EN中文

Journal of Resources and Ecology >

2019 , Vol. 10 >Issue 3: 296 - 306

DOI: https://doi.org/10.5814/j.issn.1674-764X.2019.03.008

Forest Ecosystem

Interdependent Dynamics of LAI-ET across Roofing Landscapes: the Mongolian and Tibetan Plateaus

  • TIAN Li , 1, 2, *
Expand
  • 1.Qianyanzhou Ecological Research Station, Key Laboratory of Ecosystem Network Observation and Modeling, Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences, Beijing 100101, China
  • 2.Zhongke-Ji’an Institute of Eco-Environmental Sciences, Ji’an, Jiangxi 343000, China;
*Corresponding author: Tian Li, E-mail:

Received date: 2018-10-22

  Accepted date: 2019-01-10

  Online published: 2019-05-30

Supported by

National Key Research and Development Program of China (2017YFB0503005)

The Strategic Priority Research Program of the Chinese Academy of Sciences (XDA19050501, XDA19040305)

National Natural Science Foundation of China (41601100)

The International Postdoctoral Exchange Fellowship Program 2015 by the Office of China Postdoctoral Council (No. 38 Document of OCPC, 2015).

Copyright

All rights reserved
Fold

Abstract

The Mongolian and Tibetan Plateaus have experienced warming higher than the global average in recent decades, resulting in many significant changes to ecosystem structures and functions. Among items that show change are the leaf area index (LAI) and evapotranspiration (ET), both of which play a fundamental role in shaping many causes and consequences of land surface processes and climate. This study examines the spatiotemporal changes of the LAI and ET and their relationships on these two roofing landscapes. Based on the MODIS products from 2000 through 2014, we found that there existed a general positive relationship between LAI and ET on the Mongolia Plateau (MP), while synergy did not exist on the Tibetan Plateau (TP). Overall, 49.38% (50.62%) of land areas on the TP experienced significant increases (decreases) in LAI, while on the MP the percentages of increase and decrease were 94.92% (5.09%). For ET, the increased land area was 21.70% (124100 km2) on the TP and 88.01% (341600 km2) on the MP. More importantly, the relationships varied substantially across space and over time, with mismatches found in some parts of the landscapes. Additional observational investigations and/or experimental research are needed to explore the relationships, including the influences of vegetation characteristics and disturbances.

Key words： ET; LAI; alpine; Eurasian continent; altitude and latitude; land surface process

Cite this article

TIAN Li . Interdependent Dynamics of LAI-ET across Roofing Landscapes: the Mongolian and Tibetan Plateaus[J]. Journal of Resources and Ecology, 2019 , 10(3) : 296 -306 . DOI: 10.5814/j.issn.1674-764X.2019.03.008

