EN中文

Journal of Geographical Sciences >

2023 , Vol. 33 >Issue 10: 2031 - 2051

DOI: https://doi.org/10.1007/s11442-023-2164-0

Research Articles

Spatio-temporal pattern and attribution analysis of the mass elevation effect in the Tianshan Mountains in China

  • ZHANG Mingyu , 1, 2 ,
  • ZHANG Zhengyong , 1, 2, * ,
  • LIU Lin 1, 2 ,
  • ZHANG Xueying 1, 2 ,
  • KANG Ziwei 3 ,
  • CHEN Hongjin 1, 2 ,
  • GAO Yu 1, 2 ,
  • WANG Tongxia 4 ,
  • YU Fengchen 1, 2
Expand
  • 1. School of Science, Shihezi University, Shihezi 832000, Xinjiang, China
  • 2. Key Laboratory of Oasis Town and Mountain-basin System Ecology of Xinjiang Bingtuan, Shihezi 832003, Xinjiang, China
  • 3. Xinjiang Institute of Ecology and Geography, CAS, Urumqi 830011, China
  • 4. School of Water Conservancy & Architectural Engineering, Shihezi University, Shihezi 832000, Xinjiang, China
*Zhang Zhengyong (1978-), Professor, specialized in the research of hydrological resources and climate change research. E-mail:

Zhang Mingyu (1999-), Master Candidate, specialized in the research of mountain climate change and ecological environment research. E-mail:

Received date: 2023-07-15

  Accepted date: 2023-08-20

  Online published: 2023-10-08

Supported by

National Natural Science Foundation of China(41761108)

Fold

Abstract

The mass elevation effect (MEE) is a thermal effect, in which heating produced by long wave radiation on a mountain surface generates atmospheric uplift, which has a profound impact on the hydrothermal conditions and natural geographical processes in mountainous areas. Based on multi-source remote sensing data and field observations, a spatial downscaling inversion of temperature in the Tianshan Mountains in China was conducted, and the MEE was estimated and a spatio-temporal analysis was conducted. The GeoDetector model (GDM) and a geographically weighted regression (GWR) model were applied to explore the spatial and temporal heterogeneity of the study area. Four key results can be obtained. (1) The temperature pattern is complex and diverse, and the overall temperature presented a pattern of high in the south and east, but low in the north and west. There were clear zonal features of temperature that were negatively correlated with altitude, and the temperature difference between the internal and external areas of the mountains. (2) The warming effect of mountains was prominent, and the temperature at the same altitude increased in steps from west to east and north to south. Geomorphological units, such as large valleys and intermontane basins, weakened the latitudinal zonality and altitudinal dependence of temperature at the same altitude, with the warming effect of mountains in the southern Tianshan Mountains. (3) The dominant factors affecting the overall pattern of the MEE were topography and location, among which the difference between the internal and external areas of the mountains, and the absolute elevation played a prominent role. The interaction between factors had a greater influence on the spatial differentiation of mountain effects than single factors, and there was a strong interaction between terrain and climate, precipitation, the normalized difference vegetation index (NDVI), and other factors. (4) There was a spatial heterogeneity in the direction and intensity of the spatial variation of the MEE. Absolute elevation was significantly positively correlated with the change of MEE, while precipitation and the NDVI were dominated by negative feedback. In general, topography had the largest effect on the macroscopic control of MEE, and coupled with precipitation, the underlying surface, and other factors to form a unique mountain circulation system and climate characteristics, which in turn enhanced the spatial and temporal heterogeneity of the MEE. The results of this study will be useful in the further analysis of the causes of MEE and its ecological effects.

Key words： mass elevation effect; spatial downscaling; GeoDetector; GWR; Tianshan Mountains

Cite this article

ZHANG Mingyu , ZHANG Zhengyong , LIU Lin , ZHANG Xueying , KANG Ziwei , CHEN Hongjin , GAO Yu , WANG Tongxia , YU Fengchen . Spatio-temporal pattern and attribution analysis of the mass elevation effect in the Tianshan Mountains in China[J]. Journal of Geographical Sciences, 2023 , 33(10) : 2031 -2051 . DOI: 10.1007/s11442-023-2164-0

1 Introduction

The mass elevation effect (MEE) is a phenomenon that causes higher internal than external temperatures at the same elevation in mountain ranges due to the uplift of air masses over a mountainous land surface. The hydrothermal spatial heterogeneity caused by the MEE is one of the classic elements of physical geography research (Fang et al., 2004; Zhang and Yao, 2015; Hinojosa et al., 2016). The warming effect caused by mountain uplift is widespread, with the MEE changing the geo-ecological pattern at regional and mountain system scales (Grubb et al., 1966) and affecting the distribution of biodiversity (Leakey et al., 1987). In 1904, the German scientist Quervain (Quervain, 1904) was the first to record the gradual increase of natural vertical zone boundaries from the exterior to the interior of mountain ranges during an ecogeographic study of the Alps. Early studies of MEE were mainly qualitative comparisons of the vegetation distribution boundaries inside and outside different mountain ranges, and therefore the mechanisms involved in their formation still need to be clarified. With the development of remote sensing technology, mountain base elevation (MBE), as the primary influencing factor of MEE, has been widely used in the construction of quantitative models of MEE at global and local scales (Zhao et al., 2015; Zhang and Yao, 2016). The resulting models have been applied to studies of MEE at the mountain system/plateau scale, together with hygrometric continentality and latitude factors, in areas such as the Tibetan Plateau, Andes, Alps, Colorado Rocky Mountains, and New Zealand mountains (Han et al., 2010; Zhang et al., 2015; Wang et al., 2017a). These factors were found to explain up to 60% of the MEE at global and regional scales (Zhang et al., 2016; Liu et al., 2020), accelerating the quantitative study of MEE. However, the factors studied so far only contribute to MEE, and there is an urgent need to mechanistically determine the spatial heterogeneity and genetic complexity of the effect.
The heterogeneity of MEE is related to the spatial scale. As the research scale advances to the local level, the intensity and genetic mechanism of the MEE become more complex (Li et al., 2022). Due to the apparent spatial differentiation of the mountain scale structure, climate, and underlying surface properties, the factors discussed above will cause a diversity of MEE distribution patterns and regional formation mechanisms (Navarro et al., 2005). The MEE is the climate effect caused by uplifted air masses on mountains and their surrounding environment, and studies of the climate change response to uplifted air masses in elevated terrain are also essential (Barry et al., 2008; Wang et al., 2017b). In addition to the MBE, the spatial heterogeneity of the topography is a critical factor in the temperature difference between the internal and external areas of mountainous areas (Shao et al., 2012; Kattel et al., 2013; He et al., 2016). Additionally, the reflectance of solar radiation from different substrates such as grassland, forest, bare rock, and glaciers varies greatly, and all meteorological elements participate in the energy exchange between the land surface and the free atmosphere (Zhou et al., 1999). Most previous studies of the MEE have examined the characteristics of the effects of surface types on mountain systems by constructing linear regression models among variables. However, these global regression models can neither clarify the contributions of driving factors to the spatial heterogeneity of MEE, nor express the joint control of multiple factors. Due to the significance of the MEE and the many factors that influence its distribution pattern, it is vital to develop suitable methods to quantitatively analyze the spatial heterogeneity of its drivers and the synergistic or antagonistic effects among them to reveal the differentiation pattern and causal mechanisms of the effect. GeoDetector (Wang and Xu, 2017), a new tool for measuring and characterizing spatial heterogeneity, can objectively reflect the priority of each driver and the impacts of different factors in geographic phenomena. Geographically weighted regression (GWR) models (Brunsdon et al., 1996; Gao et al., 2019) determine the local spatial regression of the relationships between independent and dependent variables, thus revealing the spatial heterogeneity of the drivers (Zhao et al., 2019). The cores incorporated into these models have complementary advantages, which can be used to explore the response mechanisms of MEE to the drivers from multiple perspectives and expand the depth of local MEE research.
Most of the mountain systems where MEE studies have been conducted were located in low-latitude or monsoon climate zones, while the Tianshan Mountains, which has an inland and mid-latitude location, is not only one of the seven major mountain systems in the world but also the most representative large mountain ecosystem in the global temperate arid zone. The quantitative analysis of the intensity, patterns, and causes of MEEs in the Tianshan Mountains is of great significance for studying the characteristics and ecological effects of MEEs in arid areas. It will also enable a more in-depth and comprehensive investigation of the spatial differentiation and mechanism of MEEs at different scales and trends, and in sea and land locations. There have been several useful studies of temperature variations in the Tianshan Mountains in China. However, most of them only consider the peaks and valleys of the Tianshan Mountains as simple “cones,” which limits the characterization of the spatial variations of temperature in the internal and external areas of the mountains and different uplift areas. Therefore, this study adopted multi-source remote sensing data to quantitatively characterize the MEE and its variations in the Tianshan Mountains based on different spatial and temporal scales. The dominant driving factors were identified using a GeoDetector model (GDM), and the results were combined with a GWR model to explore the geospatial relationship between the spatial and temporal variations of the MEE and its influencing factors, enabling the evolution of mountain ecogeography in the arid zone to be explained in a multidimensional manner. This study is a useful supplement to the existing mountain climate and mountain science research.

