RAND and Lawrence Livermore National Lab Combine High-Performance Computing and Public Policy Analysis for Demonstration of Water Resource Management
Aug 25, 2016
In November 2014, experts from the RAND Corporation and Lawrence Livermore National Laboratory conducted a one-day workshop to explore the impact of using high-performance computing to support stakeholder real-time multiscenario deliberations over natural resource management planning. The workshop, employing Robust Decision Making (RDM) methods in support of a process of deliberation with analysis, revisited work RAND conducted for the Colorado River Basin in 2012 and was attended by decisionmakers involved in the original study and others interested in advancing the state of the art in quantitative decision support. In the course of the workshop, participants developed five new project portfolios that were then evaluated over about 12,000 alternative futures using high-performance computers — a process that would have taken weeks to complete using traditional computer clusters. This document summarizes workshop results and the observations attendees made about the benefits and challenges associated with using high-performance computing in this context.