Development of a GPU Accelerated Salinity Module for the SRH-2D Platform
Can a more realistic representation of river or delta salinity via the development of a salinity module for the SRH-2D platform provide improved decision-support for addressing salinity issues? Can a parallelized GPU engine using the CUDA framework be used to accelerate a numerically-based finite-volume modelling platform reduce runtimes into a realistic time frame for decision-makers? Currently, most physically based models for salinity are difficult to use and have very long runtimes. Our goal is to an accessible and user friendly salinity module for the SRH-2D platform that overcomes the issues of accessibility and understanding, while significantly reducing runtimes and feedback via GPU acceleration from the CUDA framework.
Need and Benefit
The proposed research addresses issues regarding Water Quality, River Habitat Restoration, Water Operations Models and Decision Support Systems, and Climate Variability. More specifically, salinity is a major water quality issue in most agricultural areas. Reclamation's mission is designed around delivering usable agricultural water for the West. Therefore, being able to simulate salinity in a 2-D finite-volume model in a realistic timeframe would be of significant value to Reclamation. More-so by designing the model from the beginning with user accessibility in mind will allow for greater adoption and understanding of the modelling platform. In terms of urgency, the delta salinity issues and river salinity issues will only increase as Reclamation's operations are affected more and more by climate variability. Therefore, now is the time for Reclamation to invest in salinity module for the SRH-2D platform and to decrease the runtimes of these physically based models through GPU acceleration so that key decision-makers can make informed decisions from the best decision support systems available.
Contact the Principal Investigator for information about partners.
Contact the Principal Investigator for information about these documents.