1. Introduction
Climate change significantly impacts the activity and phenology of vegetation in the world (Rathcke and Lacey 1985; Chmielewski and Rötzer 2001; Piao et al. 2006; Wolkovich et al. 2012; Fu et al. 2015; Wang et al. 2019; Shen et al. 2021), which in turn affects the regional and global climate patterns (Peñuelas et al. 2009; Jeong et al. 2012; Richardson et al. 2013; Xu et al. 2014; Forzieri et al. 2017; Li et al. 2020; Zhang et al. 2021). Changes in the timing and magnitude of vegetation activity and phenology may amplify or dampen surface warming by affecting the surface energy budget and hydrological cycle (Bounoua et al. 2000; Jeong et al. 2012; Peng et al. 2014; Shen et al. 2016a; Li et al. 2016; Forzieri et al. 2017; Cao et al. 2019; Piao et al. 2019a). The vegetation–climate feedback is complicated because the changes in vegetation can affect many biophysical processes whose net effects on regional climate have high spatial heterogeneity (Liang et al. 2005; Jeong et al. 2012; Shen et al. 2016a). For instance, some studies found that increased vegetation productivity in the Arctic had a warming effect on temperature due to decreased surface albedo (Chapin et al. 2005; Pearson et al. 2013). In the Tibetan Plateau, however, vegetation greening attenuated the surface warming because of enhanced evapotranspiration (ET) (M. Shen et al. 2015). It is suggested that the evaporative cooling effect of increased vegetation activity on surface temperature was usually greater than the albedo warming effect in tropical and subtropical regions. Therefore, increased vegetation activity in these regions had a net cooling effect on surface temperature (Li et al. 2016; Forzieri et al. 2017). However, the net effect of vegetation activity change on surface temperature in temperate regions of the Northern Hemisphere is controversial (Piao et al. 2019b). Some studies argued that the increase of vegetation activity slowed down climate warming through the evaporative cooling effect in temperate regions like East Asia (Jeong et al. 2009). Other studies suggested that the effect of increasing vegetation activity on surface temperature in temperate regions of the Northern Hemisphere was dominated by albedo warming (Lee et al. 2011; Alkama and Cescatti 2016), and that earlier leaf-out enhanced surface warming in northern temperate regions (Xu et al. 2020). Because the influence of vegetation on the surface temperature in the northern temperate region is still subject to debate, it is necessary to further explore the impact of vegetation change on the climate in temperate regions of the Northern Hemisphere. In addition, it is suggested that vegetation has different effects on climate in different seasons. For example, some studies found that the increase of vegetation activity in the growing season often weakened the surface warming through biophysical feedback (Zeng et al. 2017), while the increase of vegetation activity in spring enhanced the surface warming in northern boreal and Arctic regions (Xu et al. 2020). Considering the complexity of the climate system, the impacts of vegetation change on regional climate in different seasons in temperate regions of the Northern Hemisphere need to be further explored. In the context of climate change, quantifying the feedback effects of vegetation activity and phenology changes on regional climate is essential to constructing accurate regional climate simulations (Peñuelas et al. 2009; Xu et al. 2020).
Temperate grasslands are one of the world’s great biomes, covering about 8% of Earth’s terrestrial surface (Carbutt et al. 2017). Temperate grasslands are highly sensitive to climate change, making them one of the most threatened ecosystems in the world (Chen et al. 2019). Climate warming is reported to enhance vegetation growth, advance the start of the growing season (SOS), delay the end of the growing season (EOS), and increase the length of the growing season (LOS) in temperate grasslands (Shen et al. 2018). The increased vegetation activity and lengthening of the growing season may attenuate warming by enhancing evapotranspiration (ET) (Bounoua et al. 2000; Jeong et al. 2009; Lee et al. 2011; M. Shen et al. 2015) but also possibly amplify warming by decreasing albedo (Chapin et al. 2005; Pearson et al. 2013). Feedback cycles between vegetation activity and growing season changes become complicated due to these opposing effects on the surface energy budget. Until recently, the possible effects of vegetation activity and growing season changes on regional climate patterns in temperate grassland regions have remained unclear.
Because there is no reliable method for separating the effects of local climate change from those of global climate change, it is difficult to quantify the feedback effects of land surface properties on the local climate change. Kalnay and Cai (2003) proposed an observation minus reanalysis (OMR) method to quantify the impacts of land surface properties on regional and global climate change. This OMR approach is effective because it compares surface observational climate data with data estimated from upper-atmospheric measurements that are not sensitive to land surface properties (Kalnay and Cai 2003). As a result, the differences from the surface observations in the reanalysis can be attributed to the land surface impacts (Kalnay and Cai 2003; Lim et al. 2005). Due to its good performance, the OMR approach has been widely used in quantifying the effects of change in land properties including vegetation on regional and global climate change in the past few years (Fall et al. 2010; Hu et al. 2010; Yang et al. 2011; Li et al. 2013; Wang et al. 2014; Shen et al. 2016a, 2017; Cao et al. 2019; T. Liu et al. 2019; Nayak and Mandal 2019; Jin et al. 2020; Prijith et al. 2020). In vegetation-covered regions, many studies have used the OMR method with reanalysis data to assess temperature sensitivity to vegetation cover change in terms of surface vegetation index (Lim et al. 2008; Yang et al. 2010; X. Shen et al. 2015; Shen 2016a,b; Jin et al. 2020). The basis of these studies is the fact that the response of surface temperature to vegetation change is not present in the National Centers for Environmental Prediction (NCEP) reanalysis data, while the surface observations data include not only local surface forcings but the large-scale atmospheric forcings (e.g., greenhouse gases, natural decadal variability, and volcanoes) (Pielke et al. 2002; Kalnay and Cai 2003; Zhou et al. 2004; Lim et al. 2005; Kalnay et al. 2006). In other words, natural climate variability caused by changes in atmospheric circulation and surface vegetation is included in both surface observations and reanalysis. Nonetheless, the vegetation change effect is captured only by surface observations (X. Shen et al. 2015; Shen et al. 2016a). As a result, the OMR trends can give an estimate of the surface temperature change signal arising from vegetation changes (Lim et al. 2008; Yang et al. 2010; X. Shen et al. 2015; Jin et al. 2020).
In this study, using the normalized difference vegetation index (NDVI) data, we first investigated the spatiotemporal changes of NDVI and vegetation phenology (including SOS, EOS, and LOS) in temperate grasslands of China. Then, we used the OMR method to investigate the relationships of surface air temperatures with NDVI and growing season duration in temperate grasslands of China. Finally, we analyzed the changes of albedo and ET induced by vegetation changes to further explain the possible effects of vegetation activity and phenology changes on seasonal surface air temperatures in temperate grasslands of China.