1 Introduction

Vegetation green-up from climate change across terrestrial surfaces has been widely recognized as a sound indicator of fundamental changes in ecosystem functioning and a source of feedback to regional and global climates (Bonan et al., 1992; Davin and de Noblet-Ducoudre 2010; Myers-Smith et al., 2015). This change is particularly pronounced in high latitude and altitude regions (IPCC, 2014)(IPCC, 2014 #18809). A simultaneous change under the warming trend is that of local-to-regional land use producing very different land covers. These changes in climate and land cover lead to further, dramatically different land-surface properties (e.g., in surface temperature, moisture, evapotranspiration (ET), and energy balances) that drive key ecosystem processes, dynamics, and functions (e.g., greenhouse gas emission), as well as to exchanges in materials and energy between the land surface and the atmosphere (Bala et al., 2007; Lee et al., 2011; Pielke, 2005; You et al., 2017). For example, altered spatiotemporal distributions in ET have been used to explain significant increases in warming potential on both regional and global scales (Loranty et al., 2014).
ET, the sum of soil evaporation, canopy evaporation, and plant transpiration, plays a key role in planetary hydrologic and energetic cycles (Oki and Kanae 2006). It regulates water and heat balance through gas and water exchanges between land surfaces and the atmosphere. ET also is a central process in the climate system and a nexus of the water, energy, and carbon cycles (Jung et al., 2010; Mu et al., 2007). It is directly altered by different land-cover types (e.g., forests vs. croplands), canopy characteristics (e.g., leaf area and distribution, foliar nitrogen (N), and moisture), and lengths of growing season (i.e., duration of green leaves), all of which result in different amounts of water being returned to the atmosphere (Planque et al., 2017). ET is also indirectly altered because of different microclimates, especially soil and surface temperatures that emit different amounts of latent and sensible heat. Recent studies have indicated a link between positive ET and increased cooling of the atmosphere; this link is associated with changes in leaf area and temperature differences between the land and atmosphere (Jassal et al., 2009; Shen et al., 2015; Tong et al., 2017). It appears that the changes in leaf area and its spatiotemporal distributions, as well as the warming/cooling of the atmosphere are the key factors required for understanding the significance of ET in time and space, and for understanding its role in key biophysical processes (e.g., thawing of frozen soil, deforestation, vegetation cover change) needed to model ecosystem dynamics at landscape, regional, and global scales (Planque et al., 2017).
Both modeling and observational studies suggest that climate warming will likely enhance vegetation growth in northern terrestrial ecosystems(Keeling, et al. 1996; Myneni, et al. 1995; Nemani, et al. 2003) (Keeling et al., 1996; Miao et al., 2017; Myneni et al.,1995; Nemani et al., 2003; Peng et al., 2013). Several recent investigations have provided further evidence of rapid changes in high altitude and high latitude areas in the northern hemisphere (Li et al., 2017a; Li et al., 2017b; Loranty et al., 2014; Peng et al., 2013; Shen et al., 2015; Tian et al., 2014). The Mongolian Plateau (MP), with its high latitude (37.7-53.3°N), and the Tibetan Plateau (TP), with its high altitude (>4000 m), have witnessed a rapid and high magnitude of changes in vegetation and climate (Lee et al., 2011). They are also the two main grassland- cover plateaus (the TP was 63% grassland and MP was 65%). In this study, we focused on the spatiotemporal changes of the relationships between ET and leaf area index (LAI) on these two roofing landscapes of the Eurasian continent and in the TP during the growing season (May-September).
The data required for ET and LAI can be obtained from different sources and approaches. Because our interest lies in understanding the long term situation on two large plateaus, data products from satellite remote sensing technology are of the most value for us to achieve our study objectives. The ET products from the Moderate Resolution Imaging Spectroradiometer satellite have a long, stable history and have been widely used to study land-cover changes, regional ecosystem functions, and to design models that track regional climate changes and human activities (Choi et al., 2017; Li et al., 2017a; Palmer et al., 2015; Paruelo et al., 2016; Shen et al., 2015; Wang et al., 2014). Specifically, we used the MODIS gap-filled snow-free product (MOD16A2 (V005) and LAI/fPAR product (MCD15A2) (Friedl et al., 2002; Wang et al., 2014) to address the following pressing questions:
(1) Have the changes in ET, LAI, and their independent relationships over the past 15 years been similar on the MT and the TP?
(2) Are there any significant inter- and intra-annual differences in ET and LAI?
(3) Are the changes spatially homogeneous within each plateau? If not, how much land area has experienced significant (vs. insignificant) changes? Where are the areas of significant change located within each plateau?
(4) Are the ET-LAI relationships similar for the two plateaus?
Both plateaus have experienced warming that is higher than the global average (John et al., 2009; Shen et al., 2015). In the case of the MP, this is due to its geospatial location in arid and semi-arid areas that have over 90% of rainfall returned to the atmosphere by ET (Cleugh et al., 2007). However, the proportion in Alpine districts may vary, especially on the TP, which is at high elevations and has high levels of glacier and permafrost coverage. Both plateaus have also experienced increased land use from grazing, urbanization, and agricultural expansion (Enkhtur et al., 2017; John et al., 2009). However, the intensity and types of land use are different across the two plateaus and have occurred at different time periods (Chen et al., 2013). Here, we hypothesize that the spatiotemporal changes in ET and LAI in recent decades are due to differences in warming trends and contrasting climates, vegetation, and historical land use. As a result, there may be “hot spots” on both plateaus where the rate of change in either ET or LAI is substantially higher, or lower, than the average. Temporally, we hypothesize that the long-term changes in annual means are comprised of uneven monthly means, with such contributions varying by year and between the two plateaus.

2 Materials and methods

2.1 Study area

The MP (87.8-126.0°E, 37.7-53.3°N) covers 2.75×106 km2, and it is part of the Central Asian Plateau. It has elevations of 1000-1500 m (Fig. 1a) and straddles two territorial jurisdictions, the country of Mongolia in the north and the Inner Mongolia Autonomous Region of China in the south. Three of the four major biomes (i.e., steppe, desert, and forest) was distributed largely along the precipitation and temperature gradient, while cropland is the exception (John et al., 2013). The steppe occupies almost 65% of the total land area (Fig. 1a). The perennial green-up onset falls between mid- and late April, and reaches maturity from early July to the end of August. In September and October, the vegetation gradually drops its leaves (i.e., there is vegetation senescence). The climate is predominantly semi-arid continental, with an annual mean temperature ranging from -1.7 °C in the meadow steppe to 5.6 °C in the desert steppe, with annual mean precipitation of 90-433 mm (John et al., 2013). Most precipitation falls from June through August (Lu et al., 2009).
Fig. 1 (a) The spatial distributions of land cover types on the two plateaus; (b) The average annual LAI during 2000-2014; (c) The average annual ET during 2000-2014; (d-e) The probability density function (PDF) for LAI and ET.
The TP (26.5-39.5°N, 78.3-103°E) covers 2.57×106 km2, extends from sub-tropical to mid-latitude regions, and has elevations of 4000-8800 m (Fig. 1a). The mean temperatures in the coldest and the warmest months are approximately -10 °C and 10 °C, respectively (Zhang et al., 1982). Temperatures and precipitation have distinct decreasing gradients from the southeast to the northwest, with the majority of rain falling in July-August. Three main types of grasslands (alpine meadow, alpine steppe, and alpine desert steppe) cover 63% of the plateau (Fig. 1a) and are distributed roughly from the southeast to the northwest along the precipitation and temperature gradients. Sparse vegetation also grows in the arid, northwestern mountains, with low canopy coverage. The growing season begins between late April and mid-May and continues until late September to early October (Zhang et al., 2013).
This study is focused on the changes in grasslands that cover >60% of the plateaus. This was necessary because extremely high or low values of ET and LAI (e.g., in glaciers in the TP and forests in the MP) have the potential to mask the mean values.

