High-performance meteorological data processing framework for real-time analysis and visualization

The current level of geoscience advancements enables analysis of the Earth as a highly complex system with many components, like hydrosphere, lithosphere, and atmosphere, which interact within and with others. Large amounts of observational data and simulation data generated by numerical models need to be analyzed in order to extract valuable knowledge from them. Visualization is a very important part of data analysis since it provides an easy and fast way to assess the data and understand, which parts of it are of particular interest and whether there are any errors in the dataset. The most efficient type of tools for meteorological data visualization are geographic information systems (GIS) because they provide great functionality for manipulating geospatial data. In this paper, we describe the processing framework, which can be efficiently used as a backend for GIS by providing rapid access to the data located at the remote storage nodes. Experimental results demonstrate that developed framework allows real-time data access and can successfully handle a large number of simultaneous requests.