2. Materials and methods
a. Study area
According to China’s vegetation regionalization, temperate grasslands are mainly distributed in the temperate grassland region of China (Fig. 1). Based on the vegetation regionalization map of China (X. Shen et al. 2015), this study focused on temperate grasslands in the temperate grassland region of China (Fig. S1 in the online supplemental material), which is mainly distributed across the Loess Plateau, the Inner Mongolian Plateau, and the Songliao Plain. The elevation of the study area ranges from 60 to 4441 m, with high elevations in the west and low elevations in the east (Shen et al. 2018). The climate in this region varies from the east semiarid climate to the west arid climate (Shen et al. 2018). Annual precipitation in this grassland region varies from 35 to 530 mm, and annual average temperatures ranges from −5° to 10°C (Chuai et al. 2013).
Spatial distributions of temperate grasslands and weather stations in the temperate grassland region of China.
Citation: Journal of Climate 35, 15; 10.1175/JCLI-D-21-0325.1
b. Data
Data used in this study include the following. 1) The Global Inventory Monitoring and Modeling Studies (GIMMS) NDVI data (GIMMS NDVI3g) from 1982 to 2015. Derived from AVHRR satellite imagery, this dataset has been processed with radiometric, geometric, and atmospheric corrections to a spatial resolution of 8 km × 8 km and a temporal resolution of 15 days (Tucker et al. 2005). In addition, we used the Moderate-Resolution Imaging Spectroradiometer (MODIS) NDVI (MOD13Q1) data from 2001 to 2015 to further test the results. The spatial and temporal resolution of MODIS NDVI data are 250 m and 16 days, respectively. 2) Monthly surface air temperature, including mean temperature (Tmean), maximum temperature (Tmax), and minimum temperature (Tmin) data from surface observations and reanalysis during 1982 to 2015. The observational temperature data are obtained from 77 weather stations distributed across the study area. The quality and consistency of these observational data have been guaranteed based on vigorous data assurance policy (Shen et al. 2016a). In addition, we also use gridded climate data including monthly Tmean, Tmax, and Tmin from 1982 to 2015 to test the results in this study. These gridded climate data have a spatial resolution of 0.5° × 0.5° and are generated from more than 800 weather stations throughout China (Shen et al. 2018). Reanalysis temperature data based on atmospheric measurements from 1982 to 2015 are obtained from the National Centers for Environmental Prediction at the U.S. Department of Energy (NCEP–DOE Reanalysis II) (Kanamitsu et al. 2002). These data have a spatial resolution of about 1.9°. 3) Two periods (the 1980s and 2015) of land use and land cover (LULC) maps covering the study area are provided by the Resource and Environment Data Cloud Platform of China. These two periods of LULC maps have the same spatial resolution of 100 m and are classified into six LULC categories: cultivated land, grassland, forest land, water area, built-up land, and barren land. 4) The white-sky albedo data at 8-day intervals from 1982 to 2015 are derived from the Global Land Surface Satellite (GLASS) products (Liang et al. 2021). These albedo data, with a spatial resolution of 0.05°, are aggregated to monthly values for each year by calculating the arithmetic mean of values within each calendar month (Li et al. 2016). 5) Monthly evapotranspiration data at 0.25° spatial resolution from 1982 to 2015 are obtained from the Global Land Evaporation Amsterdam Model (GLEAM) ET product (https://www.gleam.eu) (Martens et al. 2017). For this study, the land use map and climate data are resampled using the nearest neighbor resampling method to the same spatial resolution as NDVI data (Shen et al. 2018).
c. Methods
We used the maximum value compositing (MVC) method to produce a monthly NDVI dataset from the original NDVI data (Pettorelli et al. 2005). To reduce the influence of land use and land cover change on our results, we extracted the pixels classified as grasslands in both the 1980s and 2015 maps for use in our analysis. Specifically, we first extracted the grassland patches from two LULC maps by setting other land use categories to null, and then made a raster subtraction calculation to obtain the unchanged patches from the resulting raster (X. Shen et al. 2015). We identified 33 weather stations located in these unchanged grasslands within the study area. The corresponding NDVI value for each weather station was derived from the average value in a 3 × 3 pixel window around this station (X. Shen et al. 2015). Similarly, based on the gridded climate data and NDVI data, we extracted all the grid points in the unchanged grasslands within the study area. The NDVI value for each grid point was the value of corresponding NDVI pixel. We calculated the mean growing season NDVI (hereinafter referred to as the growing season NDVI) as the average value of the NDVI from April to October (Shen et al. 2018); we further calculated seasonal average values for spring (April and May), summer (June–August), and autumn (September and October).
Using the OMR approach (observation minus reanalysis), we calculated the difference between observation and reanalysis trends of temperature (Kalnay and Cai 2003). We followed these steps for each weather station and grid point: 1) estimating the reanalysis temperatures using a linear spatial interpolation; 2) computing monthly temperature anomalies from the 34-yr (1982–2015) mean value for both observations and reanalysis (Kalnay and Cai 2003; Lim et al. 2005; X. Shen et al. 2015); 3) calculating the monthly OMR temperature and OMR temperature trend for each weather station and grid point using calculated monthly temperature anomalies of observation and reanalysis (Shen et al. 2016a; Shen et al. 2017) (the reason for using anomalies rather than absolute values is to reduce the impact of altitude on the results; Simmons et al. 2004); and 4) identifying trends in temperature and NDVI by simple linear regression and applying the Mann–Kendall (MK) test (Mann 1945; Kendall 1975).
3. Results and discussion
a. Variations in vegetation and surface air temperatures
During the period of 1982–2015, mean growing season NDVI showed an increasing trend in most of the temperate grassland areas (Fig. 2). The most pronounced increase of growing season NDVI was found along the southern boundary of the study area. The increasing NDVI trend held across different seasons in most areas. A weak decreasing trend in NDVI was mainly concentrated in northeastern portions of the study area during spring and summer, and in the central region during autumn (Fig. 2). The average NDVI of the study area significantly increased by 0.011 decade−1 for the growing season from 1982 to 2015 (Fig. 3). Seasonally, the average rates of increase in NDVI for spring, summer, and autumn from 1982 to 2015 were 0.009, 0.011, and 0.014 decade−1, respectively. These results showed that vegetation is becoming greener in temperate grasslands of China during the past decades. Based on the MODIS NDVI data, we found that the average NDVI of the study area increased by 0.025, 0.040, and 0.003 decade−1 for spring, summer, and autumn, respectively, from 2001 to 2015. These MODIS NDVI trends were basically consistent with the corresponding GMMIS NDVI trends for spring (0.023 decade−1), summer (0.038 decade−1), and autumn (0.004 decade−1) over the temperate grasslands of China from 2001 to 2015. The spatial patterns of MODIS NDVI trends were also similar to that of MODIS NDVI trends in temperate grasslands of China from 2001 to 2015 (data not shown). Our results confirmed many studies that, in terms of interannual variability and temporal trend, GIMMS NDVI and MODIS NDVI presented high consistency in temperate grasslands of China (Fensholt et al. 2009; Chen et al. 2020).