2.2 MODIS data

From the land products of MODIS onboard NASA’s Earth Observing System’s satellite Terra (Friedl et al., 2002; Zhang et al., 2013), we employed version 005(v005) MOD16A2 eight-day Global 1km SIN grid spatial resolution land ET and MCD15A2 eight-day LAI/FPAR products with a 500× 500 m spatial resolution to analyze the spatial-temporal characteristics and factors driving land evapotranspiration during 2000-2014. For these years, we used data from May through September (i.e., five months per year).
The two (ET and LAI) products were further improved by filtering out noises resulting from cloud contamination and topographic differences, using the subset tool to restrict the analysis. Altogether, we included a total of 20 images produced between days of year (DOY) 121 and 273 (1 May to 30 September). Our initial examinations of the relationship between ET and LAI in the growing season were calculated using average monthly values—for example, May was calculated using the sum values of DOY 121, 129, 137, and 145 for each pixel—and then, for the annual value in the growing season, the annual growing season ET and LAI were calculated using average monthly values with the following formula:
${{M}_{annual}}=\underset{i=1}{\overset{5}{\mathop \sum }}\,{{M}_{month,i}}$
where${{M}_{annual}}$was the ET and LAI annual mean value of every year, and${{M}_{month,i}}$was the monthly mean value for ET and LAI, with i representing the months of May through September.
To explore the interannual temporal and spatial variations of the ET and LAI values, we divided the growing season into one month segments and analyzed the empirical relationships between LAI and ET. For analysis of the correlation between ET and LAI, the products of ET were resampled to a spatial resolution of 500 m×500 m using the Python code of Arcpy Resample management to match with the LAI data.
Using ArcGIS 10.2, a remote sensing model was employed to retrieve the pixel values of ET and LAI, formatted as raster images, and these were then used to explore intra-annual dynamics and change trends. We calculated the Z-score normalization of ET and LAI to estimate the rate of change (i.e., the slope of a linear trend) for each pixel, and we also extracted the points with the significance level P<0.05. The total pixel numbers and values of ET and LAI were retrieved by their value sorts. Then, with the software R package, the probability density function (PDF) method was used to assign the pixel distribution pattern in value grades.
Taking an average of pixel values by year, we made a line regression and determined the coefficient variation (CV) of ET and LAI for the two plateaus to compare the inter-annual range ability, and we used the Person to monitor the otherness on the two plateaus. A correlation analysis was developed for ET and LAI. We used the boxplots to explain their stability in different LAI grades by extracting various points, which were selected based on the theory that the four adjacent point types are the same. We thus extracted 50 points of each class of LAI on each plateau.