2 Study area

Located in the hinterland of Eurasia, the Tianshan Mountains are a sizeable latitudinal mountain system sandwiched between huge deserts. They represent the dividing line of the Central Asian climate system and are the origin and watershed of China's inland water system (Zhao et al., 2014). The Tianshan Mountains consist of a substantial mountain chain largely composed of three parallel folded mountain ranges, the Southern, Central, and Northern Tianshan Mountains. These three ranges account for more than two-thirds of the total length of the mountain system (Figure 1) (Hu et al., 2004). The Khantengri Mountains is the highest mountain system in the Southern Tianshan Mountains, and is divided into the South Tianshan Mountains and the Harketawu Mountains. The Central Tianshan Mountains slope gently from west to east, and include the Wusun Mountains. The Central Tianshan Mountains incorporate many mountain ranges that reach 4000 m above sea level, including the Usun Mountains, Narathi Mountains, and Erbin Mountains, with the highest peak at the Erbin Mountains (43.53°N, 81.00°E, 4835 m). The western section of the Northern Tianshan Mountains contains the Bolokonu Mountains and Yilianhabierga Mountains, while the eastern section contains the Bogda Mountains and the Balikun Mountains. The Tianshan Mountain chain in China is studded with numerous intermontane basins, such as those at 42.90°N and 86.00°E, which pass through the Bayanbuluk Basin from west to east and north to south, respectively. Influenced by its unique landform, mountainous terrain, and airflow, the Tianshan Mountains have a continental climate, with significant differences in winter and summer temperature and vertical temperature (Zhou et al., 1998; Qin et al., 2018). The relative height difference of about 4000 m provides a vast space for the vertical differentiation of the geographic environment, resulting in apparent differences in the structure of vertical natural zones in the north-south and east-west directions. Coupled with the differentiation of the water and heat conditions this has resulted in the development of a distinct vertical ecosystem in the mountains, including deserts, mountain forests/grasslands, alpine meadows, and extreme alpine glaciers/snow belts (Zhang et al., 2004).
Figure 1 Overview of the Tianshan Mountains in China

3 Data and methods

This study used a spatial downscaling and accuracy evaluation of Tianshan Mountain temperature data in China based on multi-source remote sensing data. Mathematical statistics and geographic information system (GIS) spatial analysis techniques were used to estimate and analyze the spatial pattern of the MEE in the study area and to investigate the causal mechanisms of spatial and temporal variations in the MEE in the Tianshan Mountains, together with GDM and GWR analyses.

3.1 Data resources and preprocessing

3.1.1 Remote sensing data

The remote sensing data used in the study were land surface temperature (LST), precipitation, digital elevation model (DEM), the normalized difference vegetation index (NDVI), air pressure, humidity, and wind speed. The LST was obtained from MOD11C3 daytime and nighttime data with a spatial resolution of 0.05°, and was downloaded from NASA (https://www.nasa.gov/). After multiple requester terminal (MRT) batch processing, the monthly mean LST data were synthesized by the ArcGIS software using the arithmetic mean method. The spatial resolution of the DEM data was 30 m, and was obtained from the geospatial data cloud (http://www.gscloud.cn/). The spatial resolutions of precipitation, NDVI, air pressure, humidity, and wind speed were all 1 km. The NDVI and pressure data were sourced from the Data Center for Resources and Environmental Sciences, Chinese Academy of Sciences (https://www.resdc.cn/). The rest of the data were obtained from the National Earth System Science Data Center (http://www.geodata.cn/data/). The data spanned the period of 2000-2020, and for raster calculations and statistical analysis, all data were unified to the WGS-84 coordinate system and UTM projection.

3.1.2 Observed data

The field observations were mainly meteorological station data downloaded from the National Meteorological Information Centre (http://data.cma.cn/). Considering the spatial match with other data in terms of geographical location and elevation, the measured data from 23 meteorological stations in and around the Tianshan Mountains were selected. Data from 2000-2015 were used to drive the temperature inversion model, while the data from 2016-2020 were used to test the inversion accuracy of remote sensing data.

3.1.3 Terrain factor extraction

The topographic factors considered in this study were elevation, slope, slope direction, terrain index, terrain roughness, and the internal and external degree, which were obtained with ArcGIS software using DEM data based on published methods and formulas (He et al., 2016; Wu et al., 2018) (Table 1).
Table 1 Terrain parameters used in a temperature inversion
Topographic Factors Description Method/Formula
Elevation Mountain uplift degree Acquired from DEM data
Slope Degree of surface tilt Obtained using the ArcGIS slope tool
Aspect Terrain slope orientation $\text{Trasp}=1-\cos \left[ \pi \left( aspect-20 \right)/180 \right]$
Terrain index Synthesis of spatially differentiated features of the terrain T=$\text{log}\left[ \left( \frac{E}{{\bar{E}}}+1 \right)\times \left( \frac{S}{{\bar{S}}}+1 \right) \right]$
Terrain roughness Ground roughness $R=1/\cos (s)$
Internal and external degree Distance from the edge of large terrain ArcGIS distance analysis

Note: Trasp is the slope conversion index, aspect is the slope direction; T is the terrain index, E and$\bar{E}$represent elevation and mean elevation, respectively; S and$\bar{S}$represent slope and mean slope, respectively; R is the terrain roughness.

