Actual evapotranspiration (ET_a) plays a crucial role in the water and energy cycles of the earth. An accurate estimate of the ET_a is essential for management of the water resources, agriculture, and irrigation, as well as research on atmospheric variations. Despite the importance of accurate ET_a values, estimating and mapping them remains challenging due to the physical and biological complexity of the ET process. As a novel approach for rapid and reliable estimation of ET_a, the present study develops automated deep learning (AutoDL) models that incorporate a metaheuristic optimization algorithm for image processing, architectural design, and hyperparameter tuning. The proposed AutoDL models integrate three different spatial and channel attention mechanisms, including a novel activated spatial attention mechanism (ASPAM), with the U-Net architecture. Bypassing the need for meteorological inputs, the proposed framework uses Moderate Resolution Imaging Spectrometer (MODIS) products and Digital Elevation Model (DEM) data as inputs. To evaluate the performance of the models, they are applied to three study areas in the United States with various climatic characteristics. According to the results, during the spring and summer, when the target values have higher certainty, the estimations are highly promising, with R² as high as 0.91 and MAPE as low as 6.40%. Furthermore, the proposed ASPAM module improves the accuracy of ET_a estimations compared to attention gate (AG) and squeeze and excitation (SE) attention modules. The results also indicate that the MODIS raw products and derived vegetation and water indices can predict ET_a within a reliable range of accuracy, with the addition of DEM data marginally enhancing the models' performance. The automatic workflow of this model makes it significantly easy to use, contributing to its applicability and generalizability for enhancing atmospheric research.

Rainwater harvesting (RWH) has been recognized as one of the most reliable and efficient methods for water supply, especially in arid and semi-arid regions (ASARs) facing freshwater scarcity. Nevertheless, due to the inherent uncertainty of input data and subjectivity involved in the selection of influential parameters, the identification of RWH potential areas is a challenging procedure. In this study, two approaches for locating potential RWH sites were implemented. In the first approach, a frequently-used method of the multi-criteria decision analysis and geographic information system (MCDA-GIS) was utilized, while, in the second approach, a novel strategy of integrating the soil and water assessment tool (SWAT) model as a hydrology model into an MCDA-GIS method was proposed to evaluate its performance in locating potential RWH sites. The Mashhad Plain Basin (MPB) was selected as a case study area. The developed potential RWH maps of the two approaches indicated similar patterns for potential RWH areas; in addition, the correlation coefficient (CC) between the two obtained maps were relatively high (i.e., CC = 0.914) revealing that integration of SWAT as a comprehensive hydrologic model does not necessarily result in very different outputs from the conventional method of MCDA-GIS for RWH evaluation. The overlap of developed maps of the two approaches indicated that 3394 km² of the study area, mainly located in the northern parts, was identified as high-potential RWH areas. The performed sensitivity analysis indicated that rainfall and slope criteria, with weights of 0.329 and 0.243, respectively, had the greatest sensitivity on the model in the first approach while in the second approach, the criterion of runoff coefficient (with weights of 0.358) had the highest impact. Based on results from the identification of the potential locations for conventional RWH techniques, pond and pan techniques are the most proper options, covering high-potential areas of RWH more effectively than other techniques over MPB.