Spatial distributions of temporal trends in (a) growing season NDVI, (b) spring NDVI, (c) summer NDVI, and (d) autumn NDVI in temperate grasslands of China from 1982 to 2015. The inset maps show significant (P < 0.05) increasing trends (blue) and significant (P < 0.05) decreasing trends (red).
Citation: Journal of Climate 35, 15; 10.1175/JCLI-D-21-0325.1
Time series of (a) growing season NDVI, (b) spring NDVI, (c) summer NDVI, and (d) autumn NDVI in temperate grasslands of China from 1982 to 2015.
Citation: Journal of Climate 35, 15; 10.1175/JCLI-D-21-0325.1
During the study period of 1982–2015, the SOS occurred between 114 and 126 DOY, and the EOS occurred between 275 and 292 DOY across temperate grasslands of China. On average, the SOS and EOS for temperate grasslands of China occurred on 120 DOY (corresponding to 30 April, or 29 April in leap years) and 284 DOY (corresponding to 11 October, or 10 October in leap years), respectively. As a result, the average LOS for the study area was about 164 days during the study period. Our results show changes in the timing and length of the growing season across most of the study area, with earlier SOS, later EOS, and longer LOS (Fig. 4; the smooth line shows values calculated from a nine-point binomial filter). The most obvious extension of the growing season was mainly concentrated in the southwestern portions of the study area, while the central portions showed slight decreases in LOS. Previous studies suggested that the decrease of precipitation during the growing season of 1982–2015 could account for the decreased LOS (and NDVI) in the central region of study area (Q. Liu et al. 2019; Sun et al. 2021). The regional average LOS across temperate grasslands of China has increased at a rate of 3.45 days decade−1 from 1982 to 2015, driven by the significant trends toward earlier SOS (−1.84 days decade−1; P < 0.01) and later EOS (1.61 days decade−1; P < 0.01) during the same period. From 2001 to 2015, the phenological dates calculated from the GIMMS NDVI and MODIS NDVI data showed similar trends over the temperate grasslands of China. Consistent with findings from many other studies (Piao et al. 2006; Shen et al. 2018; Zhang et al. 2020), this study found an extension of the growing season in temperate grasslands of China during the past decades, which is one of the key aspects of climate change and climate extremes (Yang et al. 2015; Liu et al. 2016; Ying et al. 2020).
Spatial distributions of temporal trends and time series of (a),(b) the start date of the growing season (SOS), (c),(d) the end date of the growing season (EOS), and (e),(f) the length of the growing season (LOS) in temperate grasslands of China from 1982 to 2015. The smooth line shows the smoothing values calculated from a nine-point binomial filter.
Citation: Journal of Climate 35, 15; 10.1175/JCLI-D-21-0325.1
Previous studies suggest that the greening we observe in temperate grasslands of China may be related to increases in surface temperatures from global warming, increasing precipitation, higher levels of atmospheric CO2, or enhanced rates of nitrogen deposition, as well as land use change (Piao et al. 2006; Forzieri et al. 2017). To examine the changes in air temperatures, we compared the spring, autumn, and growing season mean temperature anomalies from surface observations and the reanalysis averaged over all the 33 stations in our study area from 1982 to 2015. We found a good agreement between the observed and reanalyzed temperature anomalies (Fig. 5), confirming that the reanalysis temperatures capture the interannual fluctuations of observed temperatures very well (Kalnay and Cai 2003; X. Shen et al. 2015). Both the observed and reanalyzed surface temperature increased significantly from 1982 to 2015, with a larger warming rate found in the surface observations than in the reanalysis. Therefore, the OMR time series of temperatures show a positive trend (Fig. 5). Based on the principle of the OMR method, it suggests that changes in temperate grasslands of China generally correspond to an increase of the surface temperature in the study area. On average, the OMR trends in spring, summer, autumn, and growing season Tmean were all positive at rates of 0.38°, 0.24°, 0.26°, and 0.29°C decade−1, respectively, from 1982 to 2015. For the daytime or nighttime temperatures, the OMR trends in spring, summer, autumn, and growing season Tmax (Tmin) were 0.39° (0.36°), 0.25° (0.22°), 0.28° (0.23°), and 0.31°C decade−1 (0.27°C decade−1), respectively.
Time series of observational, reanalysis, and observation minus reanalysis (OMR) mean temperatures in temperate grasslands of China from 1982 to 2015. The smooth line shows the smoothing values calculated from a nine-point binomial filter.
Citation: Journal of Climate 35, 15; 10.1175/JCLI-D-21-0325.1
b. Effects of vegetation changes on surface air temperatures
To investigate the possible effects of vegetation changes on surface air temperatures, we analyzed the relationships between the OMR trends in mean temperature (OMRtrend-Tmean), maximum temperature (OMRtrend-Tmax), and minimum temperature (OMRtrend-Tmin) and the temporal trends of NDVI (NDVItrend), SOS (SOStrend), EOS (EOStrend), and LOS (LOStrend) at the 33 weather stations in temperate grasslands of China from 1982 to 2015. The results showed that the correlation between growing season OMRtrend-Tmean and growing season NDVItrend was significantly positive (Fig. 6a), which indicates that vegetation greening is associated with warming of surface air temperatures during the growing season. The relationships of growing season NDVItrend with growing season OMRtrend-Tmax and OMRtrend-Tmin were significantly positive (Figs. S2a and S3a), confirming that vegetation greening had warming effects on surface air temperatures during the growing season. Figure 6 shows the correlation coefficient (R) between the OMR temperature trend of Tmean and the trend of NDVI by season; asterisks indicate significance (*P < 0.05; **P < 0.01). The correlations for spring and autumn are significantly positive, but the summer correlation was not significant (P > 0.05). Similar results were found for both Tmax and Tmin, with larger correlation coefficients for Tmax than Tmin (Figs. S2 and S3). To test these results, we also calculated the correlations between the OMR trends in temperatures and the temporal trends of NDVI and phenological dates across all the grid points in temperate grasslands of China (Table S1). Correlation results confirmed that there were significant positive relationships of NDVItrend with OMRtrend-Tmax and OMRtrend-Tmin in both spring and autumn, but no significant correlations between NDVItrend and OMRtrend-Tmax or OMRtrend-Tmin in summer. These results indicated that the increased vegetation activity of temperate grasslands of China had warming effects on spring and autumn temperatures, but had no significant effects on summer temperatures. Consistent with the results based on the weather stations (Figs. S2 and S3), results from gridded climate data confirmed that the warming effects of vegetation greening on Tmax were larger than that on Tmin in spring and autumn (Table S1). These findings were further confirmed by the correlation results during 2001–15 (Figs. S4, S5, and S6; Table S2) calculated using the MODIS NDVI data.
