Asymmetric seasonal daytime and nighttime warming and its effects on vegetation in the Loess Plateau

Liqun Ma; Fen Qin; Hao Wang; Yaochen Qin; Haoming Xia

doi:10.1371/journal.pone.0218480

Abstract

Over the period 1982–2015, temperatures have exhibited an asymmetric warming pattern diurnally, as well as seasonally across the Loess Plateau. However, very limited research has studied the implications and effects of such seasonally heterogeneous warming across the Loess Plateau. In this study, we also analyzed the time series trends and seasonal spatial patterns of the maximum (T_max) and minimum (T_min) temperatures and evaluated how different vegetation responded to daytime and nighttime warming in the Loess Plateau from 1982 to 2015 based on the NDVI and meteorological parameters (precipitation or temperature). We found that T_max and T_min significantly increased throughout the years except for T_max in autumn, and the diurnal asymmetric warming showed some striking seasonal differences. For example, the increasing rates of T_min in spring, summer, autumn, and winter were 0.75, 1.20, 1.88, and 1.10 times larger than that of T_max, respectively. NDVI showed significantly positive correlation with T_max and T_min in spring and winter, while NDVI presented significantly positive correlation with T_min in summer and T_max in autumn across entire Loess Plateau. Furthermore, we also discovered diverse seasonal responses in terms of vegetation types to daytime and nighttime warming. For instance, Spring NDVI showed significantly positive partial correlations with T_max and T_min. In summer, grasslands and wetlands merely displayed significantly positive partial correlations with T_min. Cultivated land presented significantly positive partial correlation between the NDVI and T_max (T_min) in autumn. In winter, cultivated land, forest, and grassland exhibited significantly positive partial correlation with T_max and T_min, while only wetland showed a significantly positive partial correlation with T_max. Our results demonstrated responses of vegetation to climate extremes and enhance a better understanding of the seasonally different responses of vegetation under global climate change at different scale.

Citation: Ma L, Qin F, Wang H, Qin Y, Xia H (2019) Asymmetric seasonal daytime and nighttime warming and its effects on vegetation in the Loess Plateau. PLoS ONE 14(6): e0218480. https://doi.org/10.1371/journal.pone.0218480

Editor: Sergio Rossi, Universite du Quebec a Chicoutimi, CANADA

Received: February 1, 2019; Accepted: June 3, 2019; Published: June 24, 2019

Copyright: © 2019 Ma et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: All relevant data are within the paper.

Funding: The Program for Key Scientific Research in the University of Henan Province (18a170002) is supported by the Education Department of Henan Province. The Open Fund of China Meteorological Administration and Henan province Key Laboratory of Agrometeorological Support and Applied Technique (AMF201809) is supported by the China Meteorological Administration and Henan province Key Laboratory of Agrometeorological Support and Applied Technique. The funders had roles in the study design, data collection of the manuscript.

Competing interests: The authors have declared that no competing interests exist.

1. Introduction

Vegetation is an important part of the global terrestrial ecosystem, playing as a pivotal medium in the flows of both matter and energy across the atmosphere, hydrosphere, biosphere, and pedosphere [1]. Vegetation growth is closely related to climatic factors and hydrothermal processes [2–3]. Temperature and precipitation are two dominant climatic factors that influence vegetation growth and production [2]. The relationship between vegetation growth and climatic factors typically varies with different vegetation types and different spatiotemporal scales [2–3]. However, most of the previous studies have focused on revealing the impact of the average state of climatic factors on vegetation growth [4–5]. The effects of seasonal extreme temperature change on vegetation are poorly understood.

Recently, a growing number of studies have shown that the global climate has shown asymmetric annual and seasonal daytime and nighttime warming [3,6–7]. The trend of the nighttime minimum temperature (T_min) was faster than that of the daytime maximum temperature (T_max) [6], which is known as asymmetric warming [3]. The increasing rate of minimum temperature was about twice of the maximum temperature over global land areas from 1950 [3]. In summer, the increasing rate of temperatures in the high latitudes of the northern hemisphere is faster than that in spring and autumn, and the seasonal temperature difference shows a downward trend [8]. Accordingly, the changing characteristics of global temperature rise and seasonal daytime and nighttime asymmetric warming have been proved to affect plant photosynthesis activity and yield differently [5,6–7]. The differences in the physiological and ecological characteristics of vegetation types lead to different responses to asymmetric warming [3]. Therefore, it is necessary to study the effects of seasonal daytime and nighttime asymmetric warming on vegetation dynamic changes.

Until now, previous studies have investigated the responses of the modes of the ecosystems to annual and seasonal daytime and nighttime asymmetric warming through ground control experiments, model simulations, and remote sensing data analysis [3, 6–7]. Dynamic responses of different vegetation to asymmetric warming are intriguing, but their explanations are still uncertain. For instance, some studies have showed that T_max is beneficial to most vegetation growth in wetter and colder regions but negatively correlates with normalized difference vegetation index (NDVI) in arid and semi-arid regions, while the increasing T_min inhibits the growth of vegetation [3]. On the contrary, Xia et al. [7] have reported that T_min increases positively correlate with vegetation in the high-cold steppe and meadow steppe zones, while T_max increases are significantly positively correlated with GIMMS3g NDVI in wetter and colder regions of the Tibetan Plateau. Using MODIS, SPOT NDVIs, and MODIS EVI, Shen et al. [9] found that maximum temperature increases in summer in the Tibetan Plateau are conducive to the growth of vegetation, but the temperature increases at night restrain vegetation growth.

The response of vegetation growth to asymmetric warming also shows a seasonal variation [7]. For instance, Piao et al. [10] found that net carbon uptake of northern ecosystems is being decreased in response to autumnal warming. Using GIMMS3g NDVI, Xu et al. [11] reported that the increasing of temperature in spring enhances vegetation growth in Central Eurasia after 2002, while in summer vegetation growth is mainly driven by precipitation. Many other studies have further shown that seasonal daytime and nighttime warming has different effects on vegetation [12–13]. Wu et al. [12] found that T_max is significantly correlated with the GIMMS3g NDVI in vast boreal forests of the Northern Hemisphere in spring, summer and autumn, but T_max is negatively related with the vegetation growth in temperate dry ecosystems in summer. The T_min is positively correlated with the NDVI in spring and summer, but T_min is negatively correlated with the NDVI in autumn, using NDVI data from 1982 to 2011 from the AVHRR. Using NDVI data from 1982 to 2011 obtained from the AVHRR, Tan et al. [13] showed that the T_max increases prevent the growth of vegetation in arid regions in summer, but is conducive to the growth of vegetation in cold regions in spring. The increasing of T_min is beneficial to vegetation growth in spring and summer but restrains most vegetation growth in autumn [13]. In summary, different intensity of responses to diurnal asymmetry warming has showed complex correlations due to vegetation types, regions, and seasons.

The Loess Plateau is located in the transition zone of semi-humid and semi-arid climates [14], and it is sensitive to climate change. This area has a fragile ecological system, containing a key soil and water conservation area in the middle and upper reaches of the Yellow River [15]. The different hydrothermal distributions in different seasons lead to diverse responses of vegetation across regions and seasons. At the same time, the special natural environment of the Loess Plateau is also the place that can be used to verify the relationships between vegetation and hydrothermal climate change at local scale.

The satellite-based NDVI, which has been used successfully in long-term monitoring of vegetation dynamics at regional and global scales, has proven to be a robust indicator of vegetation growth [16]. When NDVI, with different spatial and temporal resolutions from different platforms and sensors, is used to monitor the vegetation dynamics, the results may have some differences. Therefore, it is very important to select suitable NDVI for the given research purpose. The widely used NDVI including the AVHRR, MODIS, and SPOT VEGETATION has been proven to be a robust indicator of vegetation changes at regional and global scales for long-term temporal examinations [2–3]. In order to detect long-term vegetation dynamics, the third-generation Normalized Difference Vegetation Index (NDVI3g) dataset, which released by the NASA’s Global Inventory Modeling and Mapping Studies (GIMMS) group, is selected to monitor vegetation change. At the same time, NDVI3g data has been normalized to account for sensor calibration loss, orbital drift, and atmospheric effects and has been proved its good ability for the long-term monitoring of vegetation changes.

