Hydrologic models applied to the prevention and warning of floods


Flood at Castelldefels. January, 2003


Bridge over the "Nacional II" road in Esparraguera. (June, 2000)


Raingage at Castelldefels

The new society vision based on learning to live with floods, goes a step above the simple solution of constructing infrastrucures to protect human lives and property, and it is within a more encompassing and multidisciplinary framework of risk management. Within this framework, it is of high relevance the forecasting of floods with information that allows to make correct decisions in advance. Flood forecasting at a point in a large watershed may be done reliably from river flow measurements closer to the headwaters than the point itself. This is not possible in medium size watersheds, where the transformation rainfall-runoff plays an important role and a hydrologic model, fed by rainfall and flow at different points within the watershed, needs to be used to predict floods at points of interest within the watershed. These rainfall-runoff models when integrated in a warning and forecasting system need specific hydrometeorological information in real time. This information comes from raingage networks or/and from weather radars, and flow measuring devices. Because of the widespread use of the land, flood forecasting cannot be limited to a few critical points (river flow through a city, a dam), in general, the demand of information in multiple sensitive points of the watershed will become more frequent.

The Mediterranean area meteorologic and geomorphologic characteristics and urbanization degree have particularites that make it difficult to forecast floods. The rainfall can be very variable in space and time and the watersheds have a torrential regime with short response times, which makes it necessary to use hydrological models that incorporate the spatial variability of the rainfall. Radar information, from this point of view is essential.

A wide variety of complex processes take place in a watershed. To understand and describe them is a complicated matter. The complexity of the equations that govern the hydrologic processes: surface runoff, infiltration, unsaturated flow, etc., the processes interactions and the spatial and/or temporal variability of the equations variables make it difficult to understand the system thoroughly even under the hypothesis (still debated) of an unequivocal relationship between the causes: rainfall, watershed characteristics, initial conditions, and the effect: watershed discharge.

The goal of a hydrologic model that predicts floods is to provide guidance for making decisions and giving warnings in flood sensitive areas and in infrastructures built to retain and laminate flow peaks (dams to prevent floods and for water management). This goal means that the model has to work with the watershed available data, in real time and that it has to be suitable to the needs of the entities that manage the flood warning system.

The GRAHI has a research line in hydrologic modeling in the Mediterranean area. It researches the most important processes that impact flow generation at a watershed scale. Its aim is to integrate these processes in tools for operational purposes.

We have developed a distributed parameter model. This model has been applied to several pilot watersheds in Catalonia. It predicts flow in different points of the watershed. One of the objectives of this model is to get the most profit from the radar information to improve the rainfall field data.

Would you like to know more about types of hydrologic models?

The previous text was extracted from Carles Corral's dissertation, titled: "Desenvolupament d'un model hidrològic per incorporar la informació del radar meteorològic. Aplicació operacional a la conca del riu Besòs" (Development of a hydrologic model that incorporates weather radar data. Application to the Besòs River watershed). Civil Engineering Dept. , Universitat Politècnica de Catalunya.