Investigating the role of saliency analysis with a neural network rainfall-runoff model

Software tools are available which translate neural network solutions into standard computer languages and source code. This conversion process enable s trained networks to be implemented as embedded functions within existing hydrological models or assembled into stand-alone computer programs. In add ition to this primary use, embedded functions can also provide new opportun ities for dynamic testing and for the internal investigation of the model's function. Saliency analysis, the disaggregation of a neural network soluti on in terms of its forecasting inputs, is one approach which is explored he re. Saliency analysis is used to investigate the performance of a neural ne twork one-step-ahead hydrological forecasting model using different combina tions of input data for testing and validation. (C) 2001 Elsevier Science L td. All rights reserved.