Integration of high performance computing and management of large volumes of data to the analysis and prediction of weather in the Ecuadorian Austro

Climate change can have a severe negative impact on different types of environmental services, particularly in the provision of water and hydrological regulation. Climate analysis and prediction is a tool for crucial importance in long -term decision making in the management of environmental services. This analysis and climate prediction is carried out with the help of software packages, which apply a mathematical model to a set of inputs. Weather Research & Forecasting (WRF) is one of the most popular software packages used for the analysis and/or climate prediction. As the granularity and/or the area on which the analysis and prediction of climate increases, computing requirements and input data size increase ostensibly, making the treatment of such problems not possible using hardware and common algorithms. High performance and algorithms computing for the management of large volumes of data have allowed the treatment of problems that were limited by processing capacity and data size. This research project aims to integrate high performance computing techniques and management of large volumes of data to the problem of analysis and climate prediction, in the area of the Ecuadorian Andes. Specifically, the objective of this research project is to analyze the algorithms implemented in WRF, as well as the required input data, in such a way that they can be optimized for high performance computing platforms. This analysis will be iterative and will be carried out on different high performance platforms, in order to eliminate bottlenecks in performance and obtain optimal performance. The platforms to consider will be computers -based clusters, based on blades and even based on hardware accelerators.

XII UNIVERSITY COMPETITION FOR RESEARCH PROJECTS UNIVERSITY OF CUENCA