Institutional Repository [SANDBOX]
Technical University of Crete
EN  |  EL

Search

Browse

My Space

Statistical analysis and ANN modeling for predicting hydrological extremes under climate change scenarios: The example of a small Mediterranean agro-watershed

Kourgialas Nektarios, Dokou Zoi, Karatzas Giorgos

Full record


URI: http://purl.tuc.gr/dl/dias/5DBD8E94-C9C2-4CF0-8532-2A3341D8E187
Year 2015
Type of Item Peer-Reviewed Journal Publication
License
Details
Bibliographic Citation N.N. Kourgialas , Z. Dokou, and G.P. Karatzas, "Statistical analysis and ANN modeling for predicting hydrological extremes under climate change scenarios: The example of a small Mediterranean agro-watershed," Journal of Environmental Management, vol. 154, pp. 86–101, May. 2015. doi: 10.1016/j.jenvman.2015.02.034 https://doi.org/10.1016/j.jenvman.2015.02.034
Appears in Collections

Summary

The purpose of this study was to create a modeling management tool for the simulation of extreme flow events under current and future climatic conditions. This tool is a combination of different components and can be applied in complex hydrogeological river basins, where frequent flood and drought phenomena occur. The first component is the statistical analysis of the available hydro-meteorological data. Specifically, principal components analysis was performed in order to quantify the importance of the hydro-meteorological parameters that affect the generation of extreme events. The second component is a prediction-forecasting artificial neural network (ANN) model that simulates, accurately and efficiently, river flow on an hourly basis. This model is based on a methodology that attempts to resolve a very difficult problem related to the accurate estimation of extreme flows. For this purpose, the available measurements (5 years of hourly data) were divided in two subsets: one for the dry and one for the wet periods of the hydrological year. This way, two ANNs were created, trained, tested and validated for a complex Mediterranean river basin in Crete, Greece. As part of the second management component a statistical downscaling tool was used for the creation of meteorological data according to the higher and lower emission climate change scenarios A2 and B1. These data are used as input in the ANN for the forecasting of river flow for the next two decades. The final component is the application of a meteorological index on the measured and forecasted precipitation and flow data, in order to assess the severity and duration of extreme events.

Services

Statistics