3 Results

3.1 Spatiotemporal changes in ET and LAI

Grassland was found to be the major cover type among the six cover types on both plateaus, accounting for 65% and 63% of the MP and TP, respectively (Fig. 1a). On the TP, alpine meadow is distributed in the southeast region and the alpine steppe dominates the northwest (Fig. 1a), resulting in a gradual decrease in LAI from southeast to northwest (Fig. 1b) and synergetic the ET with LAI in spatial pattern (Fig. 1c). Moving from the southeast (Inner Mongolia) to the north (Mongolia) on the MP, the Eurasian steppe includes meadow, meadow steppe, typical steppe, desert steppe and desert, before it becomes exclusively the desert (Fig. 1a). Both LAI and Et decrease along this gradient (Fig. 1b-c).
The long-term mean (±std) of LAI is 0.728 (±0.63) on the MP and 0.698 (±0.71) on the TP, and the growing season (May to September) annual average of ET is 156 (±90) and 201 (±116), respectively, on the two plateaus. The PDF indicates that both LAI and ET are left-skewed, with TP showing higher skewness for LAI (Fig. 1b). The mode value of PDF (LAI) is clearly higher on the TP than on the MP, while the mode value of PDF (ET) is lower on the TP than on the MP. For LAI, there are more pixels falling at <0.36 on the TP than on the MP, but less pixels in the LAI range of 0.36-1.22. About 38.3% and 26.1% of the pixels for the MP and TP, respectively, fall in the Q1 of the PDF (LAI) (Fig. 1d), whereas 31.35% and 83.04% of pixels fall within the Q1 of PDF (ET) (Fig. 1e). For ET, more pixels fall within 83.04-153.22 range but less within 58.5-168 range on the TP than on the MP.
The long-term changes (i.e., slopes of the linear models) of annual and monthly LAI and ET during 2000-2014 showed great spatial variations on both plateaus (Fig. 2-3). For the annual change in LAI, the regions with increasing LAI on the TP were found mostly in the northeast, while decreases were found in the southeast (Fig. 2a). Overall, we found 49.29% (50.71%) of the TP land experienced a steady increasing or decreasing trend (Table 1). On the MP, however, LAI variations showed increases on the majority of the land area (93.18%), with only a few patches showing a decreasing trend (6.42%). In addition, the increase was greater in Mongolia than in Inner Mongolia. Interestingly, these long-term changes in annual LAI received varying contributions from the monthly changes (Fig. 2, Table 1). For the TP, the LAI increase in May of 61.86×103 km2 (95.13%) was more pronounced than that during June-September (27.77- 2.99%), whereas on the MP, the increasing trend was more pronounced during June, July, and August (97.77%, 93.83%, and 90.08%, respectively) and less so in May and September (both 87.57%) (Table 1).
Fig. 2 Spatial distributions of the change trends for LAI in the growing season (May-September); the inset map shows significant increases (blue) and decreases (red) (P<0.05).
Fig. 3 Spatial distributions of the change trends for ET in the growing season (May-September); the inset map shows significant increases (blue) and decreases (red) (P<0.05).
Table 1 The average annual and standard deviation (std) of LAI and ET during 2000-2014 by average annual LAI class on the Mongolian Plateau (MP) and the Tibetan Plateau (TP)
Var Loc Sig. Area Annual May June July August September
LAI TP Total (1000 km2) 57.41 65.03 38.39 47.92 48.00 42.96
Decreasing 29.06 3.16 6.53 19.99 18.55 31.03
(%) (50.62) (4.86) (17.01) (41.72) (38.65) (72.23)
Increasing 28.35 61.86 31.86 27.93 29.45 11.93
(%) (49.38) (95.13) (82.99) (58.28) (61.35) (27.77)
MP Total (1000 km2) 101.93 57.12 100.63 80.81 62.59 61.36
Decreasing 5.19 17.25 2.24 4.99 6.21 7.63
(%) (5.09) (30.20) (2.23) (6.17) (9.92) (12.43)
Increasing 96.75 39.87 98.39 75.82 56.38 53.73
(%) (94.92) (69.80) (97.77) (93.83) (90.08) (87.57)
ET TP Total (1000 km2) 571.90 196.90 478.20 476.80 333.90 265.70
Decreasing 447.80 154.30 332.40 278.70 295.40 250.10
(%) (78.30) (78.36) (69.51) (58.46) (88.47) (94.12)
Increasing 124.10 42.60 145.80 198.00 38.50 15.60
(%) (21.70) (21.64) (30.49) (41.54) (11.53) (5.88)
MP Total (1000 km2) 388.20 86.10 343.30 496.80 224.60 49.80
Decreasing 46.50 53.80 48.40 29.40 78.10 30.30
(%) (11.99) (62.42) (14.11) (5.92) (34.79) (60.82)
Increasing 341.60 32.40 294.80 467.40 146.40 19.50
(%) (88.01) (37.58) (85.89) (94.08) (65.21) (39.18)
Large spatial variations in the long-term change trends were detected for ET on both plateaus (Fig. 3). There were significant differences in annual ET between the two plateaus (Table 1), and the rate of change of ET seemed stronger on the TP than on the MP. The amount of land that experienced a significant decreasing trend on the TP was 78.30% (447.8×103 km2) for the year, and 78.36% (154.3×103 km2), 69.51% (332.4×103 km2), 58.46% (278.7×103 km2), 88.47% (295.4×103 km2) and 94.12% (250.1×103 km2) for the months of May through September (Table 1). In terms of spatial variations, the area that experienced ET decrease was focused in the central area, and there were scattered areas of decrease distributed in the southeast area (Fig. 2a). The months of June through September contributed the most to the annual ET, with decreasing areas distributed in the central and southeast areas only in May. The area of annual ET increase was focused in the northeast and the months of May, June, and August contributed most of the increase. The proportion of increase was highest in July (41.54%, 198× 103 km2) and lowest in September (15.6×103 km2, 5.88%).
The differences among the annual and monthly ET changes were not apparent on the MP. Widespread increases in ET appeared along the desert spreading to the north (Mongolia) and southeast (Inner Mongolia) (Fig. 3a), with decreases found only in a patch in the northwest (Mongolia) and a small area in the southeast (Inner Mongolia). Contributions to annual ET were made in June through August, with May and September the only months that did not show widespread increases to ET (Fig.3a-f). The amount of land that had a significant annual increase trend on the MP was 88.01% (341.6×103 km2). This trend was highest in July (94.08%, 467.4×103 km2) and lowest in September (39.18% 19.5×103 km2), and showed 37.58% (32.40×103 km2) in May, 85.89% (294.8×103 km2) in June, and 65.21% (146.4×103 km2) in August.