3.2 Methods

3.2.1 Construction of temperature downscaling models

Previous studies have shown a significant correlation between LST and air temperature (Williamson et al., 2013). Physical and environmental factors influence the energy exchange process between the near-surface and the free atmosphere (Prihodko et al., 1997). The scarcity of meteorological stations is not conducive to the spatial representation of temperatures in mountainous areas. Remote sensing data can solve the problem of spatial continuity, but its resolution limits the precise and objective assessment of the temperature distribution. Therefore, MOD11C3 LST data, which are highly correlated with air temperature, were used as the primary data for the air temperature inversion (Zhao et al., 2020). After considering previous studies and the typical conditions that characterize the geographical location, terrain, and environment of mountainous areas, the longitude, latitude, altitude, slope, pressure, humidity, wind speed, and NDVI were selected to construct a spatial downscaling regression model of air temperature. A combination of the coefficient of determination (R2), root-mean-square error (RMSE), and bias (Bias) was used to assess the consistency of the temperature data obtained from the downscaled inversion with the measured temperature data (Yu et al., 2020).
$y=a{{x}_{1}}+b{{x}_{2}}+c{{x}_{3}}+d{{x}_{4}}+e{{x}_{5}}+f{{x}_{6}}+g{{x}_{7}}+h{{x}_{8}}+j{{x}_{9}}+\lambda $
where y is measured meteorological station data; λ is a constant; x1 is elevation; x2 is slope; x3 is longitude; x4 is latitude; x5 is ground temperature; x6 is barometric pressure; x7 is wind speed; x8 is humidity; x9 is NDVI; and a, b, c, d, e, f, g, h, and j are regression coefficients for the independent variables, respectively.
To achieve the precise expression of regional temperature, the raster data needed for downscaling should be extracted and derived in a follow-up study. First, the spatial resolution of the DEM (30 m) was defined as high resolution (HR) and the spatial resolutions of the LST (0.05°) and meteorological factor (1 km) were defined as low resolution (LR). Then the meteorological, geographical, and topographical factors corresponding to the meteorological stations at the LR scale were extracted. A temperature downscaling regression model was constructed, and factor regression values were obtained. The regression coefficients obtained in the LR were substituted into the HR factor data, and a high-resolution Tianshan temperature database was constructed with ArcGIS.

3.2.2 Estimation of the MEE

The MEE has previously been quantified by the temperature difference between the internal and external areas of mountainous areas at the same altitude (Yao et al., 2013). However, because most peaks in the Tianshan Mountains are surrounded by vast deserts, the temperature in the outer desert Gobi region is much higher than in the mountainous areas, which also have taller vegetation cover. It is not appropriate to use the temperature difference between the internal and external areas of mountainous areas at the same latitude to characterize the strength of the MEE. Additionally, the interior of the mountain system is distributed over a large area of “depressions.” Therefore, detailed measurements of the macroscopic MEE were obtained by comparing the temperature differences between the uplifted air and the other regions of the mountains, and between the inner and outer edges of the mountain ranges at the same altitude. In this study, with reference to the methods used in a previous study (Barry et al., 2008; Jobbágy et al., 2000), the temperature in the study area was uniformly converted to the that at average height of the Tianshan Mountains (2216 m).
${{T}_{ah}}={{T}_{a}}+\left( E-H \right)\times \partial $
where Tah is the air temperature at elevation h; Ta is the temperature obtained by inversion; E is the elevation of the current location; H is the average elevation of the study area; $\partial $ is the air temperature lapse rate (0.6℃/100 m in this study).
To demonstrate the fine-scale differences in the warming amplitude of each uplifted area of the Tianshan mountainous region as a whole, four profiles were selected at 42.90°N, 43.53°N, 81°E, and 86°E, which run through the typical geomorphological units of the study area. All profiles were specifically analyzed to show the MEE trend with latitude and longitude and its correlation with the degree of mountainous uplift and the temperature difference between the internal and external areas of the mountains.

3.2.3 GeoDetector

An indicator system was constructed consisting of three categories of primary indicators and eight secondary factors (Table 2). The GDM was used to explore the causal patterns of the spatial differentiation of the MEE at different time scales in the Tianshan Mountains. The factor detection module was used to test the explanatory power of each factor on the spatial and temporal differentiation patterns of the MEE, and the interaction detection module identified both the coupling patterns among the elements and the intensity of their influence on the spatial differentiation of MEE. The formula was as follows:
$q=1-\frac{\mathop{\sum }_{h=1}^{L}{{N}_{h}}{{\sigma }_{h}}^{2}}{N{{\sigma }^{2}}}=1-\frac{SSW}{SST}$
where q denotes the explanatory power of the independent variable for the spatial heterogeneity of the MEE, the larger the value, the stronger the explanatory power; h = 1, 2, 3, …; L refers to the number of classifications after discretization of the driving factors in the text; Nh and N are the numbers of cells in the stratum and the whole area respectively;$\sigma _{h}^{2}$and σ2 are the variances of Y values in the stratum h and the entire area, respectively; and SSW and SST are the within-layer and total conflicts of the whole region, respectively.
Table 2 Factors influencing the spatial differentiation of the MEE in the Tianshan Mountains
Factor type Topographic and locational factors Climate factor Underlying surface properties
Specific indicators Slope (X1) Aspect (X2) Internal and external degree (X3) Terrain index (X4) Terrain roughness (X5) Absolute elevation (X6) Precipitation (X7) NDVI (X8)

3.2.4 The GWR

The GWR is a spatial regression model based on local smoothness. Its local regression coefficients were obtained according to different geospatial division units, which can effectively estimate data with spatial autocorrelation and reflect the spatial heterogeneity of the driving factors in other regions (Han et al., 2020; Wu et al., 2022). This study used a GWR to determine the influence of the main drivers within each geospatial unit and the differences in the direction and intensity of their effects. The equation was as follows:
${{y}_{i}}={{\beta }_{0}}\left( {{u}_{i}},{{v}_{i}} \right)+\underset{k=1}{\overset{p}{\mathop \sum }}\,{{\beta }_{k}}\left( {{u}_{i}},{{v}_{i}} \right){{x}_{ik}}+{{\varepsilon }_{i}}, i=1,2,3,...,n$
where yi is the value of the dependent variable of raster i; β0 is the intercept; (ui, vi) is the constant term of the raster i; βk(ui, vi) is the coefficient of the k independent variable of raster i; xik is the k independent variable of raster i; and εi is the random error.

4 Results

4.1 Spatial and temporal patterns of the air temperature in the Tianshan Mountains

The structural features of the Tianshan Mountains, such as its significant topography and alternating peaks and valleys, result in a diverse and complex temperature pattern. In the study area (Figure 2), the annual mean temperature was 2.6℃, 11.5℃ in summer and -6.2℃ in winter. From the perspectives of longitude, latitude, altitude, and inner and external degree, it was found that the spatio-temporal distribution of temperature in the Tianshan Mountains presented a circular distribution pattern of highest in the east, declining to the west, low in the middle, high at the edge, and low in the interior, with a “V” shaped distribution from the exterior to interior regions. There was a wide range of low temperature areas in the vicinity of the large glacier of Yilenhaberga Mountain and the second mountain junction of the Tianshan Mountains, which was distributed around 84°E, resulting in opposing temperature distribution patterns in the east and west sections of this node in winter and summer (Figure 3a). There was a notable latitudinal zonality of temperature in the study area(Figure 3b). The air temperature lapse rate at different latitudes differed due to the influence of the mountainous terrain, with a trend of increasing temperature north of 42°N in the winter half of the year.
Figure 2 Spatial distribution of the annual, and summer and winter halves of year mean air temperatures (Ta) in the Tianshan Mountains
Figure 3 The variation of air temperature (Ta) and actual air temperature at the same altitude in the Tianshan Mountains
Elevation plays a crucial role in climate change in mountainous areas. Due to the differences in humidity and radiation exposure throughout the year, the air temperature lapse rate was largest in the summer half of the year. At the same time, in the winter half of the year, it was only 0.55℃/100 m (Figure 3c). The cooling rate from the outer boundary to the inner region reached a maximum of 1.39℃/10 km in the summer half of the year and then decreased to about 1.03℃/10 km in the winter half of the year. In the winter half of the year in the Tianshan Mountains, there is a widespread temperature inversion phenomenon that has a high intensity, high frequency, and an extensive depth of up to 2000 m (Hu et al., 2004). This not only caused a low air temperature lapse rate, but also a warming trend in the mountainous region 100 km from the border (Figure 3d). Overall, the spatial distribution of temperature in the Tianshan Mountains had prominent zonal characteristics. In contrast, the temperature difference between the internal and external areas of the mountains during the summer half of the year was more significant than in the winter half of the year.