Using the GIMMS3g NDVI and meteorological (temperature and precipitation) data during 1982–2015, we use the Mann-Kendall non-parametric hypothesis tests and the second-order partial correlation analysis method to determine the time series trends and spatial patterns of T_max and T_min in different seasons on the Loess Plateau as well as the responses of different vegetation types to seasonal daytime and nighttime asymmetric warming. The results can provide an evaluation of changes in vegetation and ecological health in the Loess Plateau in response to climate change.

2. Materials and methods

2.1. Study area

The Loess Plateau lies in the upper and middle reaches of the Yellow River from 100°54′E to114°33′E and from 33°43′N to 41°16′N, covering an area of approximately 624,000 km² (Fig 1). From south to north, there are two heat zones: a warm temperate zone and medium temperature zone. From east to west, the Loess Plateau includes semi-humid, semi-arid, and arid areas. It has a typical continental monsoon climate. According to the meteorological stations in the area, the average annual temperature is between 9°C and 12°C during 1982–2015 [17]. The vegetation dynamic changes and the temperature differences in the east and west are large. The Loess Plateau is characterized by a rainy summer and autumn, with less rain in the winter and spring. Precipitation is mostly restricted from July to September, and the average annual precipitation varied from 100 to 800 mm from northwest to southeast in 1982–2015 (Fig 1).

Download:

Fig 1. The altitude, meteorological stations, and vegetation types across the Loess Plateau.

https://doi.org/10.1371/journal.pone.0218480.g001

2.2. Data and processing

2.2.1. Meteorological data and preprocessing.

We obtained the meteorological data (temperature and precipitation) from the China Meteorological Data Sharing Service System (http://cdc.cma.gov.cn/home.do) [18]. Data quality control is a prerequisite for index calculations, and we performed the data quality processes using the computer program RClimDex Software Version 1.1, which is developed and maintained by researchers at the Climate Research Branch of the Meteorological Service of Canada [19]. Using the detailed principles described by Wang [20], we conducted a homogeneity test with the RHtest V4 software to identify possible change points or structural changes in the data series. The only criterion for inclusion of a meteorological station was that its recording period must cover at least 85% of the total 34-year period. After data quality control and a homogeneity assessment, 34-year periodical records at 106 stations (Fig 1) were selected (from 1 January 1982 to 31 December 2015).

According to previous research, the spatial interpolation for temperature based on a DEM (digital elevation model) provided by the NASA Shuttle Radar Topographic Mission (SRTM) (http://www.glcf.umd.edu) can effectively simulate the spatial distribution of temperature [21]. This study used the DEM as a co-variable and employed thin plate smoothing splines to interpolate the temperature and precipitation data. The final interpolation outputs were 8 km×8 km grids of data.

According to a typical practice [22], one year was divided into four seasons: spring (March-May), summer (June-August), autumn (September-November), and winter (December-February).The T_max and T_min in different seasons were calculated based on the average of the daily maximum and minimum temperature in the same season over years, respectively. The precipitation was defined as the accumulated precipitation for the corresponding season.

2.2.2. GIMMS-NDVI3g and preprocessing.

The GIMMS-NDVI3g dataset is published by Global Inventory Modeling and Mapping Studies, which is based on a synthesis method using the maximum value with a time interval of 15 days, a spatial resolution of 8 km×8 km, and a time series covering the period 1982–2015 [23]. This dataset has been calibrated for radiation and atmospheric correction and coordinate transformation; images of the daily satellite tracks are available with geometric correction, cloud-removal, and removal of bad lines for NDVI calculation and data generation [23]. To further eliminate noise and clouds [23], we used the maximum synthesis method to synthesize monthly data by IDL. The pixels with an average season NDVI value larger than 0.1 were only considered, so that the influence of soil variation on the spectral signal in bare and sparsely vegetated zones could be minimized [24].

2.2.3. Vegetation data and preprocessing.

The GlobeLand30 data are freely available and included eight types of land cover including forests, cultivated land, grasslands, shrublands, bareland, wetland, water bodies and artificial surfaces across Loess Plateau [25]. We selected forests, cultivated land, grasslands, shrublands, wetland in this study. The vegetation data were resampled to match the spatial resolution of GIMMS NDVI3g data using a majority filter method.

2.3. Data analysis

2.3.1. Mann-Kendall.

The Mann-Kendall non-parametric hypothesis test method has advantage that samples do not have to conform to a certain distribution, and abnormal values little impact on the test results [26]. Therefore, we adopt Mann-Kendall to detect the change trend of T_max, T_min, and NDVI. The slope of T_max, T_min, and NDVI were calculated using linear regression at the regional and pixel scales, respectively.

2.3.2 Partial correlation analysis.

Partial correlation yields the correlation between two variables, controlling for a third or more variables [27]. The second-order partial correlation analysis was used to distinguish the influence of interannual changes in T_max and T_min on vegetation NDVI among different seasons. The partial correlation coefficient between NDVI and T_max in each season was calculated by controlling T_min and precipitation, and likewise, the partial correlation coefficient between the NDVI and T_min in each season was calculated by controlling T_max and precipitation.

3. Results and analysis

3.1. Spatiotemporal patterns of asymmetric warming and NDVI

3.1.1. The regional analysis.

T_max increased significantly in spring, summer, and winter (p<0.05) and non-significantly in autumn (p>0.05) during the 1982–2015 period (Table 1). T_max had the highest and most significant temperature increases in spring (0.74°C/10 years) and summer (0.42°C/10 years), with a low speed rise in autumn. T_min also showed a significant increasing trend in all seasons (p<0.05) with high speed rises in spring (0.55°C/10 years) and summer (0.51°C/10 years), and the lowest speed rise in winter (0.47°C/10 years). The rate of increase of T_max in spring was 1.3 times faster than that of T_min, whereas T_min increased 1.2 times, 1.9 times, and 1.1 times more than T_max in summer, autumn, and winter, respectively.

Download:

Table 1. Statistics of trends in seasonal T_max /T_min and NDVI from 1982 to 2015.

https://doi.org/10.1371/journal.pone.0218480.t001

The seasonal NDVI showed a significant (p< 0.01) increasing trend during the study period except for winter.

3.1.2. The pixel scale analysis.

The trends of temperature always show spatial heterogeneity due to the variation characteristics of land surface [28]. The spatial of average temperature cannot effectively describe the variation characteristics in different regions [29].

T_max of the Loess Plateau in spring showed an increasing trend from 1982 to 2015 and 99.1% of the area showed a significant increasing trend (p<0.05), among which 90.9% showed the highest and most significant increasing trend (p<0.01) (Fig 2A). The areas with non-significant increasing trend of T_max in spring (p>0.05) were mainly distributed in north of this area. The area of T_min showing a significant increasing trend was 99.7% (p<0.05), among which 90.3% showing a very significant increasing trend (p<0.01). The areas showing non-significant increasing trends of T_min in spring (p>0.05) were mainly distributed in northwest of the area (Fig 2B).

Download:

Fig 2. Trends of T_max /T_min and NDVI in spring over the Loess Plateau from 1982 to 2015.

(a) the slope of T_max; (b) the slope of T_min; (c) the slope of NDVI; (d) the significance level of T_max trends; (e) the significance level of T_min trends; and (f) the significance level of NDVI trends.

https://doi.org/10.1371/journal.pone.0218480.g002

The area that has an increasing trend of NDVI in spring during the study period was 88.9%. 78.7% of the area has a significant increasing trend(p<0.05). The areas showing significant increasing trends of NDVI in spring (p>0.05) were mainly distributed in middle of this area (Fig 2C).

T_max in summer had a significant increasing trend is 85.5% (p<0.05) of the area, among which 47.8% had a highly significant increasing trend (p<0.01) (Fig 3). Areas with non-significant increasing T_max trends in summer (p>0.05) were mainly distributed in east of the area (Fig 3A). T_min had a significant increasing trend in summer over the Loess Plateau (p<0.05) (Fig 3B).

