning

The invention provides a meteorological multi-source fusion lattice point data terrain correction method and system based on deep learning. The method comprises the following steps: constructing a spatio-temporal data set; wherein the spatio-temporal data set comprises a site data set, a topographic data set and a live lattice point data set; inputting the spatio-temporal data set into a deep neural network for training to obtain a correction model; wherein the deep neural network model adopts a multi-layer full-connection neural network model; and obtaining a precipitation correction value based on the correction model. According to the invention, the technology of carrying out terrain correction on the meteorological multi-source fusion grid point data of Guangdong province by using the full-connection neural network can provide powerful support for improving the technical level of various services such as weather forecast, meteorological service, meteorological disaster risk assessment, management and earl