Combining discrete wavelet decomposition with soft computing techniques to predict monthly evapotranspiration in semi-arid Hakkâri province, Türkiye

Accurate prediction of evapotranspiration values is important in planning agricultural irrigation, crop growth research, and hydrological modeling. This study is aimed at estimating monthly evapotranspiration (ET) values in Hakkâri province by combining support vector regression, bagged tree, and boosted tree methods with wavelet transform. For this purpose, precipitation, runoff, surface net solar radiation, air temperatures, and previous ET values were divided into sub-signals with various mother wavelets such as Daubechies 4, Meyer, and Symlet 2 and presented as input to machine learning (ML) algorithms. The study's main contribution to the literature is to reveal which wavelet-based machine learning model, mother wavelet type, and combination of meteorological data show the most realistic results in ET estimation. While establishing the models, the data were divided into 80% training and 20% testing. The models' performances were based on the widely used root mean square error, mean absolute error, determination coefficient, and Taylor diagrams. As a result of the study, it was revealed that the hybrid wavelet ML, which is established with input combinations separated into subcomponents by wavelet transform, generally produces more successful predictions than the stand-alone ML model. In addition, it was revealed that the optimum ET forecasting model was obtained with the wavelet bagged tree algorithm with Symlet 2 mother wavelet. Even though the best model established is based on the precipitation and temperature inputs, it was revealed that past ET, solar radiation, and runoff values are also effective inputs in ET prediction. The results can also be used in other regions of the world with semi-arid climates, such as Hakkâri. The study's outputs provide essential resources to decision-makers and planners to manage water resources and plan agricultural irrigation.

Two-stage model for estimating the spatiotemporal distribution of hourly PM1.0 concentrations over central and east China

Widespread and severe PM_1.0 (particulate matter ≤1.0 μm) pollution in China has a significant negative influence on human health. However, knowledge of the regional spatiotemporal distribution of PM_1.0 has been hindered by sparsely distributed PM_1.0 concentration data. In this work, a two-stage model (linear mixed effect-bagged tree model) was proposed for estimating hourly PM_1.0 pollution levels from July 2015 to June 2017 over central and east China by using Himawari-8 aerosol products and coincident geographic data, meteorology, and site-based PM_1.0 concentrations from ground monitoring network. The cross-validation for the developed model displayed R² and mean absolute error value of 0.80 and 9.3 μg/m³, respectively. Validation demonstrated that the model accurately estimated hourly PM_1.0 concentrations with high R² of 0.63-0.85 and low bias of 8.7-10.1 μg/m³. The estimated PM_1.0 concentrations on daily scale showed peaks with PM_1.0 of 36.9 ± 8.4 μg/m³ at rush hours during daytime. Seasonal distribution displayed that summer was cleanest with an average PM_1.0 of 20.9 ± 6.8 μg/m³ and winter was the most polluted season with an average PM_1.0 of 45.6 ± 16.8 μg/m³. These results indicated that the proposed satellite-based model can estimate reliable spatial distribution of PM_1.0 concentrations over a large-scale region.