3.2 Interannual variations of LAI and ET

The normalized (i.e., Z-score) LAI and ET showed great interannual variations by year and month, as well between the two plateaus (Fig. 4 and Fig. 5). The LAI on the MP had a higher Z-score and variation than did that on the TP (CVMP = 0.091, 0.081, 0.157, 0.126, 0.108, and 0.192; CVTP = 0.049, 0.049, 0.119, 0.061, 0.083, and 0.121 [annual, May, June, July, August, and September, respectively]) (Fig. 4). Over the 15-year study period, there were six years when the Z-score was above or below the long-term annual LAI mean on the MP. There were also six years when the Z-score was above or below the long-term mean on the TP. The years on the TP were not the same as those on the MP (Fig. 4a). During the study years, LAI on the TP showed a decreasing trend in the annual, July, and September periods, with significant increases only detected in May (Slope 0.152, P=0.000); LAI on the MP increased annually and in all five months, especially annually and for June-August, when the trend reached significant levels (Slope 0.121, P=0.009; Slope 0.117, P= 0.013; Slope 0.105, P=0.029; Slope 0.082, P= 0.037; respectively) (Fig. 4a, c-e). When examined by month, the interannual variation (i.e., CV values) on the MP was higher than on the TP, with the long-term change of the monthly means also differing between the plateaus (Fig. 4b-f). Additionally, a large deviation in LAI was found annually on the MP for 2003, 2007, 2008, and 2012 and on the TP for 2000, 2006, and 2010. The differences for other months were not consistent on either plateau (Fig. 4b-f), such as in 2006 and 2010 for the TP and 2007 for the MP, all of which were extraordinary years for the LAI oscillations (Fig. 4a, d-e).
It is worth noting that 1) LAI was increasing on the MP, while the TP did not show synergy; 2) LAI on both plateaus was reduced in June 2001 and September 2014; 2) LAI in May-June 2014 on the MP was the highest; 3) LAI in June-August on the TP was low; and 4) LAI in June-August in 2007 was below average on the MP. The ET on the MP had a higher Z-score and variation than on the TP (CVMP = 0.111, 0.107, 0.146, 0.152, 0.126, and 0.121; CVTP = 0.063, 0.114, 0.079, 0.064, 0.090, and 0.086 (annual, May, June, July, August and September, respectively) (Fig. 5).
Fig. 4 The trends of the mean LAI on grasslands in the Mongolian and Tibetan plateaus
Fig. 5 The trends of the ET on the grasslands in the Mongolian and Tibetan plateaus
Only in May was the Z-score value on the TP higher than on the MP (Fig. 5b). Over the 15-year study period, there were five years when the Z-score was significantly above or below the long-term annual ET mean on the MP. There were also four years when the Z-score was above or below the long-term mean on the TP; again, these years were not all the same for the two study areas (Fig. 5a). During the study years, ET on the TP showed a decreasing trend in the annual, May, August, and September periods, with increase detected only in June and July (Slope 0.055 and 0.007, P=0.375, P=0.907, respectively). ET on the MP increased annually and in monthly periods, except in May (Slope -0.011, P=0.866), especially annually and in June and July, when the trend reached significant levels (Slope 0.096, P=0.02; Slope 0.102, P=0.038; Slope 0.127, P=0.027; respectively) (Fig. 5a, c-d).
When examined by month, the interannual variation (i.e., CV values) on the MP was higher than on the TP, with the long-term change of the monthly means also differing between the plateaus (Fig. 5b-f). Additionally, large annual deviations in ET on the MP were found for 2003, 2007, 2008 and 2012, and on the TP for 2000, 2006 and 2007. These differences for the months were not consistent on either plateau (Fig. 5b-f); in 2006, the large deviation showed in the annual, May, July, August, and September periods, and in June 2007, large deviations of ET were consistent for the two plateaus (Fig. 5c).
It is worth noting that 1) ET showed an increasing trend on the MP, while on the TP there was not synergy; 2) ET on both plateaus was reduced in June 2007 and September 2006; 3) ET in the annual, May, and July-September periods for 2006 on the TP was the lowest, and in 2000 was highest in both annual and monthly periods, except in May; 4) ET in the annual and June-July periods for 2007 on the MP were low; and 5) ET on the MP for the annual and all monthly periods was higher in 2003, and the annual and July-September periods were highest in 2012.

