Evapotranspiration (ET) refers to the total water vapor flux transported from the surface soil and vegetation to the atmosphere, including soil moisture evaporation and transpiration of water via vegetation (Ersi et al., 2022). As an essential component of the water and heat balances, ET is an important ecological and hydrological process that connects water, energy, and carbon cycling in terrestrial ecosystems, especially in arid and semi-arid regions (Zhang et al., 2012a). It is the most important expenditure item in the water balance of terrestrial ecosystems. Approximately 70% of global surface precipitation returns to the atmosphere through ET, with over 90% occurring in arid regions (Shi et al., 2022). Owing to the scarcity of water resources in arid and semi-arid regions, and often due to terrain, the climate, and other factors, the effect of ET is strong. This not only affects the amount of runoff and replenishment precipitation, but also affects processes such as plant productivity, soil respiration, and biogeochemical cycling (Kurc and Small, 2004).

The main surface cover in arid and semi-arid areas is grassland, which can tolerate limited-water conditions. As an important resource for animal husbandry, grasslands are a significant part of the agricultural sector (Rjili et al., 2023) and are closely related to the livelihoods of residents and sustainable development. ET plays an important role in the grassland water cycle, and is the main factor affecting the water and energy balance of grassland ecosystems (Chang et al., 2016). Calculating the actual ET of grasslands and studying their spatiotemporal distribution characteristics can help us to understand the law of grassland water dissipation and provide an important basis for water resource regulation (Zheng et al., 2013).

The Mongolian Plateau has a temperate continental climate, with most of the area comprising grassland or bare land. Grasslands cover an area of approximately 2.0156×10⁶ km², including 1.1356×10⁶ km² in Mongolia and 0.88×10⁶ km² in the Inner Mongolia Autonomous Region. It is an important component of the global grassland ecosystem (Zhao et al., 2015) and is the largest and most continuous terrestrial ecosystem in Eurasia (Dong et al., 2019). The Mongolian Plateau is one of the three main grazing grassland animal husbandry regions worldwide (Neely et al., 2009). 146.8 million heads of livestock were raised by the end of 2022, of which 71.1204 million heads were raised in Mongolia and 75.7469 million heads were raised in the Inner Mongolian Autonomous Region. Most grazing livestock are horses, cows, camels, sheep, and goats, with total output of 16.7044 million t of dairy products and 9.3219 million t of livestock products, such as beef and mutton, maintaining the livelihoods of approximately 10 million low-income people (Neely et al., 2009). However, available water resources are becoming a critical limiting factor for economic development, and with the increase in population and animal husbandry, the demand for water is rapidly increasing (Vörösmarty et al., 2010). The Mongolian Plateau has a typical arid and semi-arid climate, with an uneven distribution of water and heat, low rainfall, and limited water resources that restrict the development of animal husbandry, agriculture, and socioeconomic development in arid and semi-arid areas.

The ecological environment of the Mongolian Plateau is fragile, and the degree of grassland degradation gradually increases from the northeast to the southwest (Zhang et al., 2018). The vegetation coverage in the southwestern region is low, and desertification is severe (Meng et al., 2020). The Mongolian Plateau is affected by human activities and climatic factors such as rainfall, resulting in a prominent supply-demand contradiction of water resources and poor ecological stability. The fragility of ecosystems makes them particularly sensitive to climate change (Chen et al., 2012). ET is an important component of the surface energy and water balance, as well as a crucial driving force for the groundwater, soil, vegetation, and atmospheric water cycle (Wang, 2022). In water-deficient regions with fragile ecological environments, the ET is sensitive to vegetative dynamics and climate change (Li et al., 2015). Analyzing the trend of ET changes has important reference significance for the optimization of the management of animal husbandry, regional water resources, and the evaluation of grassland ecosystems in the Mongolian Plateau.

Research in this field has developed for over 200 years. Common models for estimating ET include the empirical statistical model, energy balance model, temperature vegetation index feature space method, Penman-Monteith model, and complementary correlation model. The empirical statistical model is only suitable for estimating ET in small areas, relying more on surface observation data, and has poor portability, making it difficult to apply to accurate estimation of ET in large areas (Zhang et al., 2012b); The complementary correlation model can generate significant errors in areas with significant terrain fluctuations (Liu et al., 2004); The temperature vegetation index feature space method is mostly suitable for flat underlying surfaces, and it is difficult for the method to accurately determine the dry and wet edges (Zhang et al., 2012b); The energy balance model requires numerous parameters and is sensitive to surface temperature anomalies. The challenge of more accurately spatializing meteorological data still needs to be overcome (Zhang et al., 2012b). The Penman-Monteith model can better understand how changes in climatic variables affect ET, including solar radiation, relative humidity, maximum and minimum temperatures, and wind speed (Sabino and de Souza, 2023). It has flexible time steps and can mimic continuous-time series ET, making it suitable for studying vegetation ET in arid and semi-arid areas. Li et al. (2023) used the Penman-Monteith model to calculate the global monthly ET from 1982 to 2018. Su et al. (2023) used the Penman-Monteith model to calculate the potential and crop ET of Chinese cotton over a long period and analyzed their spatiotemporal changes. Lan et al. (2023) used the Penman- Monteith model to calculate potential ET in the western Sichuan Plateau and analyzed the spatiotemporal variation characteristics of drought in the western Sichuan Plateau. Aschale et al. (2023) used the Penman-Monteith model to calculate the long-term potential ET in Sicily, providing a foundation for water resource management and improving irrigation efficiency. However, when most scholars use the Penman-Monteith model to estimate ET, the underlying surface conditions of certain vegetation types are relatively complex, and the surface impedance of the Penman-Monteith model is difficult to determine, requiring a large amount of meteorologied data and oversimplifying canopy impedance. There are certain limitations for areas with complex vegetation on the underlying surface; therefore, there is room for further research to improve simulation accuracy.