4.2 Spatial and temporal patterns of the MEE in the Tianshan Mountains

4.2.1 Macro-pattern analysis

There are seasonal patterns in climate change and underlying surface properties (such as vegetation and snow cover) in the Tianshan Mountains (Zhang et al., 2018). It is necessary to explore the spatial differentiation of the MEE based on the same elevation temperature patterns at different time scales. Because the Tianshan mountain system is a combination of many mountain ranges, to identify the MEE of separate uplifted landmasses in a more specific way, the geomorphological characteristics of the study area were classified into six major uplift blocks: the Southern Tianshan uplift block, the Harketawu uplift block, the Erbin uplift block, the Borokonu-Yilianhabierga-Alagou uplift block (Boyia uplift block), the Bogda uplift block, and the Balikun uplift block.
Based on the monthly air temperature at the same altitude, the spatial distribution of the annual, and winter and summer halves of the year mean air temperatures at the same height were obtained (Figure 2). Compared with the actual air temperature, there was a complex variation in the converted air temperature at the same altitude with longitude, while the zonal variation with latitude was more regular. The mountainous region was characterized by a general decrease in air temperature in the meridional direction (from west to east), with varying degrees of warming in local areas at various times (Figure 3e). A latitudinal zonality of temperature at the same altitude was evident, and the rate of decrease was often much less than the change in the actual temperature over the same period (Figure 3f). For each of the three periods of the year, there were varying degrees of warming at 39°−40°N, 42°N, and 44°-45°N, with these regions corresponding to the Southern Tianshan uplift block, the Harketawu uplift block, and the Alao Mountains, respectively. At the same time, the uplift blocks not only caused a slight increase in the trend in the air temperature lapse rate at the same altitude in the Tianshan Mountains (Figure 3g), but also reduced the temperature difference between the internal and external areas of the mountains caused by the differences in internal and external elevation (Figure 3h). The temperature changes in these uplift blocks were greater than in the Tianshan Mountains as a whole. For example, the change in the air temperature lapse rate at the same altitude in the summer half of the year was 0.15℃/100 m in the Balikun uplift block, five times higher than that of the Tianshan Mountains in the same period. The rate of temperature decline at the same altitude in the summer half of the year was only 0.096℃/10 km, while the temperature difference between the Harketawu Mountains and the Yilenhaberga Mountains in the same period was 1.5-2℃. The large number of glaciers distributed in the inner summit area caused the temperature at the same altitude to stop rising (Shen et al., 2013).
A further comparison of the temperature differences between the six uplift blocks and the same latitude at the same altitude (Table 3) revealed that two of the uplift blocks had no significant warming in the winter and summer halves of the year. However, the overall warming effect of the Tianshan Mountains was significant and universal, and spatial and temporal differences were apparent. The South Tianshan, the Erbin, and the Balikun uplift blocks were the most prominent tectonic units in the study area regarding the annual-scale mountain effect. At the same time, the MEE was not significant in the Harketawushan uplift block in the summer half of the year and the Bogda uplift block in the winter half of the year. The MEE decreased slightly and then increased from west to east, with the Southern Tianshan, Erbin and Balikun uplift blocks having the most significant MEE. The differences between the whole of the Tianshan Mountains and local regions confirmed that there were regional differences in the MEE.
Table 3 Statistics on the distribution of temperature at the same altitude in each uplift block in the Tianshan Mountains
Uplift block Mean altitude (m) Summer half of the year Winter half of the year
Mean Ta (℃) Difference with mean Ta at the
same latitude (℃)		 Mean Ta (℃) Difference with mean Ta at the same latitude (℃)
Southern Tianshan
uplift block		 3189 14.49 0.85 -0.70 1.72
Harketawu uplift block 3311 11.08 -0.31 -6.19 0.10
Erbin uplift block 3494 12.34 0.86 -5.04 1.33
Boyia uplift block 3135 10.80 0.29 -7.01 0.46
Bogda uplift block 3080 11.57 1.32 -9.81 -1.48
Balikun uplift block 3473 13.37 2.94 -6.64 1.21
In summary, there was an obvious zonality of temperature at the same altitude in the Tianshan Mountains and there were significant spatial and temporal differences. The continuity of large gullies, intermontane basins, and other geomorphological units weakened the zonality distribution, and the vertical variation of temperature at the same altitude showed a weak increasing trend.

4.2.2 Detailed analysis of the MEE characteristics

There was a clear spatial and temporal differentiation of the MEE in the Tianshan Mountains. At the same time, the warming magnitude of each uplift block differed from that of the mountainous area overall, and it was therefore necessary to further explore the fine-scale characteristics of the MEE in different geomorphic units and at different time scales (winter and summer halves of the year).
Compared with the Bayanbuluk Basin, the temperature increases in the winter and summer halves of the year at the same altitude in the Harketawu and Erbin uplift blocks, which were crossed by the 42.90°N profile (Figures 4a and 4b), were approximately 3.9 and 3.3℃, respectively. The Bogda uplift block, which was traversed by the 43.53°N (Figures 4c and 4d) profile, was approximately 0.2-2.4 and 0.2-2.6℃ warmer in winter and summer than the Yilianhabierga Mountains. The temperature at the same altitude in the Balikun uplift block was higher than in the other uplift blocks at different times of the year, most significantly during the summer half of the year. This was due to the negative feedback effect of the glacier and snow albedo in the alpine region of the Yilianhabierga Mountains, and resulted in a decrease in the warming effect of the mountain (Zhang et al., 2022).
Figure 4 The temperature profile at the same altitude for different regions of the Tianshan Mountains
Under the influence of the zonal gradient of solar radiation, the temperature at the same altitude in the study area was significantly spatially different. The temperature at the same height in the Harketawu uplift block located at 81°E (Figures 4e and 4f) increased from -8.4 and 8.1℃ in the southern outer margin in the winter and summer halves of the year, respectively, to -6.9 and 11.2℃ in the inner margin, and then fell to -9.9 and 8.8℃ in the northern outer margin, with the topographic uplift resulting in a temperature increase of1.4-5℃. The mean temperature at the same altitude in the Erbin uplift block in the southern part of 84.50°E (Figures 4g and 4h) was -5.7 and 13.1℃ in the winter and summer halves of the year, respectively, with a slight decrease in temperature in the central Bayanbuluk Basin, rising again to -6 and 12℃ in the northern Yilianhabierga uplift block, and then falling at the outer edge of the northernmost mountain to -12 and 8℃. The temperature increase inside and outside the profile area was 1.6-6.3℃. The more significant MEE in the Erbin uplift block was due to the combination of its lower latitude and the neighboring Gobi desert area to the south with its higher summer temperatures. This also indicated that the spatial differentiation of the MEE was more complex and was controlled by the underlying surface and local climate in the surrounding area.
The overall distribution of the MEE in the study area was high in the east and south, and low in the west and north. All uplift blocks displayed a warming trend from the external to internal areas of the mountains. The strength of the MEE in the uplift blocks differed, which may be related to the scale, structure, degree of uplift, geographical location, and underlying surface of the mountains.

