Climate Change Research

Climate Change Research

Modelling daily pan evaporation of Zayanderud dam station utilizing artificial intelligence and time series models

Document Type : Original Article

Authors
Ahmad reza Ghasemi 1
Marziyeh Esmaeilpour 2
1 Associate Professor, Department of Water Engineering, Shahrekord University, Shahrekord, Iran,
2 Assistant Professor, Department of Geography and Urban Planning, Maragheh University, Maragheh, Iran,
10.30488/ccr.2024.433405.1193
Abstract
The pan evaporation is used as a practical parameter in various fields, such as estimating water loss from lakes and dams, as well as estimating the Plant water requirement, especially in areas where there is no lysimeter information. Modeling this parameter can be useful in the estimation of missing data and long-term planning of water resources and agricultural development. In this research, by using an artificial intelligence model (Gene Expression Programming, GEP) and two time series models (Fourier and ARIMA), evaporation from the pan at Zayanderud dam station, was modeled in the period from 1344 to 1396 (53 years). The time series of pan evaporation on daily scale for the months of June (Khordad), July (Tir), August (Mordad), September (Shahrivar), and October (Mehr) as input to the Fourier and ARIMA models and 4 different patterns including the use of daily evaporation data 1 month, 2 months, 3 months and 4 months ago, were used as input for gene expression programming model. The results showed that the GEP model has acceptable results only in Mehr, and for other months the results are not acceptable in terms of statistical indicators. The daily evaporation estimated error was found to be 0.38 mm in Mehr. This error was acceptable based on the coefficient of determination of 0.84, the Nash-Sutcliffe coefficient of 0.83 and the Willmott's index of  agreement of 0.95. Unlike the gene expression programming model, the Fourier model provided acceptable results in all the studied months. The error values were obtained between 1.02 and 0.7 mm per day in all the studied months, which is equivalent to 5.2 to 8.8 percent. Comparing the results of the above two models with the ARIMA model showed that, the error values of the ARIMA model in all months are higher (9.4 to 19.6%) than the Fourier model, and Gene Expression Programming model. Therefore, the best model for estimating pan evaporation in October (Mehr) is the Gene Expression Programming model and in the rest of the months, the Fourier model. The evaluation of the model in estimating the extremes daily evaporation data also showed that the highest and lowest accuracy is in October (Mehr) and June (Khordad), respectively. In total, according to the statistical indices, the ability of the Fourier model to estimate the daily evaporation in the Zayanderud dam station was proved. Therefore, this model can be recommended to estimate the daily evaporation and also to reconstruct the missing data in this station
Keywords
Evaporation
Missing data
Simulation
Extreme data

Subjects

  1. Alarcon, A., Cortes, D., Alvarez, J. Gonzalez, Y. (2022). Improving Monthly Rainfall Forecast in a Watershed by Combining Neural Networks and Autoregressive Model. Environmental Processes, 9(53): 1-26.
  2. Gandomi, A., Alavi, A. Mirzahosseini M. & Moqhadas F. (2011). Nonlinear genetic-based models for prediction of flow number of asphalt mixtures. Journal of Materials in Civil Engineering, 23(4), 248-263.
  3. Gaur, S., Singh, R., Bandyopadhyay, A., & Singh, R. (2023). Diagnosis of GCM-RCM-driven rainfall patterns under changing climate through the robust selection of multi-model ensemble and sub-ensembles. Climatic Change,176(2): 1-30.
  4. Goel, A. (2009). Application of SVMs Algorithms for Prediction of Evaporation in Reservoirs. World Environmental and Water Resources Congress, Missouri, United States.
  5. Hael, M.A., Yongsheng, Y. & Saleh, B.I. (2020). Visualization of rainfall data using functional data analysis. Applied Science, 2(461). https://doi.org/ 10.1007/s42452-020-2238-x
  6. Khanal, N., Matin, M., Uddin, K., Poortinga, A., Chishtie, F., Tenneson, K., & Saah, D. (2020). Comparison of Three Temporal Smoothing Algorithms to Improve Land Cover Classification: A Case Study from NEPAL. Remote Sensing, 12, 2888.
  7. Kim, S., Shiri, J., Singh, V., Kisi, O., & Landeras, G. (2015). Predicting daily pan evaporation by soft computing models with limited climatic data. Hydrological Sciences Journal, 60(6): 1120-1136.
  8. Koza, J.R. (1993). Hierarchical Automatic Function Definition in Genetic Programming. Foundations of Genetic Algorithms, 2, 297-318.
  9. Laguardia, G. (2011). Representing the precipitation regime by means of Fourier series. International Journal of Climatology, 31: 1398–1407.
  • Li, L., Zhou, X., Li, Y., Gong, C., Lu, L., Fu, X., & Tao, W. (2017). Water absorption and water/fertilizer retention performance of vermiculite modified sulphoaluminate cementitious materials. Construction and Building Materials, 137, 224-234.
  • Mazelan, N.A., & Suhaila J. (2023). Exploring rainfall variabilities using statistical functional data analysis. Earth and Environmental Science. 1167: 1-10. doi:10.1088/1755-1315/1167/1/012007.
  • Miranda, A., Herrera, M., & Castano, V. (2019). Meteorological Temperature and Humidity Prediction from Fourier-Statistical Analysis of Hourly Data. Advances in Meteorology, 34(2), 2-13. https://doi.org/10.1155/2019/4164097
  • Moriasi, D., Arnold, J., VanLiew, M.W., Bingner, R.L., Harmel, R.D., & Veith, T. (2007). Model evaluation guidelines for systematic quantification of accuracy in watershed simulations. American Society of Agricultural and Biological Engineers, 50(3), 885-905.
  • Salas, J.D., (1993). Analysis and Modeling of Hydrologic Time Series. Pp. 1-72, In: Maidment D.R, Handbook of Hydrology, McGraw- Hill.
  • Shoaib, M., Shamseldin, A.S., Melville, B., & Muneer, M. (2015). Runoff forecasting using hybrid Wavelet Gene Expression Programming (WGEP) approach. Journal of Hydrology, 527: 326- 344.
  • Soylu, M., Lenters, John, D., Istanbulluoglu, E., & Loheide, S. (2012). On evapotranspiration and shallow groundwater fluctuations: A Fourier-based improvement to the White method. Papers in Natural Resources. 613. https://digitalcommons. unl.edu/natrespapers/613
  • Sudheer, K.P. (2000). Modeling hydrological processes using neural computing technique. PhD Thesis, Indian Institute of Technology, Delhi. India.
  • Tularam, G., & Ilahee, M. (2010). Time Series Analysis of Rainfall and Temperature Interactions in Coastal Catchments. Journal of Mathematics and Statistics. 6 (3): 372-380.
  • Ustoorikar, K., & Deo, M.C. (2008). Filling up gaps in wave data with genetic programming. Marine Structures, 21(2): 177-195.
  • Wu, X., Zhou, J., Yu, H., Li, D., Xie, K., Chen, Y., Hu, J., Sun, H., & Xing, F. (2021). The development of a hybrid wavelet-ARIMA-LSTM model for precipitation amounts and drought analysis. Atmosphere.  12: https://doi.org /10.3390/atmos12010074.
  • Yassin, M., Alazba, A., & Mattar, M. (2016). Artificial neural networks versus gene expression programming for estimating reference evapotranspiration in arid climate. Agricultural Water Management. 163: 110-124.