How applicable is the Penman-Monteith model to the Mongolian Plateau? What are the overall distribution characteristics of actual ET over the Mongolian Plateau? How do the spatiotemporal patterns and changes occur? These issues need to be addressed urgently. This study expects to analyze the spatiotemporal distribution of long-term ET in the Mongolian Plateau using the Penman-Monteith model, targeting the needs of this field and region. This study aims to enrich the ET data products of the Mongolian Plateau; enhance our understanding of the distribution of water resources in the region; help in understanding the changes in hydrology, ecology, animal husbandry, and grassland resources in the Mongolian Plateau; and promote the future regulation and sustainable development of regional water resources and animal husbandry.

ET is a key variable in the water cycle that not only reflects the feedback effect of the ecosystem on the climate system, but also provides important reference value for the utilization of water resources (Xue et al., 2023). In this study, the spatiotemporal distribution characteristics of long-term ET on the Mongolian Plateau from 2011 to 2022 were determined based on the Penman-Monteith model and data such as temperature, air pressure, and sunshine hours. The effects of precipitation and vegetation coverage of the underlying surface on ET over the Mongolian Plateau were also analyzed.

The ET is mainly composed of soil evaporation and vegetation transpiration. Transpiration is not only affected by water conditions, but also by vegetation coverage on the underlying surface (Bremer and Ham, 1999). The Mongolian Plateau is located deep inland and is an arid and semi-arid region with a vast territory and rich vegetation. The underlying surface of typical grasslands, desert grasslands, and other grasslands as land cover types accounts for approximately 95% of the Mongolian Plateau. The actual ET is mainly controlled by local water conditions, and the amount of precipitation directly affects the amount of local ET (Kang, 2023). The actual ET of the Mongolian Plateau has clear interannual spatial distribution characteristics and is highly similar, showing high values in the north and east and low values in the central and south bare land areas. The ET continued to increase from southwest to northeast, with the highest ET occurring in the Selenga River Basin in Mongolia and in the eastern part of the Inner Mongolia Autonomous Region, whereas the lowest ET occurred in the southern part of the Mongolian Plateau. As most areas of the underlying surface types in the central and southern parts of the Mongolian Plateau are bare land, and the desertification phenomenon in the southern region of Mongolia is very strong (Meng et al., 2020), precipitation is sparse, vegetation coverage is low, vegetation is sparse, and transpiration is weak. In addition, rainfall in the region is sparse, and the amount of water available for evaporation is limited; therefore, the actual ET is also low. Grasslands, such as meadow and typical grasslands, are mainly distributed in northern Mongolia and the eastern Inner Mongolia Autonomous Region, represented by the Selenga River Basin, with high precipitation, higher vegetation coverage, and vegetation transpiration, supplemented by surface water resources, such as rivers, for ET; therefore, the actual ET is higher. Areas with high vegetation coverage and abundant precipitation can reach maximum depths of >500 mm in suitable years.

The spatial distribution of mid to high annual and interannual ET on the Mongolian Plateau had a high degree of similarity, showing high levels in the north and east and low levels in the middle and southern bare land areas. In May, June, and October, the ET in the eastern part of the Inner Mongolia Autonomous Region was highest. In January, February, November, and December, the ET did not exceed 8 mm in the area with the highest value, and the minimum ET reached 0.03 mm. The area with the lowest ET in the southern desert area had an ET of 0-20 mm throughout the year, and the ET gradually increased from southwest to northeast. In terms of the temporal distribution, there was a parabolic pattern of first increasing and then decreasing, with July as the turning point. The annual ET was concentrated from May to September. During these five months, there was a large gap between the areas with the highest and lowest ET, reaching over 100 mm at the highest point and only approximately 10 mm at the lowest point. In June, the ET at the highest point was significantly different from that in most areas. The ET in most regions was approximately 25-50 mm, but the highest ET reached 90 mm.

