| چکیده انگلیسی مقاله |
Background and Purpose: Agriculture serves as the cornerstone of the global economy, providing the main source of food and raw materials for various industries. However, the rising demand for food as a consequence of population growth represents a considerable threat to food security, particularly in light of the limited access to freshwater resources. It is noteworthy that agriculture alone consumes about 70% of the world's freshwater resources, thereby emphasizing the critical need to manage and enhance irrigation efficiency to ensure sustainable food production. Therefore, the management and enhancement of irrigation efficiency are essential. At the core of determining irrigation water requirements lies the concept of actual crop evapotranspiration (ETa), which represents the combined water loss from soil evaporation and plant transpiration. Accurate estimation of ETa is crucial in optimizing irrigation methods, maximizing crop yield, and minimizing water consumption. Various models and tools have been developed to estimate ETa, aiming to provide more user-friendly and efficient methods for farmers and researchers. Given the extensive application of ET estimation models, there is a clear need to focus on the development of accurate and efficient methods for determining this parameter. Thus, this study aims to compare user-friendly ETa estimation methods, including the EEFLUX system, the METRICTOOL tool, and the automatic hot and cold pixel selection method of the SEBAL and METRIC models. Materials and Methods: The Earth Engine Evapotranspiration Flux (EEFLUX) is a version of the METRIC model that operates on the Google Earth Engine platform. METRICTOOL is a new tool in ArcGIS based on the METRIC model, offering enhanced pre-processing capabilities and automatic data identification. This tool reduces computation time by 50% and provides a user-friendly alternative to other existing METRIC model implementation platforms. The automatic hot and cold pixel selection method involves creating a binary map of eligible pixels using a rule-based classifier and a comprehensive search algorithm to identify hot and cold pixels based on defined criteria. To estimate ET using these methods, six Landsat 8 satellite images were utilized during the winter wheat crop planting period at Tehran University farms in Mohammadshahr Karaj. The evaluation of these methods was conducted using alfalfa reference evapotranspiration (ETr) calculated with the FAO-Penman-Monteith method as reference data. Results and Discussion: The Root Mean Square Error (RMSE) values for the EEFLUX system, METRICTOOL, SEBAL, and automatic METRIC tools were determined as 2.45, 0.33, 0.39, and 2.76, respectively. Despite numerical differences, the evaporation and transpiration product of the EEFLUX system showed significant correlations with other methods. For instance, the R2 between ETa estimates from the EEFLUX system and the METRICTOOL tool was found to be 0.91. Although the data from the EEFLUX system may not be precise enough for local studies due to the use of CFSV2 global meteorological data in Iran, they yield acceptable results in large or global-scale studies. The METRICTOOL tool and automatic METRIC model exhibited the highest correlation (R2=0.99) and numerical agreement with each other, with RMSE values of 0.33 and 0.39, respectively, indicating higher accuracy compared to the automatic SEBAL model. Conclusion: The results of the numerical analysis indicate that the automatic hot and cold pixel selection approach can achieve similar accuracy to that of the METRICTOOL tool. This automated approach enhances the efficiency of the model in terms of time and effectiveness, reducing the potential for human error in estimating evapotranspiration for new or inexperienced users, and making these models accessible to the public. Furthermore, EEFLUX data can be utilised for the implementation of management measures in large-scale studies. |