Spatial relationship of observation minus reanalysis (OMR) mean temperature trends with trends of (a) growing season NDVI, (b) spring NDVI, (c) summer NDVI, and (d) autumn NDVI across the 33 weather stations in temperate grasslands of China from 1982 to 2015. Each point is for one weather station; R is the correlation coefficient between the OMR temperature trend and the trend of NDVI. One asterisk (*) indicates P < 0.05; two asterisks (**) indicate P < 0.01. Correlations with no asterisk are not significant (P > 0.05).
Citation: Journal of Climate 35, 15; 10.1175/JCLI-D-21-0325.1
Figure 7 shows the correlation coefficients (R) between the OMR temperature trends in Tmean and the trends of phenological dates by season from 1982 to 2015; asterisks indicate significance (*P < 0.05; **P < 0.01). The correlation between spring OMRtrend-Tmean and SOStrend was negative while the correlation between autumn OMRtrend-Tmean and EOStrend was positive, suggesting that both the earlier SOS and later EOS could have the effect of warming surface air temperatures. Again, there was no significant relationship with summer OMRtrend-Tmean and the growing season parameters (P > 0.05), suggesting that vegetation phenology changes has no significant effects on summer air temperatures (or vice versa). In addition, we found that the relationships of OMRtrend-Tmax and OMRtrend-Tmin with the trends in phenological dates (Figs. S7 and S8) were generally consistent with the corresponding relationships between OMRtrend-Tmean and trends in phenological dates. These findings were further confirmed by the correlation results from the MODIS NDVI (Figs. S9, S10, and S11) and the grid points (Tables S1 and S2). The significant positive correlation between growing season OMR temperature trends and LOStrend suggests that the growing season extension can have the effect of increasing growing season surface air temperatures in temperate grasslands of China.
Spatial relationship of observation minus reanalysis (OMR) mean temperature trends with trends of the start date of the growing season (SOS), the end date of the growing season (EOS), and the length of the growing season (LOS) across the 33 weather stations in temperate grasslands of China from 1982 to 2015. Each point is for one weather station; R is the correlation coefficient between the OMR temperature trend and the trend of phenological dates; *P < 0.05; **P < 0.01. Correlations with no asterisk are not significant (P > 0.05).
Citation: Journal of Climate 35, 15; 10.1175/JCLI-D-21-0325.1
Increased vegetation greening and extended growing seasons in temperate grasslands may have the effect of increasing evapotranspiration (X. Shen et al. 2015) while decreasing albedo (Richardson et al. 2013), which have opposite effects on surface temperatures. The effects of ET change on surface temperatures can be masked by a stronger effect of albedo, or vice versa (Peng et al. 2014; Zhou et al. 2016; Shen et al. 2020). In this study, the significant positive correlation of growing season OMRtrend with NDVItrend and LOStrend suggests that the increased greening and extended growing season acts as a significant positive feedback to regional warming during the growing season. The increased greening and extended growing season could decrease the albedo, resulting in a warming growing season temperature due to greater absorption of incoming shortwave radiation.
To confirm this mechanism, we first examined the correlations among NDVItrend, SOStrend, or EOStrend, and the temporal trend in albedo (albedotrend) during the study period of 1982–2015 across the 33 weather stations in temperate grasslands of China (Figs. 8 and 9). The results showed that the spatial patterns of NDVItrend were significantly and negatively correlated with albedotrend during the growing season (P < 0.01; Fig. 8). As Fig. 9 shows, the spatial patterns of SOStrend are positively correlated with spring albedotrend (P < 0.01) while the patterns of EOStrend are significantly and negatively correlated with autumn albedotrend (P < 0.01). It suggests that both the earlier SOS and the delayed EOS could significantly reduce albedo associated with vegetation cover in spring and autumn, respectively. By contrast, summer albedotrend displayed no significant relationship with SOStrend, EOStrend, or LOStrend (P > 0.05), suggesting that growing season duration change has no significant effect on summer albedo in temperate grasslands of China. Considering the entire growing season, we found significant negative correlations between LOStrend and growing season albedotrend (P < 0.01).
Correlation coefficients between temporal trends in NDVI and trends of (a) albedo and (b) evapotranspiration (ET) during 1982–2015 across the 33 weather stations in temperate grasslands of China; *P < 0.05; **P < 0.01. Correlations with no asterisk are not significant (P > 0.05).
Citation: Journal of Climate 35, 15; 10.1175/JCLI-D-21-0325.1
Correlation coefficients between temporal trends in phenological indices [the start date of the growing season (SOS), the end date of the growing season (EOS), and the length of the growing season (LOS)] and trends of albedo and evapotranspiration (ET) during 1982–2015 across the 33 weather stations in temperate grasslands of China; *P < 0.05; **P < 0.01. Correlations with no asterisk are not significant (P > 0.05).
Citation: Journal of Climate 35, 15; 10.1175/JCLI-D-21-0325.1
To further explain the effects of vegetation activity and phenology changes on surface temperatures, we also analyzed the relationships between the albedotrend and OMR temperature trends for the period of 1982–2015. During the growing season, we found significantly negative correlations between albedotrend and OMR trends in Tmean, Tmax, and Tmin (Fig. 10a; Figs. S12a and S13a), and the correlations between albedotrend and OMRtrend-Tmax were larger than the corresponding correlations between albedotrend and OMRtrend-Tmin (Figs. S12a and S13a). In general, land surface absorbs and stores solar energy during the daytime and releases energy at night. Therefore, nighttime surface temperature is closely associated with the temperature during the daytime (Shen et al. 2020). Our results implied that decreased albedo had significant warming effects on surface air temperature during the daytime, and the increase of temperature during the daytime could be extended to nighttime. This could explain the larger effects of vegetation changes on Tmax as compared to Tmin in spring and autumn (Figs. S2, S3, S7, and S8). These results confirmed that the vegetation greening and growing season extension could decrease albedo, which, in turn, resulted in a warming of surface air temperature in temperate grasslands of China.
Correlation coefficients between temporal trends in OMR mean temperatures and the trends of (a) albedo and (b) evapotranspiration (ET) during 1982–2015 across the 33 weather stations in temperate grasslands of China; *P < 0.05; **P < 0.01. Correlations with no asterisk are not significant (P > 0.05).