From May to September, ET and precipitation were concentrated within the same period. The cumulative average precipitation on the Mongolian Plateau was 198.47 mm, accounting for 82.9% of the annual precipitation at the station. In May, vegetation gradually entered the greening season, with increased vegetation coverage, increased precipitation, improved vegetation and surface water resource conditions, and increased ET, which peaked in July. Subsequently, the temperature and precipitation gradually decreased. The cumulative average precipitation in January, February, October, and December was 24.58 mm, accounting for only 10.3% of the annual average precipitation; during this period, the average temperature on the Mongolian Plateau reached below 0 ℃, with a decrease in vegetation coverage. The ET of vegetation decreased, and the overall ET decreased significantly from its peak. Water, thermal, and vegetation conditions caused decreases in ET. After November, the ET in most areas was less than 2 mm, and in some areas, it was almost 0 mm.

In the research on ET related to the Mongolian Plateau region, many studies on Mongolia or the Inner Mongolia Autonomous Region alone, but much less literature has been published on the entire Mongolian Plateau is far less than on the above two regions. Therefore, studying the temporal and spatial characteristics of ET over the Mongolian Plateau provides data for water resource balance, grassland resource regulation, and animal husbandry development. The spatial distribution characteristics of the ET of Han Dianchen (Han, 2022; Han et al., 2022), and Zhao et al. (2021) on the Inner Mongolian Plateau are consistent with the results of this study, both showing low values in the northwest and high values in the northeast, and also showing a single peak change trend with July as the inflection point. When studying the spatiotemporal variation of ET in Mongolia, Liu et al. (2013) and Ersi et al. (2022) found that the spatiotemporal variation was consistent with that in this paper, with high values in the north and low values in the south. However, in a previous study conducted in southern Mongolia, the minimum annual ET was approximately 60 mm, whereas in the present study, the minimum ET was approximately 70 mm. The specific reason for this is that there are few meteorological stations in southern Mongolia; however, there are many meteorological stations in the northern region, especially in the Selenga River Basin, where ET can reach up to 500 mm. Therefore, when using ArcGIS for Kriging interpolation, the influence of the northern values is reflected in the figure, which shows that the ET in the southern region exceeded the calculated value.

In this study, owing to the sparse distribution of meteorological stations in the desert areas of the southern Mongolian Plateau, the number of stations was much lower than those in the northern and eastern watersheds and grassland areas. Therefore, when using the Penman-Monteith formula to calculate ET in arid areas with sparse stations, the accuracy of ET obtained in areas with low land coverage, such as the bare land in the southern region, is lower than that in areas with high vegetation coverage in the northern and eastern parts. The forest coverage of the Mongolian plateau is only 5%, and it is mostly dispersed in the meadow grassland (Zheng et al., 2023), therefore, in this study, we did not calculate the ET of forests as a single class. The study of actual ET using the Penman-Monteith model will play an important role in the development of animal husbandry, utilization of surface-available water resources, and sustainable development in arid and semi-arid regions, such as North Asia, Central Asia, and the Middle East. In subsequent ET research, further analysis and improvements will be conducted to improve the applicability of the Penman-Monteith model in these regions.

Taking the Mongolian Plateau as the research area, the Penman-Monteith model was used to calculate and analyze the temporal and spatial distribution patterns of ET over the Mongolian Plateau and initially reveal the ET change trend of the Mongolian Plateau ecosystem. The specific conclusions are as follows:

(1) The spatial distribution characteristics of ET in the Mongolian Plateau showed high values in the north and east and low values in the middle and south. The ET gradually decreased from the northeast to the southwest, and the regional annual ET varied significantly, which was similar to the spatial distribution pattern of the land cover. The areas with the highest ET were the Selenga River Basin and eastern part of the Mongolian Plateau, with an average annual ET of over 400 mm and the highest value reaching 500 mm each year. Due to their high land vegetation coverage and abundant grassland resources, these areas are mostly meadow grasslands and typical grasslands with abundant precipitation and strong vegetation transpiration, resulting in higher ET. The southern part of the Mongolian Plateau had the lowest ET, with the highest annual ET being approximately 100 mm and the lowest reaching 65 mm. The land cover types in this area are mostly desertified grasslands and bare land, with low vegetation coverage and sparse precipitation. The soil moisture content was low, resulting in a lower ET than that in the other regions. Overall, the ET in the northern region was higher than that in the southern region.

(2) The temporal distribution of the Mongolian Plateau showed a parabolic pattern, first increasing and then decreasing, with July as the turning point. The ET intensity was related to factors such as vegetation coverage, temperature, and precipitation. May-September was a period of high ET on the Mongolian Plateau. During this period, the temperature gradually increased, precipitation was relatively abundant, and the soil moisture content increased. The vegetation was lush, and transpiration from plants was strong. Increases in vegetation transpiration and soil moisture content led to gradual increases in ET. Owing to the large amounts of desertified grasslands and bare land in the southern part of the Mongolian Plateau, sparse vegetation, low vegetation transpiration, and low soil moisture content, the ET in the southern part of the Mongolian Plateau is relatively low, which also impacts the ecological environment. The ecological environment is fragile and the ability to prevent and control desertification is weak.