HydroMP – a computing platform for hydrodynamic simulation based on cloud computing

Abstract

Abstract Modern water management decisions are increasingly dependent on efficient numerical simulations of multiple scenarios with multi-models. In this paper, a service mode for the hydrodynamic simulation based on cloud computing is proposed, and the relevant frameworks of the Hydrologic/Hydraulic Modeling Platform (HydroMP) are designed and implemented. Various hydro-models can be integrated into HydroMP dynamically without the need of program recompiling, since it achieves the scheduling of computing resources to provide end users with the rapid computing capacity of concurrent scenario simulations in the form of a Web service. The present study focuses on the dynamic model integration, resource scheduling, system communication and data structure design. To use the present one-dimensional hydrodynamic cloud computing as a prototype, two integration methods (including the EXE integration and PIIM integration) are applied to construct the CE-QUAL-RIV1 and JPWSPC (Joint Point Water Stage Prediction and Correction) models, thereby to investigate real-time scheduling of the water transfer channels in the South-to-North Water Diversion (SNWD) project. The results showed that massive modeling scenarios by use of different hydrodynamic models, if submitted concurrently, can be processed simultaneously in the HydroMP. The data structure of the proposed framework can also be extended to two-dimensional and three-dimensional hydrodynamic situations.