3.3 The interdependent dynamics of ET and LAI

We chose about 50 points in each stratified value of LAI data and then extracted the ET value of the points by reclassifying at 500 m by spatial homogeneity with LAI. We found that ET increased with LAI classes on both plateaus during 2000-2014, following apparent linear and logarithm trends for the MP and TP, respectively (Fig. 6). The increase in ET with LAI was not significant when LAI exceeded 1.5. The median ET for the five classes was 5.60, 11.63, 16.76, 21.74, 25.28, and 34.21 mm/mon on the TP (Fig. 6A), while it was 12.88, 24.47, 27.46, 29.74, 31.67, 33.27 mm/mon on the MP. Additionally, it appeared that the ET variation for any LAI class was higher on the TP than on the MP (Fig. 6B). The boxplot showed that: 1) the ET was sensitive to LAI increase on the MP, and 2) on the TP, the ET was stable when the LAI value was over 1.
Fig. 6 Boxplots of MODIS ET retrieval as a function of MODIS-estimated leaf area index (LAI) classed (0.5 LAI steps); stratified average data for the two plateaus

4 Discussion and conclusions

ET is mainly constrained by three factors: energy, water vapor transport conditions, and water supply capacity of the medium. Because of warming on the two plateaus, we know temperature is an important element representing energy supply, and that higher temperature results in a higher ratio of the energy used for ET coming from absorbed solar energy (Gu et al., 2008). As direct reflection of surface water supply capability, the more soil moisture, vegetation cover degree that more water is available for ET (French et al., 2012; Krishnaswamy et al., 2014). For other reasons, such as precipitation advance, there were twofold effects on ET. That is, although precipitation enhances surface water supply capability, increasing air humidity reduces ET at the same time (Brulebois et al., 2015; Penatti et al., 2015).
In the growing season, the vegetation cover degree is a strong control ET (Fatichi and Ivanov,2014;Garcia et al., 2014). We employed MODIS eight-day ET and LAI products from 2000 through 2014 during the growing season in two plateaus. The results indicate that the annual mean ET during the 15-year study period varied from -0.838 to +1.237 for MP and from +1.845 to +0.261 for TP (Fig. 5), pointing to a maximum variation of >20% during the study period for yearly and monthly ET on both plateaus. Our findings of intra-annual variations are similar to the magnitudes of those found in other research on the TP (Tasumi et al., 2014; Wang et al., 2018) and on the MP (Krishnan et al., 2012; Zhang et al., 2007). We know that in the growing season, the net solar radiation (like the vegetation cover degree) is a strong control of ET; however, just as rainy and cloudy weather affects the ET, soil and air humidity also affects it on the TP (Li et al., 2016). There is an issue of spatial mismatch between vegetation and ET (Fig. 2 and Fig. 3) These patterns we find are at magnitudes similar to those found by other researchers. However, on the MP the spatial response of LAI and ET appeared spatial cooperation, it cleared that hydrothermal condition and the vegetation distribution is spatial cooperation on MP. The results of the correlation analysis between ET and LAI showed that ET increased with LAI classes on both plateaus during 2000-2014, following an apparent linear and logarithm trend for the MP and TP, respectively (Fig. 6). The increase in ET with LAI was not significant when LAI exceeded 1.5. On the TP, the ET was stable when the LAI value was over 1. In addition, like in other studies, in the extreme years and months on the two plateaus, large deviations in ET and LAI were found (Chen et al., 2013).
We compared and quantified spatiotemporal changes in ET and LAI on the two roofing plateaus in Eurasia using the MODIS products from 2000 through 2014. The long-term changes (i.e., slopes of the linear models) of annual and monthly LAI and ET during 2000-2014 showed great spatial variations on both plateaus. ET and LAI increased on the MP, while there was not synergy on the TP. The ET was sensitive to LAI increase, and a synchronous relationship was found for the change in ET and LAI over time as well as across the MG landscape gradient. On the TP, although the ET and LAI showed decreasing trends, the spatial scale was a mismatch.
Overall, 49.29% of land areas for TP and 6.42% for MP experienced significant increases in LAI, while 50.71% and 93.18% of the land areas for TP and MP, respectively, decreased in LAI. The ET increased by 78.30% in TP and by 11.99% in MP; and decreased by 21.70% in TP and 88.01% in MP. More importantly, the relationships varied substantially across space and over time. Substantial areas were mismatched with respect to changes in LAI and ET. In the extreme environments, for ET and LAI on the MP, a large deviation showed in 2007 by year and month; on the TP, the ET showed an extreme decrease in 2006 by year and month, but the LAI did not synergize.

The authors have declared that no competing interests exist.

References

Publishing order | Descend order by publishing year | Descend order by cited within
[1]
Bala G, Caldeira K, Wickett M, et al.2007. Combined climate and carbon-cycle effects of large-scale deforestation.Proceedings of the National Academy of Sciences, USA, 104(16): 6550-6555.

