Eine Plattform für die Wissenschaft: Bauingenieurwesen, Architektur und Urbanistik
Eine Plattform für die Wissenschaft: Bauingenieurwesen, Architektur und Urbanistik
Climate Teleconnections Contribution to Seasonal Precipitation Forecasts Using Hybrid Intelligent Model
Long-term precipitation forecasts can provide valuable information to help mitigate some of the outcome of floods and enhance water manage. This study aims to extract significant information from oceanic-atmospheric oscillations that could enhance seasonal precipitation forecasting. A hybrid AI-type data-driven artificial neural network model called MWD-NARX based on a non-linear autoregressive network with exogenous inputs coupled to multiresolution wavelet decomposition (MWD) is then developed in this work. First, MWD is used to decompose climatic indices and precipitation data. Then the NARX ensemble model allowed to identify the statistical links between the decomposed indices and the decomposed precipitation according to temporal scales and to predict each precipitation decomposition. Ensemble precipitation forecasts are carried out over horizons ranging from 1 to 6 months. For operational forecasting, the forecasts obtained from the decompositions are summed to represent the true precipitation forecast value. The seasonal forecasts of average precipitation by sub-basins (SBV) of the Medjerda are carried out. Large scale teleconnections ENSO, PDO, NAO, AO and as well as Mediterranean Oscillation were used as inputs to the model. The forecasting model coupled to data pre-processing method made it possible to produce very satisfactory forecasts of non-stationary data by extracting modes of variability. The results indicate that exogenous inputs like climatic indices clearly improves the accuracy of forecasts on 82% of SBVs and increases the forecast lead-time up to 6 months. This research is the first of its kind using a hybrid prediction approach by means of indices related to ocean-atmospheric oscillations in North Africa.
Climate Teleconnections Contribution to Seasonal Precipitation Forecasts Using Hybrid Intelligent Model
Long-term precipitation forecasts can provide valuable information to help mitigate some of the outcome of floods and enhance water manage. This study aims to extract significant information from oceanic-atmospheric oscillations that could enhance seasonal precipitation forecasting. A hybrid AI-type data-driven artificial neural network model called MWD-NARX based on a non-linear autoregressive network with exogenous inputs coupled to multiresolution wavelet decomposition (MWD) is then developed in this work. First, MWD is used to decompose climatic indices and precipitation data. Then the NARX ensemble model allowed to identify the statistical links between the decomposed indices and the decomposed precipitation according to temporal scales and to predict each precipitation decomposition. Ensemble precipitation forecasts are carried out over horizons ranging from 1 to 6 months. For operational forecasting, the forecasts obtained from the decompositions are summed to represent the true precipitation forecast value. The seasonal forecasts of average precipitation by sub-basins (SBV) of the Medjerda are carried out. Large scale teleconnections ENSO, PDO, NAO, AO and as well as Mediterranean Oscillation were used as inputs to the model. The forecasting model coupled to data pre-processing method made it possible to produce very satisfactory forecasts of non-stationary data by extracting modes of variability. The results indicate that exogenous inputs like climatic indices clearly improves the accuracy of forecasts on 82% of SBVs and increases the forecast lead-time up to 6 months. This research is the first of its kind using a hybrid prediction approach by means of indices related to ocean-atmospheric oscillations in North Africa.
Climate Teleconnections Contribution to Seasonal Precipitation Forecasts Using Hybrid Intelligent Model
Sustain. Civil Infrastruct.
Akhnoukh, Amin (Herausgeber:in) / Kaloush, Kamil (Herausgeber:in) / Elabyad, Magid (Herausgeber:in) / Halleman, Brendan (Herausgeber:in) / Erian, Nihal (Herausgeber:in) / Enmon II, Samuel (Herausgeber:in) / Henry, Cherylyn (Herausgeber:in) / Ouachani, Rim (Autor:in) / Bargaoui, Zoubeida (Autor:in) / Ouarda, Taha (Autor:in)
International Road Federation World Meeting & Exhibition ; 2021 ; Dubai, United Arab Emirates
21.04.2022
14 pages
Aufsatz/Kapitel (Buch)
Elektronische Ressource
Englisch
Seasonal forecasting , Precipitation , Climate teleconnection indices , Multiresolution wavelet decomposition , NARX , Neural networks Engineering , Transportation Technology and Traffic Engineering , Sustainable Development , Building Construction and Design , Building Materials , Earth and Environmental Science
Generating Synthetic Daily Precipitation Realizations for Seasonal Precipitation Forecasts
| British Library Online Contents | 2014
Generating Synthetic Daily Precipitation Realizations for Seasonal Precipitation Forecasts
| Online Contents | 2014
| British Library Conference Proceedings | 2001