Assessment of nitrogen leaching from arable land in large river basins: Part II: regionalisation using fuzzy rule based modelling |
| |
Authors: | Uwe Haberlandt, Valentina Krysanova,Andr s B rdossy |
| |
Affiliation: | Uwe Haberlandt, Valentina Krysanova,András Bárdossy |
| |
Abstract: | This study deals with fuzzy rule based modelling of nitrogen (N)-leaching from arable land. Main purpose is the elaboration of a method, which allows dynamical regionalisation of results from process-based models for large regions and can be efficiently included in metamodels or decision support systems for rapid integrated assessment of water resources. The paper is the second part of a two-part paper. In the first paper the distributed ecohydrological model SWIM had been applied to calculate and analyse nitrogen dynamics in arable soils for a set of representative natural and management conditions in the Saale River basin (Ecol. Model. (in press)). Here, in the second paper the results from those simulation experiments are used to define, train and validate fuzzy rule systems for the estimation of N-leaching. Nine fuzzy rule systems, specific for nine soil classes, were created from the simulation experiments, representing the conditions for the whole Saale River basin. The fuzzy rule systems operate on monthly time steps and consist of 15 rules and seven input variables each, which are compiled from time series of precipitation, percolation and evapotranspiration as well as from information about fertilizer and crop specific nitrogen uptake. Simulated annealing as a non-linear discrete optimisation method is used for automatic rule assessment. Validation of the fuzzy rule systems, carried out by split sampling of 30-year simulation period, shows satisfactory performance on an annual basis and good performance on the long-term basis with average correlation between SWIM-simulated and fuzzy rule-estimated N-leaching values of 0.78 and 0.94, respectively. |
| |
Keywords: | Nitrogen leaching Dynamic modelling Large scale Regionalisation Fuzzy rules Simulated annealing |
本文献已被 ScienceDirect 等数据库收录! |
|