HPC Environment on Azure Cloud for Hydrological Parameter Estimation

Abstract

High performance of data-intensive computation is required to deal with the complexity of analysis and simulation for hydrological modeling jobs like parameter estimation. The vigorously developing cloud computing has emerged as a promising platform for HPC (High Performance Computing) of science community. This paper presents our work in developing and implementing HPC environment on Azure cloud for applications of hydrological parameter estimation. According to the requirements of hydrological modeling, we design and construct a HPC environment on Azure cloud. After deploying parameter estimation applications on the HPC environment, a case study on groundwater uncertainty analysis in Heihe River Basin using the HPC environment is presented. Our work demonstrates that Azure cloud can advantageously complement traditional high performance computing infrastructure and help hydrological researchers improve model computing efficiency by handy process steps.