COMPUTATIONAL ENVIRONMENT FOR THE OPERATIONAL RAINFALL-RUNOFF MODEL

The overall objective of the ongoing work is to develop an effective tool for forecasting runoff from the catchment area for various hydro-meteorological conditions while taking into account the information available in the operating mode. Striving for operational reliability of the model and reducing the uncertainty of the generated forecasts is accomplished through the adjustment of both the internal structure of the model and the spatial representation of the computational grid to the physiographical, hydrological and climatological characteristics of a given basin. The research focused on the development of methods of selecting the optimal model structure and parameters by analysing the results obtained for different model structures. The is achieved through the computational environment in which it will be possible to implement different types of hydrological rainfall-runoff models developed both in Poland and in the world. These models have a unified system of data input, parameter optimization rules and procedures for results generation. Developed elements of computational environment, correspond to the potential of generation of models with a given structure and complexity. The formal description of the model structure is presented in the form of XML file. Based on this file the computational model is constructed by the parsing engine using the prepared programming library. Furthermore, within the framework of computational environment the following components were developed: application programming interface (API), data assimilation module, module for computational representation of a real object, module for the estimation and optimization of model parameters. The developed computational environment was applied to prepare a test version of TOPMODEL and perform hydrological validation of the model results. The hydrological validation was performed for the selected flood events in the subbasins of Nysa Klodzka river and Sola river. It is planned to further develop the computational environment and include the additional features i.e. other methods for parameter optimization, tools for building quasi-distributed models and to construct a global model that contains models of different structures and complexity.