[2]
Bonan G B, Pollard D, Thompson S L.1992. Effects of Boreal Forest Vegetation on Global Climate.Nature, 359(6397): 716-718.

[3]
Brulebois E, Castel T, Richard Y, et al.2015. Hydrological response to an abrupt shift in surface air temperature over France in 1987/88.Journal of Hydrology, 531: 892-901.

[4]
Chen J, Wan S, Henebry G, et al.2013. Dryland East Asia: Land Dynamics amid Social and Climate Change. Higher Education Press. (in Chinese)

[5]
Choi M, Mu Q Z, Kim H, et al.2017. Ecosystem-dynamics link to hydrologic variations for different land-cover types.Terrestrial Atmospheric and Oceanic Sciences, 28(3): 437-462.

[6]
Cleugh H A, Leuning R, Mu Q Z, et al.2007. Regional evaporation estimates from flux tower and MODIS satellite data.Remote Sensing of Environment, 106(3): 285-304.

[7]
Davin E L, de Noblet-Ducoudre N.2010. Climatic Impact of Global-Scale Deforestation: Radiative versus Nonradiative Processes.Journal of Climate, 23(1): 97-112.

[8]
Enkhtur K, Pfeiffer M, Lkhagva A, et al.2017. Response of moths (Lepidoptera: Heterocera) to livestock grazing in Mongolian rangelands.Ecological Indicators, 72: 667-674.

[9]
Fatichi S, Ivanov V Y.2014. Interannual variability of evapotranspiration and vegetation productivity.Water Resources Research, 50(4): 3275-3294.

[10]
French A N, Hunsaker D J, Clarke T R.2012. Forecasting spatially distributed cotton evapotranspiration by assimilating remotely sensed and ground-based observations.Journal of Irrigation and Drainage Engineering-ASCE, 138(11): 984-992.

[11]
Friedl M A, McIver D K, Hodges J C F, et al.2002. Global land cover mapping from MODIS: algorithms and early results.Remote Sensing of Environment, 83(1): 287-302.

[12]
Garcia M, Fernandez N, Villagarcia L, et al.2014. Accuracy of the Temperature-Vegetation Dryness Index using MODIS under water-limited vs. energy-limited evapotranspiration conditions.Remote Sensing of Environment, 149: 100-117.

[13]
Gu S, Tang Y H, Cui X Y, et al.2008. Characterizing evapotranspiration over a meadow ecosystem on the Qinghai-Tibetan Plateau.Journal of Geophysical Research,113. Doi: 10.1029/2007JD009173.

[14]
IPCC.2014. Climate Change 2014. Synthesis Report Summary Chapter for Policymakers.

[15]
Jassal R S, Black T A, Spittlehouse D L, et al.2009. Evapotranspiration and water use efficiency in different-aged Pacific Northwest Douglas- fir stands.Agricultural and Forest Meteorology, 149(6): 1168-1178.

[16]
John R, Chen J Q, Lu N, et al.2009. Land cover/land use change in semi-arid Inner Mongolia: 1992-2004. Environmental Research Letters, 4(4). Doi: 10.1088/1748-9326/4/4/045010

[17]
John R, Chen J Q, Ou-Yang Z T, et al.2013. Vegetation response to extreme climate events on the Mongolian Plateau from 2000 to 2010. Environmental Research Letters, 8(3). Doi:10.1088/1748-9326/8/3/035033

[18]
Jung M, Reichstein M, Ciais P, et al.2010. Recent decline in the global land evapotranspiration trend due to limited moisture supply.Nature, 467(7318): 951-954.

[19]
Keeling R F, Piper S C, Heimann M.1996. Global and hemispheric CO2 sinks deduced from changes in atmospheric O-2 concentration.Nature, 381(6579): 218-221.

[20]
Krishnan P, Meyers T P, Scott R L, et al.2012. Energy exchange and evapotranspiration over two temperate semi-arid grasslands in North America.Agricultural and Forest Meteorology, 153: 31-44.

[21]
Krishnaswamy J, John R, Joseph S.2014. Consistent response of vegetation dynamics to recent climate change in tropical mountain regions.Global Change Biology, 20(1): 203-215.

[22]
Lee X, Goulden M L, Hollinger D Y, et al.2011. Observed increase in local cooling effect of deforestation at higher latitudes.Nature, 479(7373): 384-387.

[23]
Li G, Zhang F M, Jing Y S, et al.2017. Response of evapotranspiration to changes in land use and land cover and climate in China during 2001-2013.Science of the Total Environment, 596: 256-265.

[24]
Li Q Y, Xu L, Pan X B, et al.2016. Modeling phenological responses of Inner Mongolia grassland species to regional climate change. Environmental Research Letters , 11(1). Doi:10.1088/1748-9326/11/1/015002

[25]
Li Z Y, Wu W Z, Liu X H, et al.2017b. Land use/cover change and regional climate change in an arid grassland ecosystem of Inner Mongolia, China.Ecological Modelling, 353: 86-94.