Citation: Journal of Climate 35, 15; 10.1175/JCLI-D-21-0325.1
A previous study of grasslands in the Tibetan Plateau found that enhancing vegetation growth could cool surface air temperature by increasing evapotranspiration (ET) (M. Shen et al. 2015). Compared with the cold and dry Tibetan Plateau, the temperate grassland region of China is warmer and drier (Shen et al. 2018). In contrast to the findings from the Tibetan Plateau, we found no significant correlations between growing season NDVItrend and the temporal trend in growing season ET (ETtrend) in temperate grassland region from 1982 to 2015 (Fig. 8). The correlation between NDVItrend and ETtrend was significant in summer, but not significant in spring and autumn. This suggests that the ET change induced by vegetation greening is strong in summer but weak in spring and autumn. This may be because soil moisture content is relatively low in the temperate grassland region, especially in spring and autumn, so ET changes less in response to vegetation change (Shen et al. 2018). We found that the growing season correlations between ETtrend and OMR temperature trends were weakly negative (Fig. 10b; see also Figs. S12b and S13b). Therefore, the weak correlation between ETtrend and OMR temperature trends facilitates the albedo-induced warming effects of the growing season extension on temperature. In different seasons, the significant correlations between ETtrend and OMR temperature trends were only found for the summer correlation between ETtrend and OMRtrend-Tmax, as well as the summer correlation between ETtrend and OMRtrend-Tmean (Fig. 10b and Fig. S12b). It confirms that ET is an important influencing factor for daytime temperature but not for nighttime temperature (Shen et al. 2020). In summer, which is the rainy season in temperate grasslands of China, the enhanced vegetation growth could result in the increasing ET due to relatively abundant soil water content (Shen et al. 2018). Therefore, we can conclude that the cooling effect of increased ET is balanced by the warming effect of decreased albedo resulting from vegetation greening, resulting in no significant effect on summer temperatures in temperate grasslands of China (Figs. 8 and 10). Although some studies found no significant or cooling effects of vegetation greening on temperature in temperate regions (Jeong et al. 2009; Forzieri et al. 2017; Zeng et al. 2017), our finding was consistent with other studies that found that increasing vegetation activity had a warming effect on surface temperature in temperate regions of the Northern Hemisphere (Lee et al. 2011; Alkama and Cescatti 2016). The differences in the research methods, the background climatic condition, and the extent of greening could explain the different results in the studies (Forzieri et al. 2017).
Other studies in northern ecosystems have reported that the lengthening of the vegetation growing season can strongly enhance vegetation growth (Myneni et al. 1997; Nemani et al. 2003). In temperate grasslands of China, longer growing seasons are regarded as playing an important role in enhancing vegetation growth (Zhou et al. 2020). Therefore, one would expect an increased ET to accompany the extension of the growing season that we observe. However, we found no significant correlations between LOStrend and the temporal trend in ET (ETtrend) during the growing season (Fig. 9). If change in ET is induced by vegetation change here, the signal is weak. Again, this may be due to the relatively low soil moisture content in the temperate grassland region, with the result that vegetation change has a smaller effect on ET change (Shen et al. 2018). In any case, the weak correlation between ETtrend and OMRtrend-Tmean fails to overcome the albedo-induced warming effects of the growing season extension on temperatures.
4. Conclusions
Accurately assessing the impacts of land surface vegetation change on regional climate is important for predicting future climate change and mitigating global warming. This study analyzed, for the first time, the effects of changes in both vegetation activity and phenology on surface air temperatures in temperate grasslands of China. We found that increased vegetation greening and extended growing seasons could have the effect of warming surface air temperatures in temperate grasslands of China, implying that the climate change–induced increasing vegetation activity can further aggravate regional warming. In addition, this study first revealed that the changes in vegetation activity and phenology had obvious effects on spring and autumn air temperatures, whereas summer greening and phenological change had no significant effect on summer temperatures in temperate grasslands of China. Our results suggest that the effects of vegetation activity and phenology changes on regional climate should be considered in regional or global climate models if they are to accurately simulate regional climate change in temperate grasslands. More attention should be paid to the climate feedbacks of vegetation changes, especially considering how these climate feedbacks are likely to further impact vegetation and ecosystem properties during the growing season.
We should note that there may be some limitations or uncertainties in this study. First, the NDVI data could contain some uncertainties due to the effects of atmosphere, clouds, and solar angle on remotely sensed data. Second, each NDVI pixel can only roughly reflect the information about vegetation within a certain area. Another limitation of the study is that MVC method, although very efficient at minimizing cloud cover and various other noise issues, eliminates temporal data (Holben 1986). Moreover, pixels are acquired at different times and different viewing geometries, under different atmospheric conditions. Third, although as in most previous studies we attributed the OMR temperature differences to the vegetation effects, temperature as well as other climatic variables obviously has direct effects on vegetation. The OMR method may have some uncertainty in data and assumptions (Wang et al. 2018), and the OMR differences may be affected by some other factors, including the interpolation methods and possible inaccuracies related to both the observational and reanalysis temperature data in the study. At the same time, we should note that the OMR results obtained depend partly on the dataset used. While NCEP and other similar resolution reanalysis temperature trends have often been found to be cooler than observed surface temperature trends, the opposite may be observed when higher-resolution reanalysis data are used. Further studies will be needed to reduce the range of uncertainty about the effects of vegetation change on surface air temperatures in temperate grasslands of China. Fourth, although this study focused on the local impacts of vegetation changes on climate, the large-scale change in grassland vegetation could trigger nonlocal climate impacts (Winckler et al. 2019; Xu et al. 2020). Therefore, the effects of large-scale changes in grassland vegetation on climate need to be further explored in future studies.
Acknowledgments.
We gratefully acknowledge the National Key Research and Development Program of China (2019YFC0409101), National Natural Science Foundation of China (41971065), Youth Innovation Promotion Association, CAS (2019235), Natural Science Foundation of Jilin Province (20210101104JC), and Key Research Program of Frontier Sciences, CAS (ZDBS-LY-7019) for funding this work.
Data availability statement.
The GIMMS NDVI data were provided by the U.S. National Center for Atmospheric Research (accessed June 2018 from http://ecocast.arc.nasa.gov/data/pub/gimms/3g.v1; available now at https://data.tpdc.ac.cn/en/data/9775f2b4-7370-4e5e-a537-3482c9a83d88/). The MODIS NDVI data were obtained from the Level-1 and Atmosphere Archive and Distribution System (https://ladsweb.modaps.eosdis.nasa.gov/search/order/1/MOD13Q1–6; access date: May 2019). The observational station and gridded temperature data were provided by National Meteorological Information Center through the China Meteorological Data Sharing Service Center (CMDC) (http://data.cma.cn/en; access date: October 2019). Reanalysis temperature data (NCEP-DOE Reanalysis II) originate from the NOAA Earth System Research Laboratory’s Physical Sciences Division (https://psl.noaa.gov/data/gridded/data.ncep.reanalysis2.html; access date: January 2020). Land use and land cover (LULC) maps were provided by the Resource and Environment Data Cloud Platform of China (www.resdc.cn; access date: March 2019). The surface shortwave albedo data were downloaded from the Chinese National Earth System Science Data Sharing Infrastructure initiative of the National Science and Technology Infrastructure (www.geodata.cn/data; access date: February 2020). Monthly evapotranspiration data were obtained from the Global Land Evaporation Amsterdam Model ET product (https://www.gleam.eu; access date: February 2020).