Download:

Fig 3. Trends of T_max /T_min and NDVI in summer over the Loess Plateau from 1982 to 2015.

(a) the slope of T_max; (b) the slope of T_min; (c) the slope of NDVI; (d) the significance level of T_max trends; (e) the significance level of T_min trends; and (f) the significance level of NDVI trends.

https://doi.org/10.1371/journal.pone.0218480.g003

The area that has an increasing trend of NDVI in summer from1982 to 2015 was 87.5%. 69.1% of the area has a significant increasing trend (p<0.05). The areas showing significant increasing trends of NDVI (p>0.05) were mainly distributed in middle of the area (Fig 3C).

In autumn, T_max in most areas showed an increasing trend, among which of 18.8% was a significant trend (p < 0.05), mainly in west of the area. Only 3% of the regions showed a decreasing trend (p>0.05), mainly in the northeast of the area (Fig 4A). T_min showed a significant increasing trend was 99.3% of the region, and the fastest increasing trends occurred in northeast of the area(p<0.05) (Fig 4B).

Download:

Fig 4. Trends of T_max /T_min and NDVI in autumn over the Loess Plateau from 1982 to 2015.

(a) the slope of T_max; (b) the slope of T_min; (c) the slope of NDVI; (d) the significance level of T_max trends; (e) the significance level of T_min trends; and (f) the significance level of NDVI trends.

https://doi.org/10.1371/journal.pone.0218480.g004

The area that has an increasing trend of NDVI in autumn from1982 to 2015 was 95.3%. 81.2% of the area has a significant increasing trend (p<0.05). The areas showing non-significant increasing trends of NDVI (p>0.05) were mainly distributed in south and northwest of the area (Fig 4C).

In winter, T_max showed a significant increasing trend in 26.4% (p<0.05) of the area, mainly in middle and west of area (Fig 5A). T_min showed a downward trend in only 0.5% of the region (Fig 5B), mostly in north of area. T_min showed a significant increasing trend (p<0.05) in northeast (Fig 5B).

Download:

Fig 5. Trends of T_max /T_min and NDVI in winter over the Loess Plateau from 1982 to 2015.

(a) the slope of T_max; (b) the slope of T_min; (c) the slope of NDVI; (d) the significance level of T_max trends; (e) the significance level of T_min trends; and (f) the significance level of NDVI trends.

https://doi.org/10.1371/journal.pone.0218480.g005

The area that has an increasing trend of NDVI in winter from 1982 to 2015 was 72.7%. 48.3% of the area has a significant increasing trend (p<0.05). The areas showing non-significant increasing trends of NDVI in winter (p>0.05) were mainly distributed throughout the area (Fig 5C).

3.2. Partial correlation between the NDVI and seasonal asymmetric warming

3.2.1. Region analysis.

The partial correlation coefficient between the NDVI and the seasonal T_max/T_min (Table 2) showed that the second-order partial correlation coefficient of the NDVI and T_max was positive for the area as a whole. Among them, the NDVI and T_max showed an extremely significant positive correlation in spring and winter (p < 0.01), a significant positive correlation in autumn (p < 0.05), and no correlation in summer (p > 0.05).

Download:

Table 2. Partial correlation coefficients between NDVI and T_max and T_min for different vegetation types in different season over the Loess Plateau.

https://doi.org/10.1371/journal.pone.0218480.t002

The second-order partial correlation coefficient of the NDVI and T_min passed the significance tests of p < 0.01 in spring and winter and p < 0.05 in summer and was non-significant correlated in autumn (p > 0.05). Thus, compared with the day-time temperature increase, the night-time temperature increases in summer had a more significant positive impact on the NDVI. In contrast, in autumn, the day-time temperature increase had a more significant impact on the NDVI than the night-time increase. T_max and T_min in spring had a more positive effect on the NDVI in the area.

The different types of NDVI showed partial positive correlations with T_max and T_min in spring (p<0.05) and non-significant partial positive correlations with T_max in summer (p> 0.05). In summer, grasslands and wetlands showed partial positive correlations with T_min (p < 0.05), while other types of NDVI showed non-significant partial positive correlations (p> 0.05). Except cultivated land, there was a non-significant partial positive correlation between the NDVI and T_max in autumn (p > 0.05). In autumn, cultivated land showed significant partial positive correlations with T_min (p <0.05), while other types of NDVI showed non-significant partial positive correlations (p>0.05). In winter, cultivated land, forest, and grassland showed partial positive correlations with T_max and T_min (p<0.05), shrub showed non-significant partial positive correlations with T_max and T_min (p>0.05), wetland showed a significant partial positive correlation with T_max. In conclusion, different types of vegetation had significant differences in response to daytime and nighttime warming in different seasons.

3.2.2. The pixel scale analysis.

In order to understand the spatial pattern of the correlation between the seasonal NDVI and T_max/T_min in the area, patterns for the correlation between seasonal NDVI and T_max or T_min were obtained.

In spring, the NDVI was positively correlated with T_max in 92.9% of the regions, of which 68.2% were significant at p < 0.05. Only 7.1% of the regions showed a negative correlation between the NDVI and T_max, mainly in south and northwest however, these correlations were non-significant (p > 0.05). A positive correlation between the NDVI and T_min was shown in 90.6% of the area, with 58.0% being significant (p < 0.05). 9.4% of the area showed a negative correlation between the NDVI and T_min and did not pass the significance test (p > 0.05). The distribution range was the same as that described for negative correlations between the NDVI and T_max. Thus, in general, daytime and nighttime warming in spring had positive effects on the growth of vegetation in most areas of the Loess Plateau, possibly due to a common advance of the vegetation growing season as a result of the rising temperature [30]. The area where the NDVI and T_max showed a positive correlation was slightly larger than that where the NDVI and T_min showed a positive correlation, possibly because a rising temperature in the day-time has a greater effect on the advancement of seedling and leaf development [31].

The spatial distribution of the partial correlation coefficient between the NDVI and daytime and nighttime warming in summer (Fig 6) showed that 70.3% of the regions had a positive correlation between the NDVI and T_max. 9.0% (p < 0.05) of the regions that passed the significance test were mainly distributed in the Northwestern Loess Plateau. 29.6% of the regions, mainly in the southeast of the area, showed non-significant, negative correlations between the NDVI and T_max. 78.4% of the regions where the NDVI and T_min showed positive correlations and 25.8% of the regions that passed the significance test (p < 0.05) were distributed in the middle and northwest of the study area. 21.6% of the regions that showed a negative correlation between the NDVI and T_min, mainly in the south of this area, did not pass the significance test (p > 0.05). Thus, in general, a higher proportion of areas showed a positive impact on vegetation from daytime and nighttime warming as compared with those in which it had a negative impact. Compared with spring, the proportion of NDVI and T_max showing a significant positive correlation decreased significantly in summer, possibly because the day-time temperature in summer is close to the optimal temperature for vegetation growth in most areas, leading to lower sensitivity of the vegetation to day-time temperature increase in this season [32].

Download:

Fig 6. Spatial patterns of the correlations between NDVI and corresponding T_max or T_min over the Loess Plateau from 1982 to 2015 among seasons.

