The hydrologic response of aquifers to meteorological conditions is often studied by means of time series analysis or physically based models. Modelling hydro-geological systems is however a challenging problem, often complicated by poor knowledge of basic assumptions such as hydraulic conductivity distribution, aquifer geometry and evapotraspiration rates. A better estimate of model parameters usually requires quite substantial investment and prolonged measurement campaigns. Collecting reliable data from different sources and public corporations is not always an easy task. Conversely, piezometric head and rainfall intensity data in long records are more easily available. Therefore, data-driven approaches (Ljung, 1999; Giustolisi, 2000) in groundwater hydrology and hydrogeology are attractive and potentially complementary in some cases. In the present work, the authors show some interesting results regarding the application of a recent symbolic regression technique, named evolutionary polynomial regression (Giustolisi and Savic, 2003, Mancarella, 2003, Giustolisi and Savic, 2006), in the conceptual modelling of a real hydrogeological system. The methodology’s advantage lies in the model building procedure that is entirely based on time series data, and on the possibility of conceptualizing the physical insight into the process.

La risposta idrologica degli acquiferi alle precipitazioni può essere modellata con diversi approcci. E’ frequente l’utilizzo di modelli fisicamente basati o il ricorso allo studio delle serie storiche con metodi statistici. L’utilizzo realistico di modelli fisicamente basati presuppone una buona conoscenza di base del sistema e un patrimonio di dati rilevante spesso di non facile reperibilità. La disponibilità di serie storiche idrologiche in periodi anche lunghi è invece più frequente. Pertanto gli approcci datadriven possono talvolta rappresentare una utile alternativa o uno strumento complementare importante nell’idrogeologia delle acque sotterranee. Nel presente lavoro gli autori si propongono di illustrare i risultati dell’applicazione di una recente tecnica di regressione simbolica denominata Evolutionary Polynomial Regression nella modellazione di un sistema idrogeologico reale, di cui sono disponibili le serie storiche pluviometriche e freatimetriche.

Modellazione e previsione nei sistemi idrogeologici mediante la tecnica E.P.R.. (Evolutionary Polynomial Regression) / Mancarella, D; Simeone, Vincenzo. - In: GIORNALE DI GEOLOGIA APPLICATA. - ISSN 1826-1256. - 8:1(2008), pp. 5-16. [10.1474/GGA.2008-08.1-01.0216]

Modellazione e previsione nei sistemi idrogeologici mediante la tecnica E.P.R.. (Evolutionary Polynomial Regression)

SIMEONE, Vincenzo
2008-01-01

Abstract

The hydrologic response of aquifers to meteorological conditions is often studied by means of time series analysis or physically based models. Modelling hydro-geological systems is however a challenging problem, often complicated by poor knowledge of basic assumptions such as hydraulic conductivity distribution, aquifer geometry and evapotraspiration rates. A better estimate of model parameters usually requires quite substantial investment and prolonged measurement campaigns. Collecting reliable data from different sources and public corporations is not always an easy task. Conversely, piezometric head and rainfall intensity data in long records are more easily available. Therefore, data-driven approaches (Ljung, 1999; Giustolisi, 2000) in groundwater hydrology and hydrogeology are attractive and potentially complementary in some cases. In the present work, the authors show some interesting results regarding the application of a recent symbolic regression technique, named evolutionary polynomial regression (Giustolisi and Savic, 2003, Mancarella, 2003, Giustolisi and Savic, 2006), in the conceptual modelling of a real hydrogeological system. The methodology’s advantage lies in the model building procedure that is entirely based on time series data, and on the possibility of conceptualizing the physical insight into the process.
2008
Modellazione e previsione nei sistemi idrogeologici mediante la tecnica E.P.R.. (Evolutionary Polynomial Regression) / Mancarella, D; Simeone, Vincenzo. - In: GIORNALE DI GEOLOGIA APPLICATA. - ISSN 1826-1256. - 8:1(2008), pp. 5-16. [10.1474/GGA.2008-08.1-01.0216]
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11589/11356
Citazioni
  • Scopus ND
  • ???jsp.display-item.citation.isi??? ND
social impact