REFERENCES
Alkama, R., and A. Cescatti, 2016: Biophysical climate impacts of recent changes in global forest cover. Science, 351, 600–604, https://doi.org/10.1126/science.aac8083.
Bounoua, L., and Coauthors, 2000: Sensitivity of climate to changes in NDVI. J. Climate, 13, 2277–2292, https://doi.org/10.1175/1520-0442(2000)013<2277:SOCTCI>2.0.CO;2.
Cao, Q., J. Wu, D. Yu, and W. Wang, 2019: The biophysical effects of the vegetation restoration program on regional climate metrics in the Loess Plateau, China. Agric. For. Meteor., 268, 169–180, https://doi.org/10.1016/j.agrformet.2019.01.022.
Carbutt, C., W. D. Henwood, and L. A. Gilfedder, 2017: Global plight of native temperate grasslands: Going, going, gone? Biodivers. Conserv., 26, 2911–2932, https://doi.org/10.1007/s10531-017-1398-5.
Chapin, F. S., and Coauthors, 2005: Role of land-surface changes in Arctic summer warming. Science, 310, 657–660, https://doi.org/10.1126/science.1117368.
Chen, R., Z. Hu, S. Li, and Q. Guo, 2020: Assessment of normalized difference vegetation index from different data sources in grassland of northern China. J. Geoinfo. Sci., 22, 1910–1919, https://doi.org/10.12082/dqxxkx.2020.190237.
Chen, T., A. Bao, G. Jiapaer, H. Guo, G. Zheng, L. Jiang, C. Chang, and L. Tuerhanjiang, 2019: Disentangling the relative impacts of climate change and human activities on arid and semiarid grasslands in Central Asia during 1982–2015. Sci. Total Environ., 653, 1311–1325, https://doi.org/10.1016/j.scitotenv.2018.11.058.
Chmielewski, F. M., and T. Rötzer, 2001: Response of tree phenology to climate change across Europe. Agric. For. Meteor., 108, 101–112, https://doi.org/10.1016/S0168-1923(01)00233-7.
Chuai, X., X. Huang, W. Wang, and G. Bao, 2013: NDVI, temperature and precipitation changes and their relationships with different vegetation types during 1998–2007 in Inner Mongolia, China. Int. J. Climatol., 33 1696–1706, https://doi.org/10.1002/joc.3543.
Cong, N., and Coauthors, 2012: Spring vegetation green-up in China inferred from SPOT NDVI data: A multiple model analysis. Agric. For. Meteor., 165, 104–113, https://doi.org/10.1016/j.agrformet.2012.06.009.
Cong, N., T. Wang, H. Nan, Y. Ma, X. Wang, R. B. Myneni, and S. Piao, 2013: Changes in satellite-derived spring vegetation green-up date and its linkage to climate in China from 1982 to 2010: A multimethod analysis. Global Change Biol., 19, 881–891, https://doi.org/10.1111/gcb.12077.
Fall, S., D. Niyogi, A. Gluhovsky, R. Pielke, E. Kalnaye, and G. Rochonf, 2010: Impacts of land use land cover on temperature trends over the continental United States: Assessment using the North American Regional Reanalysis. Int. J. Climatol., 30, 1980–1993, https://doi.org/10.1002/joc.1996.
Fensholt, R., K. Rasmussen, T. T. Nielsen, and C. Mbow, 2009: Evaluation of Earth observation based long term vegetation trends—Intercomparing NDVI time series trend analysis consistency of Sahel from AVHRR GIMMS, Terra MODIS and SPOT VGT data. Remote Sens. Environ., 113, 1886–1898, https://doi.org/10.1016/j.rse.2009.04.004.
Forzieri, G., R. Alkama, D. G. Miralles, and A. Cescatti, 2017: Satellites reveal contrasting responses of regional climate to the widespread greening of Earth. Science, 356, 1180–1184, https://doi.org/10.1126/science.aal1727.
Fu, Y. H., S. Piao, M. Op de Beeck, N. Cong, H. Zhao, Y. Zhang, A. Menzel, and I. A. Janssens, 2014: Recent spring phenology shifts in western Central Europe based on multiscale observations. Global Ecol. Biogeogr., 23, 1255–1263, https://doi.org/10.1111/geb.12210.
Fu, Y. H., and Coauthors, 2015: Declining global warming effects on the phenology of spring leaf unfolding. Nature, 526, 104–107, https://doi.org/10.1038/nature15402.
Holben, B. N., 1986: Characteristics of maximum-value composite images from temporal AVHRR data. Int. J. Remote Sens., 7, 1417–1434, https://doi.org/10.1080/01431168608948945.
Hu, Y., W. Dong, and Y. He, 2010: Impact of land surface forcings on mean and extreme temperature in eastern China. J. Geophys. Res., 115, D19117, https://doi.org/10.1029/2009JD013368.
Jeong, J. H., and Coauthors, 2012: Greening in the circumpolar high-latitude may amplify warming in the growing season. Climate Dyn., 38, 1421–1431, https://doi.org/10.1007/s00382-011-1142-x.
Jeong, S. J., C. H. Ho, K. Y. Kim, and J. H. Jeong, 2009: Reduction of spring warming over East Asia associated with vegetation feedback. Geophys. Res. Lett., 36, L18705, https://doi.org/10.1029/2009GL039114.
Jin, K., F. Wang, Q. Zong, P. Qin, and C. Liu, 2020: Impact of variations in vegetation on surface air temperature change over the Chinese Loess Plateau. Sci. Total Environ., 716, 136967, https://doi.org/10.1016/j.scitotenv.2020.136967.
Kalnay, E., and M. Cai, 2003: Impact of urbanization and land-use change on climate. Nature, 423, 528–531, https://doi.org/10.1038/nature01675.
Kalnay, E., M. Cai, H. Li, and J. Tobin, 2006: Estimation of the impact of land-surface forcings on temperature trends in eastern United States. J. Geophys. Res., 111, D06106, https://doi.org/10.1029/2005JD006555.
Kanamitsu, M., W. Ebisuzaki, J. Woollen, S. K. Yang, J. J. Hnilo, M. Fiorino, and G. L. Potter, 2002: NCEP–DOE AMIP-II reanalysis (R-2). Bull. Amer. Meteor. Soc., 83, 1631–1644, https://doi.org/10.1175/BAMS-83-11-1631.
Kendall, M. G., 1975: Rank Correlation Measures. Charles Griffin, 202 pp.
Lee, X., and Coauthors, 2011: Observed increase in local cooling effect of deforestation at higher latitudes. Nature, 479, 384–387, https://doi.org/10.1038/nature10588.