(1) spatial patterns of the correlations between NDVI and corresponding T_max or T_min in spring; (2) spatial patterns of the correlations between NDVI and corresponding T_max or T_min in summer; (3) spatial patterns of the correlations between NDVI and corresponding T_max or T_min in autumn; (4) spatial patterns of the correlations between NDVI and corresponding T_max or T_min in winter;(a) The partial correlation coefficients between NDVI and T_max. (b) The partial correlation coefficients between NDVI and T_min;(c) significance level of the partial correlation coefficients between NDVI and T_max; and (d) significance level of the partial correlation coefficients between NDVI and T_min.

https://doi.org/10.1371/journal.pone.0218480.g006

The spatial distribution of the partial correlation coefficient of autumn vegetation and daytime and nighttime warming (Fig 6) showed that 85.8% of the regions showed a positive correlation between the NDVI and T_max, and the 12.6% (p < 0.05) that passed the significance test were mainly distributed in the west and south. 14.2% of the regions, mainly in the southwest of the study area, showed a non-significant negative correlation between the NDVI and T_max. 75.8% of the regions, mainly in the east and south of the area, showed a positive correlation between the NDVI and T_min, with 13.6% passing the significance test (p < 0.05). 24.2% of the regions, mainly those in the northwest of the Loess Plateau, in Ordos, showed a non-significant negative correlation between the NDVI and T_min. Thus, the temperature increases in autumn had a positive effect on vegetation in most areas. Compared with spring, the area showing a significant positive correlation between the NDVI and T_max decreased in autumn, possibly because the duration of vegetation photosynthesis and the sensitivity of vegetation productivity to temperature in autumn were lower [33].

The spatial distribution of the partial correlation coefficient of winter vegetation and daytime and nighttime warming (Fig 6) showed a positive correlation between the NDVI and T_max in 94.0% of the regions, and only 18.3% (p < 0.05) passed the significance test, mainly those in the southwest of the Loess Plateau (p < 0.05). 6.0% of the regions, mainly those in the northwest, showed a non-significant negative correlation between the NDVI and T_max. The 80.9% of the regions where the NDVI and T_min showed a positive correlation, and the 4.7% of the regions that passed the significance test (p < 0.05) were mainly distributed in the south of the area. 19.1% of the regions, mainly those in the east of this area, showed a non-significant negative correlation between the NDVI and T_min. Thus, daytime and nighttime warming in winter was conducive to vegetation growth in most areas, possibly because temperature, as a stress factor of vegetation growth in winter, is conducive to vegetation photosynthesis.

4. Discussions

In this study, daytime and nighttime warming is asymmetric among different seasons, which coincides with some previous studies [34]. This study found that seasonal NDVI in the Loess Plateau showed increasing trends, which is in agreement with by previous studies [35]. Fensholt et al [36] assessed the accuracy of the long term GIMMS3g NDVI data using MODIS NDVI for an overlapping period of 2000–2010 and concluded that the temporal trends derived from GIMMS NDVI are in overall good agreement with trends from MODIS NDVI data in most of the world's semi-arid, dry sub-humid, and sub-humid areas.

We found that daytime and nighttime warming over the Loess Plateau had a positive effect on NDVI during 1982–2015. Specifically, the response of the NDVI to daytime and nighttime warming was significantly different in different seasons. Among all the seasons, spring shows the largest area with a significantly positive relationship between T_max and the NDVI (Fig 6). The availability of soil nitrogen was improved in spring [37]. There are earlier and longer growing seasons driven by temperature, which can promote vegetation photosynthetic activity [30]. The area with a significantly positive relationship between T_max and the NDVI during winter was smaller than that in the other seasons, because of lower soil moisture and surface soil frost in some regions [38]. Winter does not include growing seasons (April-October). As such, the temperature sensitivities of the photosynthetic activity and productivity of winter are both smaller than that of the other three seasons. Summer shows a smaller area, with a significantly positive relationship between T_max and the NDVI (Fig 6), compared to spring and autumn. The analysis suggests that T_max of summer in most areas is close to the optimal value for plant photosynthetic activity [30]. Fensholt, et al [36] gave an explanation for the smaller response between the NDVI and T_max in summer than that in spring and autumn, i.e., it is associated with the saturation of the NDVI over dense vegetation in summer. Similar to T_max, T_min may have impacted on vegetation photosynthetic activity in opposite ways, through the action of different mechanisms [39–40]. The study showed that the positive correlation between T_min and the NDVI is mainly evident in spring and summer, implying that the dominant ecophysiological effects of T_min vary seasonally.

In most regions, T_max and T_min had promoting effects on the NDVI, and in a small number of regions the day-night temperature rise had an inhibiting effect on the NDVI. Most vegetation photosynthesizes during the daytime [3]. In semi-humid areas with excess moisture and areas in which temperature is the main limiting factor to vegetation growth, an increase in T_max can promote the opening of stoma in vegetation leaves and vegetation transpiration, therefore also increasing the probability of CO₂ and improving the availability of soil nitrogen. This promotes the growth of vegetation, characterized by an NDVI increase [41]. An increase in T_max in the arid areas of the Loess Plateau can accelerate leaf transpiration, leading to increased evaporation, reduced soil moisture, exposure to water stress, and inhibition of growth and photosynthesis. This is consistent with the research results of Peng et al. [3], who found that an increase of T_max was beneficial to vegetation growth and its ecosystem carbon sink function in most cold and humid regions, but not to the vegetation growth in temperate arid and semi-arid regions.

An increase in T_min at night can decrease the volume of endosperm cells in the mature stage of vegetation [42] and shorten the grain-filling stage of vegetation [43] by increasing the rate of vegetation autotrophic respiration [44]. This has a negative effect on the promotion of the NDVI in vegetation. The increase in T_min can improve vegetation resistance to drought [45] and regulate the content of carbohydrates in plant leaves [46] by reducing the frequency of frost disasters [35]. This has a positive impact on vegetation productivity.

Since the photosynthesis of most vegetation takes place during the day and the respiration of vegetation runs throughout the day, asymmetric warming will inevitably affect the carbon absorption and consumption of vegetation [47]. The various types of vegetation showed partial positive correlations with T_max and T_min in spring (p < 0.05), likely due to the advancement of seedling establishment and sprouting leaves. There was a non-significant partial correlation between various types of NDVI and T_max in summer. Except for cultivated land, there was a non-significant partial correlation between other types of vegetation and T_max in autumn. In most areas, the day-time temperature in summer and autumn was close to the optimum for vegetation growth, reducing the sensitivity of vegetation to day-time warming. Cultivated land, forest, and grassland all showed positive partial correlations with T_max and T_min in winter (p < 0.05), shrub showed a non-significant partial correlation with T_max and T_min in winter (p > 0.05), wetland showed a significant positive partial correlation with T_max in winter. Different types of vegetation had significant differences in daytime and nighttime warming in winter. In winter, the rising temperature was conducive to the photosynthesis of large vegetation. Meanwhile, the decrease in rainfall and water impaired the growth of vegetation.

The NDVI is a comprehensive response to climate change. This paper mainly discussed precipitation and temperature under hydrothermal climate conditions. Future research should comprehensively consider the impacts of humidity, evaporation, sunshine duration, and other meteorological factors on the NDVI. In recent years, ecological projects have been carried out in the Loess Plateau, such as returning grazing land to grassland, which will lead to changes in land cover and subsequent changes in the NDVI [48]. Determining how to quantitatively eliminate the dynamic changes in vegetation caused by land cover change is an important issue for further research. This paper used 8 km×8 km mixed pixel NDVI data, which comprehensively reflected the spectrum of all land cover. However, the next step is to combine other higher resolution image data, such as MODIS NDVI. Data-fusion that was carried out on images from 1982–2000 and high resolution remote sensing image data from 2000–2015 can be combined to obtain nearly 34 years of high resolution image datasets to allow further study on the responses of vegetation dynamics to seasonal daytime and night-time asymmetric warming. Meanwhile, such effects can be further quantified by combining the observation experimental data of ground stations and model simulation analyses to provide a more reliable scientific basis for ecological environment monitoring and adaptation to climate change.

5. Conclusion

In the Loess Plateau, the daytime and nighttime temperatures in each season showed significant (p< 0.05) increasing trends over the period 1982-2015except for T_max in autumn. The increasing speed of T_max and T_min were asymmetrical across Different seasons, and have not been evenly distributed in this area. For example, the increasing rates of T_min in spring, summer, autumn, and winter were 0.75, 1.20, 1.88, and 1.10 times larger than that of T_max, respectively. Secondly, the NDVI in the Loess Plateau was extremely significant (p< 0.01) positively correlated with T_max and T_min in spring and winter, and with T_min in summer and T_max in autumn across entire Loess Plateau. Meanwhile, the different vegetation had different responses to seasonal daytime and nighttime asymmetric warming. Whether it increasing of the T_max or T_min in spring, it will promote the growth of vegetation. Only grasslands and wetland benefitted from increasing of T_min in summer. Autumnal increasing of T_max or T_min merely helped cultivated land growth. Winter increasing of T_max or T_min could promote growth of cultivated land, forest, and grassland, while only wetland showed a significantly positive partial correlation with T_max in winter.