4.3 Attribution of spatial variability to the MEE in the Tianshan Mountains

4.3.1 Attribution of spatial patterns to the MEE in mountainous regions

A previous study found that the MEE in the winter and summer halves of the year in the Tianshan Mountains were significantly different at the scale of the mountain system and uplift region. Therefore, the present study screened the possible influencing factors related to the distribution of the MEE, assessed the contribution of each element to the distribution of the MEE based on the q value using the GDM, and explored the interaction mechanism of each factor on the MEE. All eight factors selected (Table 2) passed the confidence test, and were considered to have a significant impact on the distribution of the MEE in the Tianshan Mountains.
The factor detection results showed differences in the intensity of each driving factor on the MEE in different periods, and the overall contribution was ranked as follows: topographic and locational factors > climatic factors > underlying surface properties. The explanatory power of topographic factors in the winter and summer halves of the year was 57% and 77%, respectively, with the internal and external degree playing a prominent role in the spatial differentiation of the MEE, especially in the summer half of the year with a single factor explanatory power of 17.40%. The absolute elevation directly represented the scale and degree of mountain uplift, with an explanatory power of 11.21% and 10.5% for the heterogeneity of MEE in the winter and summer halves of the year, respectively. Additionally, the terrain roughness and terrain index both characterized the mountain's local structural features, and the comprehensive contributions of the two factors to the differentiation of MEE in the winter and summer halves of the year were 13.95% and 23.72%, respectively. This was because the more significant the surface relief, the rougher the surface, and the greater the depth of cuttings, the more difficult it is for surface heat to exchange with the free atmosphere, which affects the magnitude of the MEE. Slope and aspect were the core factors describing the microtopographic scale, with combined contributions of 18.07% and 25.5% in the winter and summer halves of the year. They affect the MEE at the local scale by controlling the amount of solar radiation and the angle of the prevailing wind direction (Yeh et al., 1982), and the higher the latitude, the greater the influence of local topographic factors on solar radiation (Zeng et al., 2005). Precipitation was the primary factor influencing the spatial variation of the MEE in the winter half of the year, with an explanatory power of 24% and 12% in the winter and summer halves of the year, respectively. The NDVI can express the spatial heterogeneity of the underlying surface properties in mountainous areas. The contributions of the NDVI in the winter and summer halves of the year were 19.16% and 10.65%, respectively. The vegetation pattern and climate change resulting from long-term mutual influence and adaptation, and the seasonal shift of subsurface properties will cause changes in roughness, cloud cover, and wind speed. Furthermore, the processes of near-surface solar radiation, latent heat, and turbulent transport resulted in significant seasonal and spatial differences in the regional temperature distribution (Molnar et al., 1999; Chen et al., 2018; Wu et al., 2019), which in turn contributed to the complexity of the spatial pattern of the MEE in the Tianshan Mountains.
Previous studies of the MEE emphasized the fundamental role of uplift blocks on the warming effect (Kitoh et al., 1997), and it is also necessary to consider how various factors such as terrain, climate, and the underlying surface jointly control the distribution of the MEE. The results of an interaction detection (Figure 5) showed that the explanatory ability of the combination of multiple driving factors is more important than that of a single element, and the heterogeneity of the MEE can be enhanced through the two-factor nonlinearity among factors. The interaction between topography and climate in both the winter and summer half of the year had the greatest explanatory power for the spatial heterogeneity of the MEE. The interaction between absolute elevation and precipitation had the greatest explanatory power for the spatial distribution of MEE, with contribution rates of 0.33 and 0.44 in the winter and summer halves of the year, respectively, indicating that the spatial differences in precipitation within the same elevation zone significantly enhanced the spatial heterogeneity of the MEE. The next most important interactions were those between absolute elevation and internal and external degree, and precipitation and NDVI The absolute elevation had little influence on the single-factor sounding. The interaction between precipitation and NDVI was prominent due to the coordination and superposition of the absolute elevation and the two factors jointly formed unique mountain circulation systems and climate characteristics, which explained the MEE's local spatial and temporal heterogeneity and complexity. It was further confirmed that the degree of mountain uplift played a decisive role in the warming effect of the mountains.
Figure 5 The results of an interaction detection showing the spatial differentiation of the drivers of temperature at the same altitude in the Tianshan Mountains
In general, there were differences in the contributions and dominant types of single factors and between the contributions of factors at different periods. The factor detection results indicated that topographic and locational factors had the most significant influence on the spatio-temporal variation of the MEE, and the dominant roles of the internal and external degree and elevation were decisive. The interaction between topographic and climate factors was prominent, and the interactions between absolute elevation and the internal and external degree, and precipitation and NDVI were particularly evident in the spatial and temporal divergence of the MEE.

4.3.2 Attribution of spatial heterogeneity in the MEE across the uplift blocks

Based on the GDM results, driving factors such as internal and external degree, absolute elevation, precipitation, and NDVI were selected to explain the spatial heterogeneity of MEE in the study area. Using the GWR model, the spatial differences in the direction and intensity of the main driving factors were identified.
The GWR correlation coefficients indicated that the drivers were all spatially non-stationary, with varying degrees of variability and different characteristics (Table 4). The effect of absolute elevation (Figure 6b) on the MEE was significantly spatially non-stationary (winter: -12.11 to 45.17, summer: -21.75 to 29.81), with an overall predominantly positive effect. The absolute elevation effectively contributed to the warming of the uplift block, with the MEE in the Boyia and Erbin uplift blocks having the greatest positive feedback from the absolute elevation. The influence of precipitation (Figure 6c) and the NDVI (Figure 6d) on the MEE in the winter and summer halves of the year was dominated by negative feedback, which was more prominent for precipitation (-11.26) than for the NDVI (-0.96). Generally, the lower the precipitation, the stronger the MEE, and the latent heat of condensation released by precipitation from topographically induced cumulus clouds also contributed to the uplift block's heating effect. The scarcity of rainfall in the central and eastern parts of the study areas played a positive role in the warming of the uplift blocks, with the most substantial negative feedback in the Balikun uplift block, and the weak MEE in the west was related to the humid climate of the region.
Table 4 Statistical details of the main driving factors in the uplift blocks in the Tianshan Mountains
Period Summer half of the year Winter half of the year
Driving factors Internal and external
degree		 Absolute elevation Precipitation NDVI Internal and external degree Absolute elevation Precipitation NDVI
Tianshan Mountains 3.82 11.90 -7.73 -0.47 -2.99 6.39 -11.26 -0.72
Southern Tianshan uplift block 15.01 -11.56 13.76 -0.79 0.21 2.71 0.38 -0.35
Harketawu uplift block 0.93 10.34 -8.21 -2.25 -2.93 3.02 -6.61 -0.59
Erbinin uplift block -1.21 26.16 -17.58 -0.48 -5.15 11.90 -11.63 0.12
Boyia uplift block -0.25 21.76 -13.59 0.43 -2.60 12.22 -15.22 1.21
Bogda uplift block 1.46 23.34 -17.45 -0.06 -3.59 5.19 -22.15 -2.50
Balikun uplift block 2.99 22.72 -21.31 -0.15 -4.22 5.93 -25.28 -3.74
Figure 6 The impact of the main driving factors on the spatial differentiation of temperature patterns at the same altitude in the Tianshan Mountains
Overall, the NDVI had a weak negative feedback effect, which was consistent with the conclusion that the lower the NDVI, the more significant the MEE. The negative feedback of the NDVI on the MEE was more effective in the summer half of the year in the Harketawu uplift block, and the control of the NDVI was most decisive in the winter half of the year in the Bogda and Balikun uplift blocks. The influence of internal and external degree on the MEE varied (Figure 6a), with the summer half of the year dominated by a positive influence, while in the winter half of the year the influence was mainly negative. The MEE in the Southern Tianshan Mountains uplift block was clearly affected by the internal and external degree. The mountain interior became drier due to the inaccessibility of the humid airflow, and the lower latitude received more solar radiation resulting in a strong MEE. Overall, the higher absolute elevation, lower precipitation, and sparser vegetation cover in the central and eastern uplift blocks favored mountain warming, while warming in the southern uplift block was more closely related to the internal and external degree.