Li, Y., and Coauthors, 2013: Urbanization impact on temperature change in China with emphasis on land cover change and human activity. J. Climate, 26, 8765–8780, https://doi.org/10.1175/JCLI-D-12-00698.1.
Li, Y., and Coauthors, 2016: Potential and actual impacts of deforestation and afforestation on land surface temperature. J. Geophys. Res. Atmos., 121, 14 372–14 386, https://doi.org/10.1002/2016JD024969.
Li, Y., and Coauthors, 2020: Local and teleconnected temperature effects of afforestation and vegetation greening in China. Natl. Sci. Rev., 7, 897–912, https://doi.org/10.1093/nsr/nwz132.
Liang, S., and Coauthors, 2021: The Global Land Surface Satellite (GLASS) product suite. Bull. Amer. Meteor. Soc., 102, 323–337, https://doi.org/10.1175/BAMS-D-18-0341.1.
Liang, X. Z., and Coauthors, 2005: Development of land surface albedo parameterization based on Moderate Resolution Imaging Spectroradiometer (MODIS) data. J. Geophys. Res., 110, D11107, https://doi.org/10.1029/2004JD005579.
Lim, Y.-K., M. Cai, E. Kalnay, and L. Zhou, 2005: Observational evidence of sensitivity of surface climate changes to land types and urbanization. Geophys. Res. Lett., 32, L22712, https://doi.org/10.1029/2005GL024267.
Lim, Y.-K., and Coauthors, 2008: Impact of vegetation types on surface temperature change. J. Appl. Meteor. Climatol., 47, 411–424, https://doi.org/10.1175/2007JAMC1494.1.
Liu, Q., Y. Fu, Z. Zeng, M. Huang, X. Li, and S. Piao, 2016: Temperature, precipitation, and insolation effects on autumn vegetation phenology in temperate China. Global Change Biol., 22, 644–655, https://doi.org/10.1111/gcb.13081.
Liu, T., L. Yu, and S. Zhang, 2019: Impacts of wetland reclamation and paddy field expansion on observed local temperature trends in the Sanjiang Plain of China. J. Geophys. Res. Earth Surface, 124, 414–426, https://doi.org/10.1029/2018JF004846.
Liu, X., Z. Tian, A. Zhang, A. Zhao, and H. Liu, 2019: Impacts of climate on spatiotemporal variations in vegetation NDVI from 1982–2015 in Inner Mongolia, China. Sustainability, 11, 768, https://doi.org/10.3390/su11030768.
Mann, H. B., 1945: Nonparametric tests against trend. Econometrica, 13, 245–259, https://doi.org/10.2307/1907187.
Martens, B., and Coauthors, 2017: GLEAM v3: Satellite-based land evaporation and root-zone soil moisture. Geosci. Model. Dev., 10, 1903–1925, https://doi.org/10.5194/gmd-10-1903-2017.
Myneni, R. B., C. D. Keeling, C. J. Tucker, G. Asrar, and R. R. Nemani, 1997: Increased plant growth in the northern high latitudes from 1981 to 1991. Nature, 386, 698–702, https://doi.org/10.1038/386698a0.
Nayak, S., and M. Mandal, 2019: Impact of land use and land cover changes on temperature trends over India. Land Use Policy, 89, 104238, https://doi.org/10.1016/j.landusepol.2019.104238.
Nemani, R. R., C. D. Keeling, H. Hashimoto, W. M. Jolly, S. C. Piper, C. J. Tucker, R. B. Myneni, and S. W. Running, 2003: Climate-driven increases in global terrestrial net primary production from 1982 to 1999. Science, 300, 1560–1563, https://doi.org/10.1126/science.1082750.
Pearson, R. G., S. J. Phillips, M. M. Loranty, P. S. A. Beck, T. Damoulas, S. J. Knight, and S. J. Goetz, 2013: Shifts in Arctic vegetation and associated feedbacks under climate change. Nat. Climate Change, 3, 673–677, https://doi.org/10.1038/nclimate1858.
Peng, S., and Coauthors, 2014: Afforestation in China cools local land surface temperature. Proc. Natl. Acad. Sci. USA, 111, 2915–2919, https://doi.org/10.1073/pnas.1315126111.
Peñuelas, J., T. Rutishauser, and I. Filella, 2009: Phenology feedbacks on climate change. Science, 324, 887–888, https://doi.org/10.1126/science.1173004.
Pettorelli, N., J. O. Vik, A. Mysterud, J.-M. Gaillard, and C. J. Tucker, 2005: Using the satellite-derived NDVI to assess ecological responses to environmental change. Trends Ecol. Evol., 20, 503–510, https://doi.org/10.1016/j.tree.2005.05.011.
Piao, S., J. Fang, L. Zhou, P. Ciais, and B. Zhu, 2006: Variations in satellite-derived phenology in China’s temperate vegetation. Global Change Biol., 12, 672–685, https://doi.org/10.1111/j.1365-2486.2006.01123.x.
Piao, S., and Coauthors, 2019a: Plant phenology and global climate change: Current progresses and challenges. Global Change Biol., 25, 1922–1940, https://doi.org/10.1111/gcb.14619.
Piao, S., and Coauthors, 2019b: Characteristics, drivers and feedbacks of global greening. Nat. Rev. Earth Environ., 1, 14–27, https://doi.org/10.1038/s43017-019-0001-x.
Pielke, R. A., Sr., and Coauthors, 2002: The influence of land-use change and landscape dynamics on the climate system: Relevance to climate-change policy beyond the radiative effect of greenhouse gases. Philos. Trans. Roy. Soc., 360, 1705–1719, https://doi.org/10.1098/rsta.2002.1027.
Prijith, S. S. , and Coauthors, 2020: Effects of land use/land cover alterations on regional meteorology over northwest India. Sci. Total Environ., 765, 142678, https://doi.org/10.1016/j.scitotenv.2020.142678.
Rathcke, B., and E. P. Lacey, 1985: Phenological patterns of terrestrial plants. Annu. Rev. Ecol. Syst., 16, 179–214, https://doi.org/10.1146/annurev.es.16.110185.001143.
Richardson, A. D., T. F. Keenan, M. Migliavacca, Y. Ryu, O. Sonnentag, and M. Toomey, 2013: Climate change, phenology, and phenological control of vegetation feedbacks to the climate system. Agric. For. Meteor., 169, 156–173, https://doi.org/10.1016/j.agrformet.2012.09.012.
Shen, M., and Coauthors, 2015: Evaporative cooling over the Tibetan Plateau induced by vegetation growth. Proc. Natl. Acad. Sci. USA, 112, 9299–9304, https://doi.org/10.1073/pnas.1504418112.
Shen, X., B. Liu, G. Li, and D. Zhou, 2015: Impacts of grassland types and vegetation cover changes on surface air temperature in the regions of temperate grassland of China. Theor. Appl. Climatol., 126, 141–150, https://doi.org/10.1007/s00704-015-1567-y.