The results suggested that vegetation variability to asymmetric warming is important to understand the changes in vegetation photosynthetic activity under global warming and ultimately influences regional and hemispheric-scale carbon balances. Given climate warming is projected to further increase in the future, our results, which addresses an improved characterization vegetation variability to asymmetric warming under the temporal and spatial scales, provide a valuable reference for assessing and predicting underlying mechanisms of vegetation variability to global climate change.

Acknowledgments

All authors thank the anonymous reviewers and the editor for the constructive comments on an earlier version of this manuscript. Qinming Meng, who work in department of Geosciences of Mississippi State University, contributed to revise the final paper and modify English writing.

References

1. Yi L, Ren Z, Zhang C, Liu W. Vegetation cover, climate and human activities on the Loess Plateau. Resources Sciences. 2014; 36(1): 166–174. https://doi.org/1007-7588(2014)01-0166-09.
- View Article
- Google Scholar
2. He B, Chen AF, Wang HL, Wang QF. Dynamic response of satellite-derived vegetation growth to climate change in the three north shelter forest region in China. Remote Sensing. 2015; 7(8): 9998–10016. https://doi.org/10.3390/rs70809998
- View Article
- Google Scholar
3. Peng SS, Piao SL, Ciais P, Myneni RB, Chen A, Chevallier F et al. Asymmetric effects of daytime and night-time warming on Northern Hemisphere vegetation. Nature. 2013; 501:88–93. https://doi.org/10.1038/nature12434 pmid:24005415
- View Article
- PubMed/NCBI
- Google Scholar
4. Zhang C, Ren ZY, Wei ZF. Vegetation responses to intra-annual hydrothermal conditions on the loess plateau. Resources Science. 2013;35(10):2017–2023.
- View Article
- Google Scholar
5. Wang XC, Li J, Wang HN, Hao MD. Simulation of the response of soil water in winter wheat field and winter wheat yield to rainfall and temperature change on the loess plateau. Journal of Natural Resources. 2017; 32(8): 1398–1410.
- View Article
- Google Scholar
6. Vose RS, Easterling DR, Gleason B. Maximum and minimum temperature trends for the globe: An update through 2004. Geophysical Research Letters. 2005; 32(23): L23822. https://doi.org/10.1029/2005GL024379.
- View Article
- Google Scholar
7. Xia HM, Li AN, Feng G, Li Y, Qin Y, Lei G. et al. The effects of asymmetric diurnal warming on vegetation growth of the Tibetan Plateau over the past three decades. Sustainability. 2018; 10: 1103. https://doi.org/10.3390/su10041103
- View Article
- Google Scholar
8. Xu L, Myneni RB, Chapin FS, Callaghan TV, Pinzon JE, Tucker CJ, et al. Temperature and vegetation seasonality diminishment over northern lands. Nature climate change. 2013; 3:581–586. https://doi.org/10.1038/NCLIMATE1836.
- View Article
- Google Scholar
9. Shen M, Piao S, Chen X, An S, Fu YH, Wang S, et al. Strong impacts of daily minimum temperature on the green-up date and summer greenness of the Tibetan Plateau. Glob Chang Biol.2016;22(9):3057–3066. https://doi.org/10.1111/gcb.13301 pmid:27103613
- View Article
- PubMed/NCBI
- Google Scholar
10. Piao SL, Ciais P, Friedlingstein P, Peylin P, Reichstein M, Luyssaert S, et al. Net carbon dioxide losses of northern ecosystems in response to autumn warming. Nature. 2008; 451(7174): 49–52. https://doi.org/10.1038/nature06444 pmid:18172494
- View Article
- PubMed/NCBI
- Google Scholar
11. Xu H, Wang X, Yang T. Trend shifts in satellite-derived vegetation growth in Central Eurasia, 1982–2013. Sci Total Environ. 2017; 579: 1658–1674. pmid:27919557
- View Article
- PubMed/NCBI
- Google Scholar
12. Wu X, Liu H, Li X, Liang E, Beck PSA, Huang YM. Seasonal divergence in the interannual responses of northern hemisphere vegetation activity to variations in diurnal climate. Sci Rep.2016; 6: 19000. https://doi.org/10.1038/srep19000 pmid:26751166
- View Article
- PubMed/NCBI
- Google Scholar
13. Tan J, Piao S, Chen A, Zeng Z, Ciais P, Janssens IA, et al. Seasonally different response of photosynthetic activity to day-time and night-time warming in the Northern Hemisphere. Glob Chang Biol. 2015; 21:377–387. https://doi.org/10.1111/gcb.12724 pmid:25163596
- View Article
- PubMed/NCBI
- Google Scholar
14. Yan LB. Characteristics of temperature and precipitation on the Loess Plateau from 1961 to 2014. Journal of Earth Environment. 2015; 6: 276–282.
- View Article
- Google Scholar
15. Zhang C, Ren ZY. Temporal and spatial differences and its trends in vegetation cover change over the loess plateau. Resour Sci. 2011; 33:2143–2149.
- View Article
- Google Scholar
16. Rouse JW. Monitoring the vernal advancement and retrogradation of natural vegetation. NASA/GSFCT Type II Report. Greenbelt, MD, USA.;1973.
17. Ma LQ, Qin F, Sun JL, Wang H, Xia HM. Diurnal asymmetry of temperature and its effect on vegetation dynamics in Loess Plateau. J Resources Science.2018; 40(8):1684–1692.
- View Article
- Google Scholar
18. http://data.cma.cn/data/detail/DataCode/A.0012.0001.html. China Meteorological Data Service Center. Hourly Observation Data from China Ground Weather Station. (accessed on 21 April 2018).
19. Hyndman R, Fan Y. Sample quantiles in statistical packages. Am Stat. 1996; 50: 361–367.
- View Article
- Google Scholar
20. Wang XL, Wen QH, Wu Y. Penalized maximal t test for detecting undocumented mean change in climate data series. J Appl. Meteorol Climatol. 2007; 46:916–931.
- View Article
- Google Scholar
21. Li YC, He ZM, Liu CX. Review on spatial interpolation methods of temperature data from meteorological stations. Progr Geogr. 2014; 33: 1019–1028.
- View Article
- Google Scholar
22. Ma LQ, Xia HM, Meng QM. Spatiotemporal Variability of Asymmetric Daytime and Night-Time Warming and Its Effects on Vegetation in the Yellow River Basin from 1982 to 2015. Sensors. 2019; 19(8):1832. https://doi.org/10.3390/s19081832.
- View Article
- Google Scholar
23. Holben B.N. Characterization of maximum value composites from temporal AVHRR data. Int J Remote Sensing. 1986; 7:1417–1434.
- View Article
- Google Scholar
24. Piao S, Cui M, Chen A, Wang X, Ciais P, Liu J, Tang Y. Altitude and temperature dependence of change in the spring vegetation green-up date from 1982 to 2006 in the Qinghai-Xizang Plateau. Agric For Meteorol. 2011; 151(12):1599–1608. https://doi.org/10.1016/j.agrformet.2011.06.016
- View Article
- Google Scholar
25. http://www.globallandcover.com/Chinese/GLC30Download/index.aspx. Institute of Remote Sensing and Digital Earth Chinese Academy of Sciences. (accessed on 20 August 2018).
26. Mann HB. Non-parametric tests against trend. Econometrics. 1945;13(3):245–259. http://www.jstor.org/stable/1907187.
- View Article
- Google Scholar
27. Baba K, Shibata R, Sibuya M. Partial correlation and conditional correlation as measures of conditional independence. Australian New Zealand Journal of Statistics. 2004; 46(4): 657–664. https://doi.org/10.1111/j.1467-842X.2004.00360.x
- View Article
- Google Scholar
28. Wu ZL, Jia WX, Zhao Z, Zhang YS, Liu YR, Chen JH. Spatial-temporal variations of vegetation and its correlation with climatic factors in Qilian Mountains from 2000 to 2012. Arid Land Geogr. 2015; 38: 1241–1252. https://doi.org/10.13826/j.cnki.cn65-1103/x.2015.06.017
- View Article
- Google Scholar
29. Liu XF, Zhu XF, Pan YZ, Li YZ, Zhao AZ. Spatiotemporal changes in vegetation coverage in China during 1982–2012. Acta Ecol Sin. 2015; 35: 5331–5342. https://doi.org/10.5846/stxb201404150731.
- View Article
- Google Scholar
30. Piao SL, Friedlingstein P, Ciais P, Viovy N, Demarty J. Growing season extension and its impact on terrestrial carbon cycle in the Northern Hemisphere over the past 2 decades. Glob Biogeochem Cycles. 2007; 21: GB3018. https://doi.org/10.1029/2006GB002888.
- View Article
- Google Scholar
31. Piao SL, Tan J, Chen A, Fu YH, Ciais P, Liu Q, et al. Leaf onset in the northern hemisphere triggered by day-time temperature. Nat Commun. 2015; 6: 6911. https://doi.org/10.1038/ncomms7911 pmid:25903224
- View Article
- PubMed/NCBI
- Google Scholar
32. Wan S, Hui D, Wallace L, Luo Y. Direct and indirect effects of experimental warming on ecosystem carbon processes in a tallgrass prairie. Glob Biogeochem Cycles. 2005; 19: GB2014. https://doi.org/10.1029/2004GB002315
- View Article
- Google Scholar
33. Graven HD, Keeling RF, Piper SC, Patra PK, Stephens BB, Wofsy SC, et al. Enhanced seasonal exchange of CO2 by northern ecosystems since 1960. Science. 2013; 341(6150): 1085–1089.https://doi.org/10.1126/science.1239207 pmid:23929948
- View Article
- PubMed/NCBI
- Google Scholar
34. Intergovernmental Panel on Climate Change (IPCC). Climate Change 2013: The Physical Science Basis.Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change;Cambridge University Press: Cambridge, UK; New York, NY, USA, 2013.
35. Zhang CS, Hu Y,Shi XL. Analysis of Spatial-Temporal Evolution of vegetation cover in Loess Plateau in recent 33 years based on AVHRR NDVI and MODIS NDVI. Journal of applied sciences electronics and information engineering. 2006;34(6):702–712. https://doi.org/10.3969/j.issn.0255-8297.2016.06.006
- View Article
- Google Scholar
36. Fensholt RP, Simon R. Evaluation of Earth Observation based global long term vegetation trends-Comparing GIMMS and MODIS global NDVI time series. Remote Sensing of Environment. 2012; 119:131–147. https://doi.org/10.1016/j.rse.2011.12.015
- View Article
- Google Scholar
37. Melillo JM, Butler S, Johnson J, Mohan J, Steudler P, Lux H, et al. Soil warming, carbon–nitrogen interactions, and forest carbon budgets. Proceedings of the National Academy of Sciences. 2011;108:9508–9512. https://doi.org/10.1073/pnas.1018189108.
- View Article
- Google Scholar
38. Robock A, Vinnikov KY, Srinivasan G, Entin JK, Hollinger SE, Speranskaya NA, et al.The global soil moisture data bank. Bulletin of the American Meteorological Society. 2000; 81(6): 1281–1299. https://doi.org/10.1175/1520-0477
- View Article
- Google Scholar
39. Wan SQ, Xia JY, Liu WX, Niu SL. Photosynthetic overcompensation under nocturnal warming enhances grassland carbon sequestration. Ecology. 2009; 90(10): 2700–2710. https://doi.org/10.1890/08-2026.1 pmid:19886480
- View Article
- PubMed/NCBI
- Google Scholar
40. Beier C, Emmett B, Gundersen P, Tietema A, Penuelas J, Estiarte M, et al. Novel approaches to study climate change effects on terrestrial ecosystems in the field: drought and passive nighttime warming. Ecosystems. 2004; 7:583–597. https://doi.org/10.1007/s10021-004-0178-8
- View Article
- Google Scholar
41. Liang CL, Yu QZ, Liu YJ, Zhang Z. Effects of air temperature circadian on the NDVI of Nansi Lake wetland vegetation. Trop Geogr. 2015; 35(3):422–426. https://doi.org/10.13284/j.cnki.rddl.002746
- View Article
- Google Scholar
42. Alward RD, Detling JK, Milchunas DG. Grassland vegetation changes and nocturnal global warming. Science. 1999; 283: 229–231. https://doi.org/10.1126/science.283.5399.229 pmid:9880257
- View Article
- PubMed/NCBI
- Google Scholar
43. Morita S, Yonemaru J, Takanashi J. Grain growth and endosperm cell size under high night temperatures in rice (Oryza sativa L.). Ann Bot. 2005; 95: 695–701. https://doi.org/10.1093/aob/mci071 pmid:15655104
- View Article
- PubMed/NCBI
- Google Scholar
44. Welch JR, Vincent JR, Auffhammer M, Moya PF, Dobermann A, Dawe D. Rice yields in tropical/subtropical Asia exhibit large but opposing sensitivities to minimum and maximum temperatures. Proc Natl Acad Sci. 2010; 107: 14562–14567. https://doi.org/10.1073/pnas.1001222107 pmid:20696908
- View Article
- PubMed/NCBI
- Google Scholar
45. Nicholls N. Increased Australian wheat yield due to recent climate trends. Nature.1997;387:484–485. https://doi.org/10.1038/387484a0.
- View Article
- Google Scholar
46. Yang Z, Jiang L, Su F, Zhang Q, Xia J, Wan S. Night-time warming enhances drought resistance of plant communities in a temperate steppe. Sci Rep. 2016; 6:23267. https://doi.org/10.1038/srep23267 pmid:26987482
- View Article
- PubMed/NCBI
- Google Scholar
47. Atkin OK, Turnbull MH, Zaragoza-Castells J, Fyllas NM, Lloyd J, Meir P, et al. Light inhibition of leaf respiration as soil fertility declines along a post-glacial chronosequence in New Zealand, an analysis using the Kok method. Plant Soil. 2013; 367: 163–182. https://doi.org/10.1007/s11104-013-1686-0.
- View Article
- Google Scholar
48. Zhao AZ, Zhang AB, Liu HX, Liu YX, Wang HF, Wang DL. Spatiotemporal variation of vegetation coverage before and after implementation of grain for Green Project in the Loess Plateau. J Nat Resour. 2017; 32: 449–460. https://doi.org/10.11849/zrzyxb.20160411
- View Article
- Google Scholar

