Uncertainty Modelling in Rainfall-Runoff Simulations Based on Parallel Monte Carlo Method

Abstract

This article describes statistical evaluation of the computational model for precipitation forecast and proposes a method for uncertainty modelling of rainfall-runoff models in the Floreon + system based on this evaluation. The Monte-Carlo simulation method is used for estimating possible river discharge and provides several confidence intervals that can support the decisions in operational disaster management. Experiments with other parameters of the model and their influence on final river discharge are also discussed.

Harnessing Performance Variability: A HPC-Oriented Application Scenario

“…In this paper we exploit a rainfall-runoff (RR) model, developed as part of the Floreon+ system [6]. The model is an example of multi-threaded scientific application, running on HPC system, made by twelve 6-core AMD Opteron 8425HE processors, for a total amount of 48 CPU cores.…”

“…The experiments simulates the Moravian-Silesian region catchments [6]. The simulation is based on four instances of an Uncertainty model, concurrently executed.…”

Flood Prediction Model Simulation With Heterogeneous Trade-Offs In High Performance Computing Framework

“…This method runs the uncertainty model with the parameters set in the onSetup and onConfigure methods (for more information about the experimental uncertainty model see (Golasowski et al 2015)). Aside from the standard uncertainty percentiles computed by the model, the method also computes the precision of the current execution and saves it for monitoring and control purposes in the onMonitor method.…”

“…A model used in this paper is a rainfall-runoff model developed as part of the Floreon + system (Golasowski, Litschmannova, Kuchar, Podhoranyi & Martinovic 2015), (Vavrik, Theuer, Golasowski, Kuchar, Podhoranyi & Vondrak 2015). The RR models transforms precipitation to water discharge levels by modelling individual parts of the rainfall-runoff process.…”

“…To validate the use of the proposed solution, we have deployed a hybrid OpenMP (Board 2013) and MPI (Forum 2012) implementation of the uncertainty modelling (Golasowski et al 2015) to the Anselm supercomputer operated by IT4Innovations Czech National Supercomputing Center (Sliva & Stanek 2013). The cluster contains 209 nodes each with 2 Intel(R) Sandy Bridge E5-2665 @ 2.4GHz CPU sockets with 8 cores and 20480 KB L2 cache.…”