Modelos predictivos del comportamiento del nivel piezométrico de la laguna Charco del Toro (Parque Nacional de Doñana, Huelva, sw España) mediante técnicas de análisis multivariante

The performance of the intra-annual piezometric level of the Charco del Toro small-lake (Doñana National Park, Huelva, Spain) is modeled using multivariate analysis techniques. There are two main purposes on this paper: interpreting the current performance of the piezometric level of the wetland from different environmental variables and predicting climate change in the small-lake by the end of the 21st century. Thus, an explanatory and a predictive model were developed. Although the methodological approach for both types of models is similar, based on multivariate linear regression analysis, the models developed for each of the purposes were different, due to the greater availability of data for environmental variables in the recent past than for making predictions based on climate change scenarios. The results from the explanatory model of the current performance of the small-lake indicate that the piezometric level depends on the accumulated precipitation over long periods of time (from 6 to 10 months), as well as the potential evapotranspiration recorded in the previous 10 months. The models developed for predicting the future intra-annual performance, based on different scenarios and climatic models, suggest that the piezometric level of the lagoon will fall drastically 1-1.5 meters regardless the model of climate change or the emssions scenario.