[26]
Loranty M M, Berner L T, Goetz S J, et al.2014. Vegetation controls on northern high latitude snow-albedo feedback: observations and CMIP5 model simulations.Global Change Biology, 20(2), 594-606.

[27]
Lu N, Wilske B, Ni J, et al.2009. Climate change in Inner Mongolia from 1955 to 2005-trends at regional, biome and local scales. Environmental Research Letters, 4(4). Doi:10.1088/1748-9326/4/4/045006

[28]
Miao L J, Muller D, Cui X F, et al.2017. Changes in vegetation phenology on the Mongolian Plateau and their climatic determinants. Plos One, 12(12). Doi: 10.1371/journal.pone.0190313

[29]
Mu Q, Heinsch F A, Zhao M, et al.2007. Development of a global evapotranspiration algorithm based on MODIS and global meteorology data.Remote Sensing of Environment, 111(4): 519-536.

[30]
Myers-Smith I H, Elmendorf S C, Beck P S A, et al.2015. Climate sensitivity of shrub growth across the tundra biome.Nature Climate Change, 5(9): 887-892.

[31]
Myneni R B, Hall F G, Sellers P J, et al.1995. The Interpretation of Spectral Vegetation Indexes.IEEE Transactions on Geoscience and Remote Sensing, 33(2): 481-486.

[32]
Nemani R R, Keeling C D, Hashimoto,Het al.2003. Climate-driven increases in global terrestrial net primary production from 1982 to 1999.Science, 300(5626): 1560-1563.

[33]
Ni J.2004. Estimating net primary productivity of grasslands from field biomass measurements in temperate northern China.Plant Ecology, 174(2): 217-234.

[34]
Oki T, Kanae S.2006. Global hydrological cycles and world water resources.Science, 313(5790): 1068-1072.

[35]
Palmer A R, Weideman C, Finca A, et al.2015. Modelling annual evapotranspiration in a semi-arid, African savanna: functional convergence theory, MODIS LAI and the Penman-Monteith equation. African Journal of Range & Forage Science, 32(1): 33-39.

[36]
Peng S S, Piao S L, Ciais P, et al.2013. Asymmetric effects of daytime and night-time warming on Northern Hemisphere vegetation.Nature. DOI: 10.1038/nature12434

[37]
Pielke R A.2005. Land use and climate change.Science, 310(5754): 1625-1626.

[38]
Planque C, Carrer D, Roujean J L.2017. Analysis of MODIS albedo changes over steady woody covers in France during the period of 2001-2013.Remote Sensing of Environment, 191(15): 13-29.

[39]
Shen M G, Piao S L, Jeong S J, et al.2015. Evaporative cooling over the Tibetan Plateau induced by vegetation growth.Proceedings of the National Academy of Sciences, USA, 112(30): 9299-9304.

[40]
Tasumi M, Hirakawa K, Hasegawa N, et al.2014. Application of MODIS Land Products to Assessment of Land Degradation of Alpine Rangeland in Northern India with Limited Ground-Based Information.Remote Sensing, 6: 9260-9276.

[41]
Tian L, Zhang Y J, Zhu J T.2014. Decreased surface albedo driven by denser vegetation on the Tibetan Plateau. Environmental Research Letters, 9(10). Doi: 10.1088/1748-9326/9/10/104001

[42]
Tong X J, Zhang J S, Meng P, Li J, et al.2017. Environmental controls of evapotranspiration in a mixed plantation in North China.International Journal of Biometeorology, 61(2): 227-238.

[43]
Wang W G, Li J X, Yu Z B, et al.2018. Satellite retrieval of actual evapotranspiration in the Tibetan Plateau: Components partitioning, multidecadal trends and dominated factors identifying.Journal of Hydrology, 559: 471-485.

[44]
Wang Z S, Schaaf C B, Strahler A H, et al.2014. Evaluation of MODIS albedo product (MCD43A) over grassland, agriculture and forest surface types during dormant and snow-covered periods.Remote Sensing of Environment, 140: 60-77.

[45]
You Q G, Xue X, Peng F, et al.2017. Surface water and heat exchange comparison between alpine meadow and bare land in a permafrost region of the Tibetan Plateau.Agricultural and Forest Meteorology, 232: 48-65.

[46]
Zhang G L, Zhang Y J, Dong J W, et al.2013. Green-up dates in the Tibetan Plateau have continuously advanced from 1982 to 2011. Proceedings of the National Academy of Sciences, USA, 110(11): 4309-4314.

[47]
Zhang R Z, Yang Q Y.1982.Physical Geography of Xizang. Beijing: Science Press .(in Chinese)

[48]
Zhang Y, Kadota T, Ohata T, et al.2007. Environmental controls on evapotranspiration from sparse grassland in Mongolia.Hydrological Processes, 21: 2016-2027.

Options
Outlines

/

〈 〉