[ref1] 1. Yi L, Ren Z, Zhang C, Liu W. Vegetation cover, climate and human activities on the Loess Plateau. Resources Sciences. 2014; 36(1): 166–174. https://doi.org/1007-7588(2014)01-0166-09.
View Article
Google Scholar

[2] View Article

[3] Google Scholar

[ref2] 2. He B, Chen AF, Wang HL, Wang QF. Dynamic response of satellite-derived vegetation growth to climate change in the three north shelter forest region in China. Remote Sensing. 2015; 7(8): 9998–10016. https://doi.org/10.3390/rs70809998
View Article
Google Scholar

[5] View Article

[6] Google Scholar

[ref3] 3. Peng SS, Piao SL, Ciais P, Myneni RB, Chen A, Chevallier F et al. Asymmetric effects of daytime and night-time warming on Northern Hemisphere vegetation. Nature. 2013; 501:88–93. https://doi.org/10.1038/nature12434 pmid:24005415
View Article
PubMed/NCBI
Google Scholar

[8] View Article

[9] PubMed/NCBI

[10] Google Scholar

[ref4] 4. Zhang C, Ren ZY, Wei ZF. Vegetation responses to intra-annual hydrothermal conditions on the loess plateau. Resources Science. 2013;35(10):2017–2023.
View Article
Google Scholar

