High Performance Computing for Hydrological Modelling
Abstract
A hydrological model represents a simplification of a variety of complex mechanisms occurring in nature at different scales. Some physical phenomena at certain scales are not considered in the set of parameters either because they are deemed unimportant or because they cannot be measured. Hidden variables can be represented by applying stochastic modeling which relates parameters to probability distributions. By this approach, methods of improving parameter sets have been developed. So, a possible way of determining the optimal combination of parameters is to simulate an important set of possible parameters. This requires a considerable number of simulations that exceeds the capabilities of traditional computation. For example, the systematic exploration of the objective function structure of the fourparameter model MEDOR( a daily RainfallRunoff Model), specific to the Mediterranean climate, requires 1,476,800 simulations which needs 29.9 hours using a single processor where as this computation only requires 58 minutes when 40 processors are used. The importance of modeling in hydrology is to develop aspects of hydrological models using High Performance Computing (HPC), to utilize the most efficient algorithms on the parallel target architecture in order to combine the algorithmic and the hardware speedup in the optimal way. High Performance Computing, called the "Grand Challenges", requires interdisciplinarity. The combination of MultiProcessor Vector Machines, Massively Parallel Processors, and Parallelism can be exploited by using networks of workstations creating an opportunity to couple singlemedia models and explore data assimilation issues at the same time. The question is how existing methods and models, specifically hydrological spatial simulations, could be adapted or modified to work in such an environment. A theoretical design is proposed to integrate simulation modules and existing visualisation tools in a heterogeneous parallel computing network to create a dynamic interactive system where users can display and manipulate remote data in a transparent fashion.
 Publication:

AGU Spring Meeting Abstracts
 Pub Date:
 May 2004
 Bibcode:
 2004AGUSM.H52A..04T
 Keywords:

 1800 HYDROLOGY;
 1848 Networks