**1. Introduction**

Among the interactions amid various spheres of the climate system, the land–atmosphere interaction plays an important role in influencing the evolution of weather and climate [1]. The land surface is closely linked to the atmosphere through energy and water cycles, causing increases in the temperature variability [2] and the frequency of high-temperature heat waves [3], and exacerbating compound soil and atmospheric drought intensity [4]. The land–atmosphere coupling strength (CS) is a key indicator to characterize the land– atmosphere interaction. Areas with a stronger CS imply a greater influence of land surface on regional weather and climate. The global strong land–atmosphere coupling zone is

**Citation:** Yang, Z.; Zhang, Q.; Zhang, Y.; Yue, P.; Zhang, L.; Zeng, J.; Qi, Y. Hydrothermal Factors Influence on Spatial-Temporal Variation of Evapotranspiration-Precipitation Coupling over Climate Transition Zone of North China. *Remote Sens.* **2022**, *14*, 1448. https://doi.org/ 10.3390/rs14061448

Academic Editor: Nicola Montaldo

Received: 7 February 2022 Accepted: 15 March 2022 Published: 17 March 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

mostly located in the arid-humid climate transition zone [5], including west-central North America, parts of Eurasia, Australia, Argentina, the Sahel region of North Africa, and South Africa [6,7]. Accurate acquisition of land–atmosphere forcing signals in these land– atmosphere coupling "hotspots" is important for improving the forecasting capabilities of the weather and climate [5,8,9].

Land–atmosphere coupling includes a series of complex processes: land surface state anomalies first cause changes in surface fluxes, which in turn affect precipitation through feedbacks from the land surface to the atmosphere [7,10]. Regulated by the surface energy balance, sensible heat fluxes change synergistically with evapotranspiration. Thus, evapotranspiration can regulate sensible heat fluxes via the Bowen ratio. Therefore, evapotranspiration is often considered as a key process in land–atmosphere coupling processes [11,12]. However, studies have shown that the influence of evapotranspiration on precipitation has the greatest uncertainty in land–atmosphere coupling processes [13]. Generally, evapotranspiration can affect precipitation in three ways. First, evapotranspiration directly affects atmospheric precipitation through water recycling. Evapotranspiration can return approximately 70% of precipitation to the atmosphere [14], and atmospheric precipitable water directly influences precipitation. This mechanism is more prominent in water-scarce areas [15,16]. Second, changes in evapotranspiration can also alter the regional pressure field, which can cause adjustments in atmospheric circulation and lead to large-scale precipitation changes [17]. Moreover, evapotranspiration and the sensible heat fluxes regulated by it affect the atmospheric stability state by altering the atmospheric temperature and humidity profiles, thus affecting convective precipitation [18]. The last of these pathways, evapotranspiration–precipitation local coupling, is the most important method for evapotranspiration to influence precipitation [13,19,20]. However, due to the complexity of the influence of evapotranspiration on atmospheric stability, evapotranspiration-precipitation local coupling has significant uncertainty and becomes a challenging problem in the current land–atmosphere coupling research.

Local evapotranspiration–precipitation coupling is controlled by many factors and the influence mechanism is very complex. Water and energy cycles are the key physical processes throughout the coupled land–atmosphere interaction [21]. Moisture and thermal factors directly affect the evapotranspiration process: in dry areas, evapotranspiration is controlled mainly by moisture factors, while in humid areas evapotranspiration is controlled mainly by thermal factors [22,23]. In turn, evapotranspiration affects the structural characteristics of the boundary layer through the transport of water and heat, and sufficient surface moisture can lead to lowered boundary layer height (BLH) and lifting condensation level (LCL) and increased moist static energy (MSE); in contrast, limited surface water and adequate thermal conditions raise the BLH and LCL. In general, lower BLH and LCL and larger MSE can lead to an increase in convective available potential energy (CAPE) and a higher probability of convective precipitation [20]. However, observation studies have found that arid conditions can also promote physical mechanisms that favor the generation of convective precipitation [24–27]. This is attributed to the large sensible heat flux caused by strong thermal factors in arid regions that weakens the convective inhibition energy (CIN) [28], thereby increasing the probability of convective precipitation; there is also a negative feedback of evapotranspiration on precipitation. Therefore, both positive and negative feedbacks between evapotranspiration and precipitation are closely related to moisture and thermal properties, i.e., moisture and thermal factors are the most critical forcing factors affecting the land–atmosphere coupling strength.

