Keywords
rainfall-runoff models, uncertainty, performance metrics, information
Start Date
1-7-2008 12:00 AM
Abstract
The inclusion of additional information in a model should improve the model’s performance and reduce associated uncertainties. The additional information may take the form of higher temporal and/or spatial resolution of data already used, or additional types of data (including soft data). However, additional data is not necessarily equivalent with additional information as a model may only be able to make use of a fraction of the information in the data, or the approach chosen to extract data might be inefficient. Consideration of the information-to-noise ratio of the data is needed in order to evaluate whether the information will indeed improve the model's performance or reduce uncertainty. Evaluating the information-to-noise ratio and the information content of data in general are often difficult tasks. One simple but effective approach to this problem is to compare the performance and uncertainty of a model with and without additional information. This workshop will focus on the theory and use of additional information in improving the performance of and reducing the uncertainty in hydrological models. This includes consideration of the uncertainty in datasets as well as techniques to evaluate the effectiveness of the additional information in improving models.
Evaluating the information content of data for uncertainty reduction in hydrological modelling
The inclusion of additional information in a model should improve the model’s performance and reduce associated uncertainties. The additional information may take the form of higher temporal and/or spatial resolution of data already used, or additional types of data (including soft data). However, additional data is not necessarily equivalent with additional information as a model may only be able to make use of a fraction of the information in the data, or the approach chosen to extract data might be inefficient. Consideration of the information-to-noise ratio of the data is needed in order to evaluate whether the information will indeed improve the model's performance or reduce uncertainty. Evaluating the information-to-noise ratio and the information content of data in general are often difficult tasks. One simple but effective approach to this problem is to compare the performance and uncertainty of a model with and without additional information. This workshop will focus on the theory and use of additional information in improving the performance of and reducing the uncertainty in hydrological models. This includes consideration of the uncertainty in datasets as well as techniques to evaluate the effectiveness of the additional information in improving models.
