Assimilating uncertain, dynamic and intermittent streamflow observations in hydrological models. (September 2015)

• Record Type: Journal Article
• Authors: Mazzoleni, Maurizio; Alfonso, Leonardo; Chacon-Hurtado, Juan; Solomatine, Dimitri
• Is Part Of: Advances in water resources. Volume 83(2015)
• Page Start: 323
• Page End: 339
• Publication Date: 2015-09
• Subjects: Data assimilation -- Uncertain observation -- Dynamic observations -- Intermittent observations -- Hydrological modeling; Hydrology -- Periodicals; Hydrodynamics -- Periodicals; Hydraulic engineering -- Periodicals; 551.48
• DOI: 10.1016/j.advwatres.2015.07.004
• Languages: English
• ISSNs: 0309-1708

Abstract

Highlights: We compare model performance obtained assimilating streamflow observations from static sensors and dynamic sensors. Different structures of semi-distributed hydrological models affect in different ways the assimilation performances. Analysis of the model performance assimilating intermittent observations with variable uncertainty in space and time. Evaluation of the model improvement assimilating static, dynamic and intermittent uncertain observations of discharge in realistic scenarios of sensors location. Abstract: Catastrophic floods cause significant socio-economical losses. Non-structural measures, such as real-time flood forecasting, can potentially reduce flood risk. To this end, data assimilation methods have been used to improve flood forecasts by integrating static ground observations, and in some cases also remote sensing observations, within water models. Current hydrologic and hydraulic research works consider assimilation of observations coming from traditional, static sensors. At the same time, low-cost, mobile sensors and mobile communication devices are becoming also increasingly available. The main goal and innovation of this study is to demonstrate the usefulness of assimilating uncertain streamflow observations that are dynamic in space and intermittent in time in the context of two different semi-distributed hydrological model structures. The developed method is applied to the Brue basin, where the dynamic observations are imitated by the synthetic observations of discharge. The results of this study show how model structures and sensors locations affect in different ways the assimilation of streamflow observations. In addition, it proves how assimilation of such uncertain observations from dynamic sensors can provide model improvements similar to those of streamflow observations coming from a non-optimal network of static physical sensors. This can be a potential application of recent efforts to build citizen observatories of water, which can make the citizens an active part in information capturing, evaluation and communication, helping simultaneously to improvement of model-based flood forecasting.