Shen, X., B. Liu, and D. Zhou, 2016a: Using GIMMS NDVI time series to estimate the impacts of grassland vegetation cover on surface air temperatures in the temperate grassland region of China. Remote Sens. Lett., 7, 229–238, https://doi.org/10.1080/2150704X.2015.1128131.
Shen, X., B. Liu, D. Zhou, and X. Lu, 2016b: Effect of grassland vegetation on diurnal temperature range in China’s temperate grassland region. Ecol. Eng., 97, 292–296, https://doi.org/10.1016/j.ecoleng.2016.10.014.
Shen, X., B. Liu, and X. Lu, 2017: Effects of land use/land cover on diurnal temperature range in the temperate grassland region of China. Sci. Total Environ., 575, 1211–1218, https://doi.org/10.1016/j.scitotenv.2016.09.187.
Shen, X., B. Liu, M. Henderson, L. Wang, Z. Wu, H. Wu, M. Jiang, and X. Lu, 2018: Asymmetric effects of daytime and nighttime warming on spring phenology in the temperate grasslands of China. Agric. For. Meteor., 259, 240–249, https://doi.org/10.1016/j.agrformet.2018.05.006.
Shen, X., B. Liu, M. Jiang, and X. Lu, 2020: Marshland loss warms local land surface temperature in China. Geophys. Res. Lett., 47, e2020GL087648, https://doi.org/10.1029/2020GL087648.
Shen, X., and Coauthors, 2021: Aboveground biomass and its spatial distribution pattern of herbaceous marsh vegetation in China. Sci. China Earth Sci., 64, 1115–1125, https://doi.org/10.1007/s11430-020-9778-7.
Simmons, A. J., and Coauthors, 2004: Comparison of trends and low-frequency variability in CRU, ERA-40, and NCEP/NCAR analyses of surface air temperature. J. Geophys. Res., 109, D24115, https://doi.org/10.1029/2004JD005306.
Sun, R., S. Chen, and H. Su, 2021: Climate dynamics of the spatiotemporal changes of vegetation NDVI in northern China from 1982 to 2015. Remote Sens., 13, 187, https://doi.org/10.3390/rs13020187.
Tucker, C. J., J. E. Pinzon, M. E. Brown, and D. A. Slayback, 2005: An extended AVHRR 8-km NDVI dataset compatible with MODIS and SPOT vegetation NDVI data. Int. J. Remote Sens., 26, 4485–4498, https://doi.org/10.1080/01431160500168686.
Wang, J., Z. Yan, and J. Feng, 2018: Exaggerated effect of urbanization in the diurnal temperature range via “observation minus reanalysis” and the physical causes. J. Geophys. Res. Atmos., 123, 7223–7237, https://doi.org/10.1029/2018JD028325.
Wang, Q., D. Riemann, S. Vogt, and R. Glaser, 2014: Impacts of land cover changes on climate trends in Jiangxi province China. Int. J. Biometeor., 58, 645–660, https://doi.org/10.1007/s00484-013-0645-z.
Wang, X., and Coauthors, 2019: No trends in spring and autumn phenology during the global warming hiatus. Nat. Commun., 10, 2389, https://doi.org/10.1038/s41467-019-10235-8.
Winckler, J., Q. Lejeune, C. H. Reick, and J. Pongratz, 2019: Nonlocal effects dominate the global mean surface temperature response to the biogeophysical effects of deforestation. Geophys. Res. Lett., 46, 745–755, https://doi.org/10.1029/2018GL080211.
Wolkovich, E. M., and Coauthors, 2012: Warming experiments underpredict plant phenological responses to climate change. Nature, 485, 494–497, https://doi.org/10.1038/nature11014.
Xu, M., X. Z. Liang, A. Samel, and W. Gao, 2014: MODIS consistent vegetation parameter specifications and their impacts on regional climate simulations. J. Climate, 27, 8578–8596, https://doi.org/10.1175/JCLI-D-14-00082.1.
Xu, X., W. J. Riley, C. D. Koven, G. Jia, and X. Zhang, 2020: Earlier leaf-out warms air in the north. Nat. Climate Change, 10, 370–375, https://doi.org/10.1038/s41558-020-0713-4.
Yang, X., and Coauthors, 2010: Observational evidence of the impact of vegetation cover on surface air temperature change in China. Chin. J. Geophys., 53, 261–269, https://doi.org/10.1002/cjg2.1493.
Yang, X., Y. Hou, and B. Chen, 2011: Observed surface warming induced by urbanization in East China. J. Geophys. Res., 116, D14113, https://doi.org/10.1029/2010JD015452.
Yang, Y., H. Guan, M. Shen, W. Liang, and L. Jiang, 2015: Changes in autumn vegetation dormancy onset date and the climate controls across temperate ecosystems in China from 1982 to 2010. Global Change Biol., 21, 652–665, https://doi.org/10.1111/gcb.12778.
Ying, H., H. Zhang, J. Zhao, Y. Shan, Z. Zhang, X. Guo, R. Wu, and G. Deng, 2020: Effects of spring and summer extreme climate events on the autumn phenology of different vegetation types of Inner Mongolia, China, from 1982 to 2015. Ecol. Indic., 111, 105974, https://doi.org/10.1016/j.ecolind.2019.105974.
Zeng, Z., and Coauthors, 2017: Climate mitigation from vegetation biophysical feedbacks during the past three decades. Nat. Climate Change, 7, 432–436, https://doi.org/10.1038/nclimate3299.
Zhang, C., Y. Zhang, Z. Wang, J. Li, and I. Odeh, 2020: Monitoring phenology in the temperate grasslands of China from 1982 to 2015 and its relation to net primary productivity. Sustainability, 12, 12, https://doi.org/10.3390/su12010012.
Zhang, M., Y. Liu, J. Zhang, and Q. Wen, 2021: AMOC and climate responses to dust reduction and greening of the Sahara during the mid-Holocene. J. Climate, 34, 4893–4912, https://doi.org/10.1175/JCLI-D-20-0628.1.
Zhou, D., D. Li, G. Sun, L. Zhang, Y. Liu, and L. Hao, 2016: Contrasting effects of urbanization and agriculture on surface temperature in eastern China. J. Geophys. Res. Atmos., 121, 9597–9606, https://doi.org/10.1002/2016JD025359.
Zhou, L., and Coauthors, 2004: Evidence for a significant urbanization effect on climate in China. Proc. Natl. Acad. Sci. USA, 101, 9540–9544, https://doi.org/10.1073/pnas.0400357101.
Zhou, X., X. Geng, G. Yin, H. Hänninen, F. Hao, X. Zhang, and Y. H. Fu, 2020: Legacy effect of spring phenology on vegetation growth in temperate China. Agric. For. Meteor., 281, 107845, https://doi.org/10.1016/j.agrformet.2019.107845.