5 Discussion

5.1 Comparison of the MEE between the Tianshan Mountains and other mountain ranges

The scale/structure, geographic location, and mountain base elevation of the many mountains around the globe result in different strengths and weaknesses of thermal effects within each mountain range. To better understand the systematicity and completeness of the MEE, published results for representative mountain ranges worldwide were compared with the characteristics of the MEE in the Tianshan Mountains identified in this study.
Among the world's mountainous areas the “third pole,” i.e., the Tibetan Plateau, has the strongest MEE, with a maximum temperature difference of 10℃ between the exterior and interior of the mountain range (Yao et al., 2013). The Qinling Mountains (Liu et al., 2020), Andes (He et al., 2016), Alps (Zhang et al., 2015), and Rocky Mountains (Wang et al., 2017a) have the next most significant increases. The maximum increase in the Tianshan uplift region was about 7°C in the winter half of the year and 5℃ in the summer half of the year. The MEE in the Tianshan Mountains is therefore relatively weak. The width and average elevation of the Tianshan Mountains are more significant than those of the Alps, resulting in a smaller MEE. The intensity of the MEE is consistent with the distribution of mountain base heights. The base height of the Tibetan Plateau (Han et al., 2014) ranges from 2000 to 5000 m, with the western base where the MEE is strongest being highest at 5000 m. The base heights of the Qinling Mountains (Liu et al., 2018) and the Andes (Zhang et al., 2015) range from 147 to 4000 m, and for the Alps and the Rocky Mountains (Wang et al., 2017a) the range is from 6 to 2342 m, whereas the base height of the Tianshan Mountains (Han et al., 2014) is significantly lower (0−2000 m). Additionally, the land and sea position, structure, trend, prevailing wind direction, and micro-topography of mountain ranges also affects regional atmospheric circulation and forms the local climate. These energy exchange processes will affect the macro and local patterns of the MEE. Under the influence of southern westerlies, the western Bolivarian Plateau in the middle of the Andes, with a north-south trend, has a strong continentality (Arias et al., 2021), and therefore its MEE is the most significant of the world's major mountain ranges. The Rocky Mountains, which also run north-south, block the moist air brought by the prevailing westerly winds, and the topographic features of the staggered mountain ranges and huge ravines, together with the continental climate, lead to the warming of the mountain range (Wang et al., 2017a). The strength of the MEE is also related to hygrometric continentality. As a mid-latitude mountain range with an east-west trend and deep inland location, the Tianshan Mountains have a mild and humid climate with many basins and valley floors. The north slope of the Tianshan Mountains receives more precipitation from the cold and wet air from the Arctic Ocean, which has led to north-south differences in the strength of its MEE. All the mountain ranges considered had a significant MEE, but differed in their respective sizes, orientations, geographic locations, atmospheric circulation, and geographic factors at all scales.

5.2 Intensity of human activity intensity and its characterization

The mountainous land-temperature process is affected by the combination of multiple elements, but only some meteorological and topographical factors, such as barometric pressure, wind speed and humidity, were considered in this study. Other factors that were not considered could also influence mountain climate characteristics (Zhang et al., 2014). For example, aspect plays a decisive role in the distribution of solar radiant energy in mountainous areas. It significantly impacts the heat, wind speed, and precipitation distribution in mountain ranges (Wei et al., 2017). Topographic relief and topographic shading strongly influence the regional temperature and solar radiation divergence (Weng et al., 1990). From the surface radiation budget and energy balance perspective, surface albedo is crucial. At the same time, changes in subsurface conditions, precipitation, glacier and snow cover will affect the surface flux, and thus change the regional climate (Tian et al., 2001; Lu et al., 2022). Solar radiation, which is the primary energy source on the Earth's surface, is also spatially and temporally heterogeneous in complex terrain (Chen et al., 2014). In complex topographic areas, the topographic shadow is an important factor that directly and indirectly affects solar radiation (Zhang et al., 2019). Additionally, cloudy weather generates significant anisotropic radiation (Zhang et al., 2015). From an energy balance perspective, carefully considering topographic and environmental factors and their coupling effects on radiation warming in mountainous areas may lead to a more direct understanding of the MEE. From this perspective, there were subjective and objective limitations in selecting factors when exploring the spatial differentiation of MEE in this study. In addition to selecting elements, it was necessary to incorporate the differences in the effect scale of the different factors influencing the ground temperature transformation to improve the accuracy of air temperature inversions. The MEE of the Tianshan Mountains was investigated based on inversion temperatures, and the main driving factors affecting the MEE were determined, which enabled the spatial non-stationarity of the dominant warming factors in different uplift blocks to be simulated. However, the main driving factors of MEE vary at different scales, and spatial scale changes may change the correlation and intensity between the magnitude of mountain warming and the drivers, coupled with the fact that data from some factors are difficult to obtain and quantify. Therefore, future studies must combine multi-temporal scales and multi-source data to comprehensively investigate the spatial differentiation of the MEE.