[12] View Article

[13] Google Scholar

[ref5] 5. Wang XC, Li J, Wang HN, Hao MD. Simulation of the response of soil water in winter wheat field and winter wheat yield to rainfall and temperature change on the loess plateau. Journal of Natural Resources. 2017; 32(8): 1398–1410.
View Article
Google Scholar

[15] View Article

[16] Google Scholar

[ref6] 6. Vose RS, Easterling DR, Gleason B. Maximum and minimum temperature trends for the globe: An update through 2004. Geophysical Research Letters. 2005; 32(23): L23822. https://doi.org/10.1029/2005GL024379.
View Article
Google Scholar

[18] View Article

[19] Google Scholar

[ref7] 7. Xia HM, Li AN, Feng G, Li Y, Qin Y, Lei G. et al. The effects of asymmetric diurnal warming on vegetation growth of the Tibetan Plateau over the past three decades. Sustainability. 2018; 10: 1103. https://doi.org/10.3390/su10041103
View Article
Google Scholar

[21] View Article

[22] Google Scholar

[ref8] 8. Xu L, Myneni RB, Chapin FS, Callaghan TV, Pinzon JE, Tucker CJ, et al. Temperature and vegetation seasonality diminishment over northern lands. Nature climate change. 2013; 3:581–586. https://doi.org/10.1038/NCLIMATE1836.
View Article
Google Scholar

[24] View Article

[25] Google Scholar

[ref9] 9. Shen M, Piao S, Chen X, An S, Fu YH, Wang S, et al. Strong impacts of daily minimum temperature on the green-up date and summer greenness of the Tibetan Plateau. Glob Chang Biol.2016;22(9):3057–3066. https://doi.org/10.1111/gcb.13301 pmid:27103613
View Article
PubMed/NCBI
Google Scholar

[27] View Article

[28] PubMed/NCBI

[29] Google Scholar

[ref10] 10. Piao SL, Ciais P, Friedlingstein P, Peylin P, Reichstein M, Luyssaert S, et al. Net carbon dioxide losses of northern ecosystems in response to autumn warming. Nature. 2008; 451(7174): 49–52. https://doi.org/10.1038/nature06444 pmid:18172494
View Article
PubMed/NCBI
Google Scholar

[31] View Article

[32] PubMed/NCBI

[33] Google Scholar

[ref11] 11. Xu H, Wang X, Yang T. Trend shifts in satellite-derived vegetation growth in Central Eurasia, 1982–2013. Sci Total Environ. 2017; 579: 1658–1674. pmid:27919557
View Article
PubMed/NCBI
Google Scholar

[35] View Article

[36] PubMed/NCBI

[37] Google Scholar

[ref12] 12. Wu X, Liu H, Li X, Liang E, Beck PSA, Huang YM. Seasonal divergence in the interannual responses of northern hemisphere vegetation activity to variations in diurnal climate. Sci Rep.2016; 6: 19000. https://doi.org/10.1038/srep19000 pmid:26751166
View Article
PubMed/NCBI
Google Scholar

[39] View Article

[40] PubMed/NCBI

[41] Google Scholar

[ref13] 13. Tan J, Piao S, Chen A, Zeng Z, Ciais P, Janssens IA, et al. Seasonally different response of photosynthetic activity to day-time and night-time warming in the Northern Hemisphere. Glob Chang Biol. 2015; 21:377–387. https://doi.org/10.1111/gcb.12724 pmid:25163596
View Article
PubMed/NCBI
Google Scholar

[43] View Article

[44] PubMed/NCBI

[45] Google Scholar

[ref14] 14. Yan LB. Characteristics of temperature and precipitation on the Loess Plateau from 1961 to 2014. Journal of Earth Environment. 2015; 6: 276–282.
View Article
Google Scholar

[47] View Article

[48] Google Scholar

[ref15] 15. Zhang C, Ren ZY. Temporal and spatial differences and its trends in vegetation cover change over the loess plateau. Resour Sci. 2011; 33:2143–2149.
View Article
Google Scholar

[50] View Article

[51] Google Scholar

[ref16] 16. Rouse JW. Monitoring the vernal advancement and retrogradation of natural vegetation. NASA/GSFCT Type II Report. Greenbelt, MD, USA.;1973.

[ref17] 17. Ma LQ, Qin F, Sun JL, Wang H, Xia HM. Diurnal asymmetry of temperature and its effect on vegetation dynamics in Loess Plateau. J Resources Science.2018; 40(8):1684–1692.
View Article
Google Scholar

[54] View Article

[55] Google Scholar

[ref18] 18. http://data.cma.cn/data/detail/DataCode/A.0012.0001.html. China Meteorological Data Service Center. Hourly Observation Data from China Ground Weather Station. (accessed on 21 April 2018).

[ref19] 19. Hyndman R, Fan Y. Sample quantiles in statistical packages. Am Stat. 1996; 50: 361–367.
View Article
Google Scholar

[58] View Article

[59] Google Scholar

[ref20] 20. Wang XL, Wen QH, Wu Y. Penalized maximal t test for detecting undocumented mean change in climate data series. J Appl. Meteorol Climatol. 2007; 46:916–931.
View Article
Google Scholar

[61] View Article

[62] Google Scholar

[ref21] 21. Li YC, He ZM, Liu CX. Review on spatial interpolation methods of temperature data from meteorological stations. Progr Geogr. 2014; 33: 1019–1028.
View Article
Google Scholar

[64] View Article

[65] Google Scholar

[ref22] 22. Ma LQ, Xia HM, Meng QM. Spatiotemporal Variability of Asymmetric Daytime and Night-Time Warming and Its Effects on Vegetation in the Yellow River Basin from 1982 to 2015. Sensors. 2019; 19(8):1832. https://doi.org/10.3390/s19081832.
View Article
Google Scholar

[67] View Article

[68] Google Scholar

[ref23] 23. Holben B.N. Characterization of maximum value composites from temporal AVHRR data. Int J Remote Sensing. 1986; 7:1417–1434.
View Article
Google Scholar

[70] View Article

[71] Google Scholar

[ref24] 24. Piao S, Cui M, Chen A, Wang X, Ciais P, Liu J, Tang Y. Altitude and temperature dependence of change in the spring vegetation green-up date from 1982 to 2006 in the Qinghai-Xizang Plateau. Agric For Meteorol. 2011; 151(12):1599–1608. https://doi.org/10.1016/j.agrformet.2011.06.016
View Article
Google Scholar

[73] View Article

[74] Google Scholar

[ref25] 25. http://www.globallandcover.com/Chinese/GLC30Download/index.aspx. Institute of Remote Sensing and Digital Earth Chinese Academy of Sciences. (accessed on 20 August 2018).

[ref26] 26. Mann HB. Non-parametric tests against trend. Econometrics. 1945;13(3):245–259. http://www.jstor.org/stable/1907187.
View Article
Google Scholar

[77] View Article

[78] Google Scholar

[ref27] 27. Baba K, Shibata R, Sibuya M. Partial correlation and conditional correlation as measures of conditional independence. Australian New Zealand Journal of Statistics. 2004; 46(4): 657–664. https://doi.org/10.1111/j.1467-842X.2004.00360.x
View Article
Google Scholar

[80] View Article

[81] Google Scholar

[ref28] 28. Wu ZL, Jia WX, Zhao Z, Zhang YS, Liu YR, Chen JH. Spatial-temporal variations of vegetation and its correlation with climatic factors in Qilian Mountains from 2000 to 2012. Arid Land Geogr. 2015; 38: 1241–1252. https://doi.org/10.13826/j.cnki.cn65-1103/x.2015.06.017
View Article
Google Scholar

[83] View Article

[84] Google Scholar

[ref29] 29. Liu XF, Zhu XF, Pan YZ, Li YZ, Zhao AZ. Spatiotemporal changes in vegetation coverage in China during 1982–2012. Acta Ecol Sin. 2015; 35: 5331–5342. https://doi.org/10.5846/stxb201404150731.
View Article
Google Scholar