In the past 20 years, a large number of studies have paid attention to the spatial and temporal distribution of land–atmosphere coupling and its intrinsic mechanism. The Global Land-Atmosphere Coupling Experiment (GLACE) found that the strong land– atmosphere coupling regions are mostly located in semi-arid and semi-humid climate transition zones [5], and other diagnostic studies based on observational data have also verified this conclusion [6,7,29]. Due to the apparent spatial variability of land–atmosphere CS, some typical regions have attracted wide attention. In North America, the spatial distribu-

tion of land–atmosphere coupling was correlated with the multi-year average soil moisture, and the strong coupling area was mainly distributed in areas with soil moisture ranging from 0.4 to 0.55 [30]. Due to interannual variations in soil moisture, land–atmosphere coupling shows significant interannual fluctuations [31,32]. In southeastern South America, the spatial and temporal distribution of land–atmosphere coupling is correlated not only with soil moisture but also with wet static energy and its vertical gradient that is controlled by soil moisture [33]. Moreover, under future climate warming and humidification, the intensity of land–atmosphere coupling will be significantly weakened due to the gradual shift of evapotranspiration from moisture limitation to energy limitation [34]. In Europe, the northward expansion of the Hadley circulation has caused a northward shift of the climate transition zone, leading to a northward shift of the strong land–atmosphere coupling zone [2]. In Africa, land–atmosphere coupling was negatively correlated with the spatial and temporal distribution of soil moisture, with areas of lower soil moisture and periods of dry moisture exhibiting stronger land–atmosphere coupling. Spatial differences in the soil moisture lead to enhanced sensible heat fluxes in the dry zone and reduced sensible heat fluxes in the wet zone, which in turn trigger mesoscale circulation, and the upward branch of this circulation in the dry zone is an important factor in triggering deep convection [26]. In East Asia, land–atmosphere coupling is strong in north China, where soil moisture is low [35]; land–atmosphere coupling is strong in southwest China in spring, when soil moisture is low [36]; land–atmosphere coupling is strongly influenced by the snow cover in the dry season and by the soil moisture in the rainy season over Tibetan Plateau [37]; the land–atmosphere coupling degree is closely related to the state of surface vegetation in northwest China, and the improvement of vegetation state can improve the surface moisture condition, reduce the land surface evapotranspiration, and decrease the strength of land–atmosphere coupling [38].

The above studies mainly focused on the spatial distribution and temporal variation in the land–atmosphere CS and its relationship with soil moisture, and there is lack of research on the role of thermal factors on the land–atmosphere CS. Theoretically, the thermal properties also play a substantial role in the land–atmosphere coupling. The role of moisture and thermal factors in regulating the land–atmosphere coupling is similar to that of evapotranspiration, as it is regulated by moisture in water-scarce areas and by thermal and energy in water-sufficient areas [39].

Most of northern China makes up a dry-wet climate transitional zone, with dramatic spatial and temporal variations in water and heat characteristics. From the northwest to the southeast, moisture availability decreases, and energy availability increases, while the evapotranspiration control factor gradually changes from moisture to energy limitation [23]. This inevitably affects the spatial and temporal variation of regional land–atmosphere coupling. However, the spatial and temporal distribution of land–atmosphere coupling in north China's climate transition zone remains unclear, and it is also unknown how hydrothermal factors affect the spatial-temporal variation in land–atmosphere coupling.

In this study, an evapotranspiration–precipitation coupling index proposed by Zeng et al. [6] was used to diagnose the land–atmosphere CS in the climate transitional zone of north China, and the main objectives were to (i) analyze the spatial distribution and temporal evolution characteristics of the land–atmosphere coupling and (ii) explain the impacts of moisture and thermal factors on the land–atmosphere CS. The results of the study are expected to enhance the understanding of land–atmosphere coupling mechanisms in the climate transitional zone of northern China.

### **2. Data and Methods**

### *2.1. Study Area and Sites*

The climate transitional zone of northern China (CTZNC) is selected as the study area in this work, with the spatial extent of the region between 33◦N and 45◦N and 100◦E and 125◦E. The geographical area and the climatic background are shown in the black box

of Figure 1. The climate background is classified using the dryness index defined by the United Nations Environment Programme:

$$\text{AI} = \text{P} / \text{PET}\_{\text{\textquotedblleft}} \tag{1}$$

where, AI is the dryness index, P is the average annual precipitation in mm, and PET is the average annual potential evapotranspiration in mm. AI < 0.05 is classified as a hyper-arid zone, 0.05 < AI < 0.2 as an arid zone, 0.2 < AI < 0.5 as semi-arid zone, 0.5 < AI < 0.65 as a sub-humid zone, and AI > 0.65 as a humid zone. As shown in Figure 1, the study region mainly includes arid, semi-arid, sub-humid, and humid climate. It is not only a climate transition region but also a typical ecological transition zone, as well as a major activity area of the northern edge of the East Asian summer monsoon.

**Figure 1.** Climatic background of climate transitional zone of northern China (hyper-arid: AI < 0.05, arid: 0.05 < AI < 0.2, semi-arid: 0.2 < AI < 0.5, sub-humid: 0.5 < AI < 0.65, and humid: AI > 0.65). The soil moisture observation sites are indicated by stars.

Four sites are selected for study, the underlying surfaces of which are either grassland or cropland. The locations of the sites are shown in Figure 1, and the climate and environmental background are given in Table 1.


**Table 1.** Brief description of the soil moisture observatories.