6 Conclusions

This study constructed a mountain temperature inversion model based on multi-source remote sensing data and measured meteorological data, analyzed the spatial and temporal divergence characteristics of the MEE in Tianshan Mountains, and used the GDM and a GWR model to identify the driving mechanisms of the spatial divergence of MEE at different spatial and temporal scales. The following conclusions were obtained.
(1) The temperature pattern in the Tianshan Mountains is complex and diverse, with an overall circular distribution pattern of the highest temperatures in the east, declining to the west, low in the middle, high at the edges, and low in the interior. The temperature was negatively correlated with altitude and internal and external degree, with a maximum air temperature lapse rate of 0.63℃/100 m in the summer half of the year and a cooling rate 1.39℃/10 km from the periphery to the internal region. In the winter half of the year, a widespread inversion led to a lapse rate of only 0.55℃/100 m, and the cooling rate from the periphery to the internal region decreased to 1.03℃/10 km.
(2) The warming effect of the mountains in the study area was widespread and significant. The warming effect of the uplift blocks was stronger than the overall mean value of the Tianshan Mountains, with the most prominent warming occurring in the southern Tianshan Mountains, the Erbin uplift block, and the Balikun uplift block. The intervention of geomorphological units such as continuous large gullies and mountain basins slowed the rate of latitudinal regression in parts of the uplift. It weakened the altitudinal dependence of mountain temperatures. From west to east, the temperature increases from the external to internal areas of the mountains in the winter and summer halves of the year were 0.2-3.9 and 0.2-3.3℃, respectively. From north to south, the temperature increased by 1.4-6.3℃ from the external to internal areas of the mountains in the winter and summer halves of the year, and the warming effect was significantly enhanced from the edge of the mountain range to the mountainous interior.
(3) The spatial and temporal variability of the MEE in the Tianshan Mountains was dominated by topographic and locational factors, among which the internal and external degree and elevation were the key factors affecting the variability of the MEE. The interaction between factors had a more significant influence on the spatial differentiation of MEE than any single factor. Topography and climate were the main interaction types, and absolute elevation and precipitation was the dominant combination that enhanced the spatial heterogeneity of MEE in the winter and summer halves of the year.
(4) There was evident spatial heterogeneity in the direction and intensity of the driving factors of the spatial variation of MEE in the Tianshan Mountains. The absolute elevation was significantly and positively correlated with changes in the MEE, while precipitation and NDVI dominated the negative feedback. Topography had the greatest macroscopic control on the MEE. Coupled with precipitation, the underlying surface, and other factors it jointly formed a unique mountain circulation system and local climate characteristics, thus enhancing the spatial and temporal heterogeneity of the MEE in the study area.

Acknowledgements

We are grateful to the Geospatial Data Cloud (http://www.gscloud.cn/), Data Center for Resources and Environmental Sciences, Chinese Academy of Sciences (RESDC) (http://www.resdc.cn), National Earth System Science Data Center (http://www.geodata.cn/data/), National Meteorological Information Centre (http://data.cma.cn/).

Conflict of interest statement

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

References

Publishing order | Descend order by publishing year | Descend order by cited within
[1]
Arias P A, Garreaud R, Poveda G et al., 2021. Hydroclimate of the Andes Part (II): Hydroclimate variability and sub-continental patterns. Frontiers in Earth Science, 8: 505467.

DOI

[2]
Barry R G, 2008. Mountain Weather and Climate. New York: Cambridge University Press.

[3]
Brunsdon C, Fotheringham A S, Charlton M E, 1996. Geographically weighted regression: A method for exploring spatial nonstationarity. Geographical Analysis, 28(4): 281-298.

DOI

[4]
Chen D T, Huang F R, Li Q et al., 2018. Spatial variation of humidity and its influencing factors in the north and south slopes of the Tianshan Mountains, China during 1966-2015. Climate Change Research, 14(6): 562-572. (in Chinese)

[5]
Chen M, Zhuang Q L, He Y J, 2014. An efficient method of estimating downward solar radiation based on the MODIS observations for the use of land surface modeling. Remote Sensing, 6(8): 7136-7157.

DOI

[6]
Fang J Y, Shen Z H, Cui H T, 2004. Ecological characteristics of mountains and research issues of mountain ecology. Biodiversity Science, 12(1): 10-19. (in Chinese)

DOI

[7]
Gao J B, Jiao K W, Wu S H, 2019. Revealing the climatic impacts on spatial heterogeneity of NDVI in China during 1982-2013. Acta Geographica Sinica, 74(3): 534-543. (in Chinese)

DOI

[8]
Grubb P J, Whitmore T C, 1966. A comparison of montane and lowland rain forest in Ecuador: II. The climate and its effects on the distribution and physiognomy of the forests. The Journal of Ecology, 54(2): 303-333.

DOI

[9]
Han F, Zhang B P, Tan J et al., 2010. The effect of mountain base elevation on the altitude of timberline in the southeastern Eurasia: A study on the quantification of mass elevation effect. Acta Geographica Sinica, 65(7): 781-788. (in Chinese)

DOI

[10]
Han F, Zhang B P, Tan J et al., 2014. The effect of mountain basal elevation on the distribution of snowline with different mountain basal elevations in Tibetan Plateau and its surrounding areas. Geographical Research, 33(1): 23-30. (in Chinese)

DOI

[11]
Han J, Rui Y, Yang K et al., 2020. Quantitative attribution of national key town layout based on geodetector and the geographically weighted regression model. Progress in Geography, 39(10): 1687-1697. (in Chinese)

DOI

[12]
He W H, 2016. Mass elevation effect simulation based on mountain belts in the Tibetan Plateau and the central Andes[D]. Beijing: University of Chinese Academy of Sciences. (in Chinese)

[13]
Hinojosa L, Napoléone C, Moulery M et al., 2016. The “mountain effect” in the abandonment of grasslands: Insights from the French Southern Alps. Agriculture, Ecosystems & Environment, 221: 115-124.

DOI

[14]
Hu R J, 2004. Physical Geography of the Tianshan Mountains in China. Beijing: China Environmental Science Press. (in Chinese)

[15]
Jobbágy E G, Jackson R B, 2000. Global controls of forest line elevation in the northern and southern hemispheres. Global Ecology and Biogeography, 9(3): 253-268.

DOI

[16]
Kattel D B, Yao T, Yang K et al., 2013. Temperature lapse rate in complex mountain terrain on the southern slope of the central Himalayas. Theoretical and Applied Climatology, 113: 671-682.

DOI

[17]
Kitoh A, 1997. Mountain uplift and surface temperature changes. Geophysical Research Letters, 24(2): 185-188.

DOI

[18]
Leakey R J G, Proctor J, 1987. Invertebrates in the litter and soil at a range of altitudes on Gunung Silam, a small ultrabasic mountain in Sabah. Journal of Tropical Ecology, 3(2): 119-129.

DOI

[19]
Li W Y, Lan X C, Tang J L et al., 2022. Influence of albedo and evapotranspiration on the mass elevation effect in the Qinling-Daba Mountains of China. Mountain Research, 40(4): 614-625. (in Chinese)

[20]
Liu J J, Pan Z W, Qin F et al., 2020. Estimation of air temperature based on MODIS and analysis of mass elevation effect in the Qinling-Daba Mountains. Geographical Research, 39(3): 735-748. (in Chinese)

DOI

[21]
Liu J J, Qin F, Zhao F et al., 2018. Extraction and distribution of mountain base elevation in the Qinling-Daba Mountains. Journal of Geo-information Science, 20(10): 1457-1466. (in Chinese)

[22]
Lu Y B, Wang L C, Niu Z G et al., 2022. Variations of land surface albedo and its influencing factors in China from 2000 to 2017. Geographical Research, 41(2): 562-579. (in Chinese)

DOI

[23]
Molnar P, Emanuel K A, 1999. Temperature profiles in radiative-convective equilibrium above surfaces at different heights. Journal of Geophysical Research: Atmospheres, 104(D20): 24265-24271.

DOI

[24]
Navarro G, Molina J A, Barra N D, 2005. Classification of the high-Andean Polylepis forests in Bolivia. Plant Ecology, 176(1): 113-130.

DOI

[25]
Prihodko L, Goward S N, 1997. Estimation of air temperature from remotely sensed surface observations. Remote Sensing of Environment, 60(3): 335-346.

