Seasonal rainfall hindcasting using ensemble multi-stage genetic programming
- PDF / 3,010,994 Bytes
- 12 Pages / 595.276 x 790.866 pts Page_size
- 42 Downloads / 158 Views
ORIGINAL PAPER
Seasonal rainfall hindcasting using ensemble multi-stage genetic programming Ali Danandeh Mehr 1 Received: 3 July 2020 / Accepted: 14 October 2020 # Springer-Verlag GmbH Austria, part of Springer Nature 2020
Abstract Rainfall hindcasting is one of the most challenging tasks in the hydrometeorological forecasting community. The current ad hoc data-driven approaches appear to be insufficient for forecasting rainfall. The task becomes more difficult, when the forecasts are over a long period of time. To increase the accuracy of seasonal rainfall hindcasting, this paper introduces an ensemble evolutionary model that integrates two genetic programming techniques: gene expression programming (GEP) and multi-stage genetic programming (MSGP). To demonstrate the development and validation procedures of the new model, the rainfall data from the Antalya meteorology station was used. The model performance was evaluated in terms of different statistical measures and compared with that of the state-of-the-art gradient boosted decision tree (GBT) model developed as a reference model in this study. The performance results during testing showed that the proposed ensemble model has increased the seasonal forecasting accuracy of the GEP and MSGP models up to 30%. The GBT was found comparable to the proposed model during training period; however, it drastically underestimated extreme wet seasons during testing. Abbreviations ACF Autocorrelation function ANN Artificial neural networks DM Data mining EGEP Ensemble gene expression programming EMSGP Evolutionary ensemble multi-stage genetic programing GBT Gradient boosted tree GEP Gene expression programming GP Genetic programing LEM Linear ensemble model MAPE Mean absolute percentage of error MLR Multi-variable linear regression MSGP Multi-stage genetic programming NSE Nash-Sutcliff Efficiency OAM Ordinary arithmetic mean PACF Partial autocorrelation function RMSE Root mean squared error SR Seasonal rainfall SSA Singular spectrum analysis SVM Support vector machine
* Ali Danandeh Mehr [email protected] 1
Department of Civil Engineering, Antalya Bilim University, Antalya, Turkey
1 Introduction The increase in the concentration of atmospheric carbon dioxide and the global temperature have significantly affected rainfall patterns in both catchment and regional scales (Wang et al. 2013; Danandeh Mehr and Kahya 2017; Zhang et al. 2019; Homsi et al. 2020; Qian et al. 2020). Rainfall process is highly random due to the stochastic attributes of its triggering factors, such as temperature, wind speed, and humidity, which results in inaccurate prediction of its occurrence and magnitude. The level of uncertainty in rainfall pattern is significantly higher than the other hydrological parameters such as temperature, and streamflow, and therefore, rainfall forecasting is more challenging in water resources management (Fawcett and Stone 2010; Vano et al. 2019; Ni et al. 2020). Consequently, loss of life and damage to the properties remain inevitable due to unexpected heavy
Data Loading...