[86] View Article

[87] Google Scholar

[ref30] 30. Piao SL, Friedlingstein P, Ciais P, Viovy N, Demarty J. Growing season extension and its impact on terrestrial carbon cycle in the Northern Hemisphere over the past 2 decades. Glob Biogeochem Cycles. 2007; 21: GB3018. https://doi.org/10.1029/2006GB002888.
View Article
Google Scholar

[89] View Article

[90] Google Scholar

[ref31] 31. Piao SL, Tan J, Chen A, Fu YH, Ciais P, Liu Q, et al. Leaf onset in the northern hemisphere triggered by day-time temperature. Nat Commun. 2015; 6: 6911. https://doi.org/10.1038/ncomms7911 pmid:25903224
View Article
PubMed/NCBI
Google Scholar

[92] View Article

[93] PubMed/NCBI

[94] Google Scholar

[ref32] 32. Wan S, Hui D, Wallace L, Luo Y. Direct and indirect effects of experimental warming on ecosystem carbon processes in a tallgrass prairie. Glob Biogeochem Cycles. 2005; 19: GB2014. https://doi.org/10.1029/2004GB002315
View Article
Google Scholar

[96] View Article

[97] Google Scholar

[ref33] 33. Graven HD, Keeling RF, Piper SC, Patra PK, Stephens BB, Wofsy SC, et al. Enhanced seasonal exchange of CO2 by northern ecosystems since 1960. Science. 2013; 341(6150): 1085–1089.https://doi.org/10.1126/science.1239207 pmid:23929948
View Article
PubMed/NCBI
Google Scholar

[99] View Article

[100] PubMed/NCBI

[101] Google Scholar

[ref34] 34. Intergovernmental Panel on Climate Change (IPCC). Climate Change 2013: The Physical Science Basis.Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change;Cambridge University Press: Cambridge, UK; New York, NY, USA, 2013.

[ref35] 35. Zhang CS, Hu Y,Shi XL. Analysis of Spatial-Temporal Evolution of vegetation cover in Loess Plateau in recent 33 years based on AVHRR NDVI and MODIS NDVI. Journal of applied sciences electronics and information engineering. 2006;34(6):702–712. https://doi.org/10.3969/j.issn.0255-8297.2016.06.006
View Article
Google Scholar

[104] View Article

[105] Google Scholar

[ref36] 36. Fensholt RP, Simon R. Evaluation of Earth Observation based global long term vegetation trends-Comparing GIMMS and MODIS global NDVI time series. Remote Sensing of Environment. 2012; 119:131–147. https://doi.org/10.1016/j.rse.2011.12.015
View Article
Google Scholar

[107] View Article

[108] Google Scholar

[ref37] 37. Melillo JM, Butler S, Johnson J, Mohan J, Steudler P, Lux H, et al. Soil warming, carbon–nitrogen interactions, and forest carbon budgets. Proceedings of the National Academy of Sciences. 2011;108:9508–9512. https://doi.org/10.1073/pnas.1018189108.
View Article
Google Scholar

[110] View Article

[111] Google Scholar

[ref38] 38. Robock A, Vinnikov KY, Srinivasan G, Entin JK, Hollinger SE, Speranskaya NA, et al.The global soil moisture data bank. Bulletin of the American Meteorological Society. 2000; 81(6): 1281–1299. https://doi.org/10.1175/1520-0477
View Article
Google Scholar

[113] View Article

[114] Google Scholar

[ref39] 39. Wan SQ, Xia JY, Liu WX, Niu SL. Photosynthetic overcompensation under nocturnal warming enhances grassland carbon sequestration. Ecology. 2009; 90(10): 2700–2710. https://doi.org/10.1890/08-2026.1 pmid:19886480
View Article
PubMed/NCBI
Google Scholar

[116] View Article

[117] PubMed/NCBI

[118] Google Scholar

[ref40] 40. Beier C, Emmett B, Gundersen P, Tietema A, Penuelas J, Estiarte M, et al. Novel approaches to study climate change effects on terrestrial ecosystems in the field: drought and passive nighttime warming. Ecosystems. 2004; 7:583–597. https://doi.org/10.1007/s10021-004-0178-8
View Article
Google Scholar

[120] View Article

[121] Google Scholar

[ref41] 41. Liang CL, Yu QZ, Liu YJ, Zhang Z. Effects of air temperature circadian on the NDVI of Nansi Lake wetland vegetation. Trop Geogr. 2015; 35(3):422–426. https://doi.org/10.13284/j.cnki.rddl.002746
View Article
Google Scholar

[123] View Article

[124] Google Scholar

[ref42] 42. Alward RD, Detling JK, Milchunas DG. Grassland vegetation changes and nocturnal global warming. Science. 1999; 283: 229–231. https://doi.org/10.1126/science.283.5399.229 pmid:9880257
View Article
PubMed/NCBI
Google Scholar

[126] View Article

[127] PubMed/NCBI

[128] Google Scholar

[ref43] 43. Morita S, Yonemaru J, Takanashi J. Grain growth and endosperm cell size under high night temperatures in rice (Oryza sativa L.). Ann Bot. 2005; 95: 695–701. https://doi.org/10.1093/aob/mci071 pmid:15655104
View Article
PubMed/NCBI
Google Scholar

[130] View Article

[131] PubMed/NCBI

[132] Google Scholar

[ref44] 44. Welch JR, Vincent JR, Auffhammer M, Moya PF, Dobermann A, Dawe D. Rice yields in tropical/subtropical Asia exhibit large but opposing sensitivities to minimum and maximum temperatures. Proc Natl Acad Sci. 2010; 107: 14562–14567. https://doi.org/10.1073/pnas.1001222107 pmid:20696908
View Article
PubMed/NCBI
Google Scholar

[134] View Article

[135] PubMed/NCBI

[136] Google Scholar

[ref45] 45. Nicholls N. Increased Australian wheat yield due to recent climate trends. Nature.1997;387:484–485. https://doi.org/10.1038/387484a0.
View Article
Google Scholar

[138] View Article

[139] Google Scholar

[ref46] 46. Yang Z, Jiang L, Su F, Zhang Q, Xia J, Wan S. Night-time warming enhances drought resistance of plant communities in a temperate steppe. Sci Rep. 2016; 6:23267. https://doi.org/10.1038/srep23267 pmid:26987482
View Article
PubMed/NCBI
Google Scholar

[141] View Article

[142] PubMed/NCBI

[143] Google Scholar

[ref47] 47. Atkin OK, Turnbull MH, Zaragoza-Castells J, Fyllas NM, Lloyd J, Meir P, et al. Light inhibition of leaf respiration as soil fertility declines along a post-glacial chronosequence in New Zealand, an analysis using the Kok method. Plant Soil. 2013; 367: 163–182. https://doi.org/10.1007/s11104-013-1686-0.
View Article
Google Scholar

[145] View Article

[146] Google Scholar

[ref48] 48. Zhao AZ, Zhang AB, Liu HX, Liu YX, Wang HF, Wang DL. Spatiotemporal variation of vegetation coverage before and after implementation of grain for Green Project in the Loess Plateau. J Nat Resour. 2017; 32: 449–460. https://doi.org/10.11849/zrzyxb.20160411
View Article
Google Scholar

[148] View Article

[149] Google Scholar

Figures

Abstract

1. Introduction

2. Materials and methods

2.1. Study area

2.2. Data and processing

2.2.1. Meteorological data and preprocessing.

2.2.2. GIMMS-NDVI3g and preprocessing.

2.2.3. Vegetation data and preprocessing.

2.3. Data analysis

2.3.1. Mann-Kendall.

2.3.2 Partial correlation analysis.

3. Results and analysis

3.1. Spatiotemporal patterns of asymmetric warming and NDVI

3.1.1. The regional analysis.

3.1.2. The pixel scale analysis.

3.2. Partial correlation between the NDVI and seasonal asymmetric warming

3.2.1. Region analysis.

3.2.2. The pixel scale analysis.

4. Discussions

5. Conclusion

Acknowledgments

References