DOI

[26]
Qin Y, Ding J L, Zhao Q D et al., 2018. Spatial-temporal variation of snow cover in the Tianshan Mountains from 2001 to 2015, and its relation to temperature and precipitation. Journal of Glaciology and Geocryology, 40(2): 249-260. (in Chinese)

[27]
Quervain A, 1904. Die Hebung der atmosphärischen lsothermenin der Schweizer Alpen und ihre Beziehung zu den Höhengrenzen. Gerlands Beitr age zur Geophys, 6: 481-533.

[28]
Shao J A, Li Y B, Ni J P, 2012. The characteristics of temperature variability with terrain, latitude and longitude in Sichuan-Chongqing region. Journal of Geographical Sciences, 22(2): 223-244.

DOI

[29]
Shen Y P, Su H C, Wang G Y et al., 2013. The responses of glaciers and snow cover to climate change in Xinjiang (I): Hydrological effect. Journal of Glaciology and Geocryology, 35(3): 513-527. (in Chinese)

[30]
Tian Y Q, Davies-Colley R J, Gong P et al., 2001. Estimating solar radiation on slopes of arbitrary aspect. Agricultural and Forest Meteorology, 109(1): 67-74.

DOI

[31]
Wang J, Zhang B P, Zhang W J et al., 2017a. Quantitative research of mass elevation effect in Colorado Rocky Mountains. Geographical Research, 36(8): 1467-1477. (in Chinese)

[32]
Wang J F, Xu C D, 2017. Geodetector: Principle and Prospective. Acta Geographica Sinica, 72(1): 116-134. (in Chinese)

DOI

[33]
Wang Y X, Ding K, Li M B et al., 2017b. Spatial distribution modeling of temperature increase for the uplifted mountain terrains and its characteristics in Southwest China. Journal of Mountain Science, 14(11): 2270-2283.

DOI

[34]
Wei S L, Chen Z B, Chen Z Q et al., 2017. Simulation of the total solar radiation over micro-landform and correlation between the solar radiation and the land surface temperature. Remote Sensing for Land & Resources, 29(1): 129-135. (in Chinese).

[35]
Weng D M, Luo Z X, 1990. Topographical Climate in Mountainous Areas. Beijing: China Meteorological Press, 1990. (in Chinese)

[36]
Williamson S N, Hik D S, Gamon J A et al., 2013. Evaluating cloud contamination in clear-sky MODIS terra daytime land surface temperatures using ground-based meteorology station observations. Journal of Climate, 26(5): 1551-1560.

DOI

[37]
Wu A B, Qin Y J, Zhao Y X, 2018. Terrain composite index and its application in terrain gradient effect analysis of land use change: A case study of Taihang hilly areas. Geography and Geo-Information Science, 34(6): 93-99, 118. (in Chinese)

[38]
Wu C Y, Cao G C, Chen K L et al., 2022. Spatio-temporal variation in soil conservation service and its influencing factors in the upper reaches of the Yellow River. Journal of Soil and Water Conservation, 36(4): 143-150. (in Chinese)

[39]
Wu P F, Zhang J Y, Tan Jiao, 2019. Temporal and spatial variation of sunshine duration and related driving forces in Tianshan Mountain during 1961-2015. Meteorological Science and Technology, 2019, 47(3): 450-459. (in Chinese)

[40]
Yao Y H, Zhang B P, 2013. MODIS-based estimation of air temperature and heating-up effect of the Tibetan Plateau. Acta Geographica Sinica, 68(1): 95-107. (in Chinese)

DOI

[41]
Yeh T C, 1982. Some aspects of the thermal influences of the Qinghai-Tibetan Plateau on the atmospheric circulation. Archives for Meteorology, Geophysics, and Bioclimatology, Series A, 31(3): 205-220.

DOI

[42]
Yu H Z, Li L J, Li J Y, 2020. Establishment of comprehensive drought monitoring model based on downscaling TRMM and MODIS data. Journal of Natural Resources, 35(10): 2553-2568. (in Chinese)

DOI

[43]
Zeng Y, Qiu X F, Liu C M et al., 2005. Distributed modelling of direct solar radiation of rugged terrain over the Yellow River Basin. Acta Geographica Sinica, 60(4): 680-688. (in Chinese)

DOI

[44]
Zhang B P, Tan Y, Mo S G, 2004. Digital spectrum and analysis of altitudinal belts in the Tianshan Mountains. Journal of Mountain Science, 22(2): 184-192. (in Chinese)

[45]
Zhang B P, Yao Y H, 2015. Studies on Mass Elevation Effect. Beijing: China Environmental Science Press. (in Chinese)

[46]
Zhang B P, Yao Y H, 2016. Implications of mass elevation effect for the altitudinal patterns of global ecology. Journal of Geographical Sciences, 26(7): 871-877.

DOI

[47]
Zhang H B, Immerzeel W W, Zhang F et al., 2022. Snow cover persistence reverses the altitudinal patterns of warming above and below 5000 m on the Tibetan Plateau. Science of the Total Environment, 2022, 803: 149889.

DOI

[48]
Zhang L W, Huang J F, Wang X Z, 2014. A review on air temperature estimation by satellite thermal infrared remote sensing. Journal of Natural Resources, 29(3): 540-552. (in Chinese)

[49]
Zhang S, 2015. Quantitative study of mass elevation effect based on its main forming-factors[D]. Beijing: University of Chinese Academy of Sciences. (in Chinese)

[50]
Zhang S H, Li X G, She J F et al., 2019. Assimilating remote sensing data into GIS-based all sky solar radiation modeling for mountain terrain. Remote Sensing of Environment, 231: 111239.

DOI

[51]
Zhang S, Zhang B P, Yao Y H et al., 2016. Magnitude and forming factors of mass elevation effect on Qinghai-Tibet Plateau. Chinese Geographical Science, 26: 745-754.

DOI

[52]
Zhang Y L, Li X, Bai Y L, 2015. An integrated approach to estimate shortwave solar radiation on clear-sky days in rugged terrain using MODIS atmospheric products. Solar Energy, 113: 347-357.

DOI

[53]
Zhang Z Y, 2018. Modeling hydrological processes in main runoff generating area of Manasi River Basin, Xinjiang[D]. Shihezi: Shihezi University.

[54]
Zhao Fang, 2015. A Quantitative study of the mass mountain effect based on the theoretical and real distribution of mountain altitudinal belts[D]. Beijing: University of Chinese Academy of Sciences. (in Chinese)

[55]
Zhao Gang, 2014. The study of glacier changes in typical regions, China Tianshan Mountains[D]. Lanzhou: Lanzhou University. (in Chinese)

[56]
Zhao G N, Zhang Z Y, Liu Lin et al., 2020. Changes of glacier mass balance in Manas river basin based on multi-source remote sensing data. Acta Geographica Sinica, 75(1): 98-112. (in Chinese)

DOI

[57]
Zhao M S, Liu B Y, Lu H L et al., 2019. Spatial modeling of soil organic matter over low relief areas based on geographically weighted regression. Transactions of the Chinese Society of Agricultural Engineering, 35(20): 102-110. (in Chinese)

[58]
Zhou G S, Wang Y H, 1999. The feedback of land use/cover change on climate. Journal of Natural Resources, 14(4): 318-322. (in Chinese)

DOI

[59]
Zhou X, Chen D J, 1998. Study on vertical change features of climate in the southern of Tianshan Mountains. Mountain Research, 16(1): 47-52. (in Chinese)

Options
Outlines

/

〈 〉