Abstract

Changes in snow cover over the Qinghai–Tibetan Plateau have attracted much attention in recent years owing to climate change. Because of the limitations of in situ observations, only a few studies have analyzed the dynamics of snow cover. Using observations from 103 meteorological stations across the Qinghai–Tibetan Plateau, this study investigated the spatial and temporal variability of snow depth and the number of snow-cover days. The results show a very weak negative trend for the snow depth and the number of snow-cover days in spring and winter from 1961 to 2010, but two different trends were found: an initial increase followed by a decrease. In summer and autumn, snow depth and the number of snow-cover days show a significant decreasing trend for most sites. The duration of snow cover exhibits a significant decreasing trend (−3.5 ± 1.2 days decade−1), which was jointly controlled by a later snow starting time (1.6 ± 0.8 days decade−1) and an earlier snow ending time (−1.9 ± 0.8 days decade−1) consistent with a response to climate change. This study highlights the competing effects of rising temperatures and changing precipitation, which remain an important challenge in understanding and interpreting the observed changes in snow depth and the number of snow-cover days for the Qinghai–Tibetan Plateau.

1. Introduction

Snow cover plays an important role in regulating regional and global climate, especially in the Qinghai–Tibetan Plateau, because of its high surface albedo and heat-insulation effect, which influences the energy exchange between the land surface and atmosphere (Barnett et al. 1988; Yang et al. 2001; Chapin et al. 2005; Euskirchen et al. 2007). More than a century ago, Blanford (1884) suggested that an inverse relationship existed between summer rainfall over northwestern India and the mean snow cover in the western Himalayas. In recent years, numerous studies have reported a negative correlation between snow cover over the Tibetan Plateau and summer rainfall in India (Yasunari et al. 1991; Zhao and Moore 2004) and positive correlations between snow cover over the Tibetan Plateau and summer rainfall in southern China (Bamzai and Shukla 1999; Robock et al. 2003; Zhao and Moore 2004). Moreover, snow cover also strongly influences hydrological processes, vegetation phenology, and the carbon cycle (Goulden et al. 1998; Barnett et al. 2005; Monson et al. 2006; Rawlins et al. 2006; Dorrepaal et al. 2009; Peng et al. 2013).

The Qinghai–Tibetan Plateau is the highest plateau on Earth, covering more than 2.0 × 106 km2 at an average elevation exceeding 4500 m above mean sea level (MSL). Considerable heterogeneity in the topography and climate has created complex spatial and temporal snow cover patterns. Because of limited in situ observations, most studies have reported snow cover distribution and its variability using satellite-based observations (Zhang et al. 2004; Che et al. 2008; Gao et al. 2012; Marchane et al. 2015). However, the satellite data related to snow cover are lacking with regard to snow depth and do not provide a sufficient time series length. Using surface observations of snow depth, a few studies have shown an increase over the Qinghai–Tibetan Plateau (Qin et al. 2006; Ma and Qin 2012). A previous study reported that the annual cumulative daily snow depth increased by 23% decade−1 from 1957 to 1998 over the plateau, even though winter temperatures have been warmer (Qin et al. 2006). Intriguingly, after 2000, the mean air temperature in 2001–10 increased by 1.1°C compared with the 1990s (Li et al. 2012), and this warming rate was higher than that of other global regions. Snow cover responses need to be urgently evaluated using longer-term observations following Qin et al. (2006). In particular, the changes in snow cover during the current decade under intense climate warming need to be investigated.

In this study, we investigated the spatial and temporal variations and trends in snow depth and the number of snow-cover days from 1961 to 2010 based on observations from 103 stations across the Qinghai–Tibetan Plateau. More specifically, we aimed 1) to compare the differences in the spatial and temporal patterns in snow depth and the number of snow-cover days for the Qinghai–Tibetan Plateau, 2) to assess the interannual variability in snow depth and the number of snow-cover days, 3) to analyze the relationship between snow depth and climatic factors (i.e., precipitation and temperature) over the various seasons, and 4) to investigate the changes in snow starting time, snow ending time, and snow-cover duration in response to climate change.

2. Data and methods

a. Snow cover and climate data

Daily snow depth data and meteorological data for the Qinghai–Tibetan Plateau were collected from 103 national meteorological stations of the China Meteorological Administration (Fig. 1). Missing records of daily snow depth data were filled using meteorological data (i.e., daily precipitation and daily air temperature) according to Yuan et al. (2015, 2016). When the missing records were less than 15 consecutive days, if the daily average air temperature was below 0°C, then daily precipitation was added to the daily snow depth, and if the daily average air temperature was between 0° and 2°C, then half of the precipitation was added to the snow depth. Following strict quality control, the station records were integrated into snow-cover year (from August to the next July). Then, the snow-cover year data were divided into the four seasons: spring (March–May), summer (June–August), autumn (September–November), and winter (December–February) from 1961 to 2010. In this study, the mean daily snow depth is the average depth in all days during a certain period (i.e., spring, summer, autumn, winter, and snow-cover year).

Fig. 1.

Distribution of meteorological stations over the Qinghai–Tibetan Plateau in China. The hatching indicates the region of the Qinghai–Tibetan Plateau, and the dots show the meteorological stations.

Fig. 1.

Distribution of meteorological stations over the Qinghai–Tibetan Plateau in China. The hatching indicates the region of the Qinghai–Tibetan Plateau, and the dots show the meteorological stations.

After gap filling the snow depth data and the number of snow-cover days, we identified the snow starting time as the first day with snow depth ≥1 cm at each station in the latter half of the year (August–December), while the snow ending time was defined as the last day with snow depth ≥1 cm in the first half of the year (January–July). The duration of snow cover was defined as the number of days from the snow starting to snow ending time. To accurately detect the snow cover starting and ending times and to avoid errors resulting from missing data, the years with missing records between 1 August and the snow starting time or the snow ending time and 31 July were excluded from the analysis of snow cover phenology. Furthermore, if both snow starting time and snow ending time occurred during a half-year period (January–July or August–December), the corresponding year at this site was not considered. A total of 20.7% of the years were excluded from the analyses.

b. Statistical analysis

Linear analysis was used to analyze the trends in the mean daily snow depth, the number of snow-cover days, and snow phenological variables (i.e., snow starting time, snow ending time, and snow-cover duration) for each station from 1961 to 2010. We used a linear model yt = β1xt + β0 (Zhang et al. 2009; Chen et al. 2014) to model yt against time t. We used the interval estimation of parameters to calculate the standard deviation of β1. For the linear model, and were the estimates of β0 and β1, respectively. The interval estimation of (i = 0, 1) was

 
formula

so that

 
formula

or

 
formula

where α, n, E1, and E2 are the confidence level, sample size, and lower and upper limits of the interval estimation, respectively. The statistic β1/SE(β1) [where SE(β1) is the standard deviation of β1] has the Student’s t distribution, and the Student’s t test was used to analyze and classify the trend significance into weak, moderate, and strong. When |β1/SE(β1)| < 1.0, β1 was within one standard deviation, and the trend was classified as weak; when 1.0 ≤ |β1/SE(β1)| < t0.10, where t0.10 is the 10% critical value of the Student’s t distribution, the trend was classified as moderate; when |β1/SE(β1)| ≥ t0.10, the trend was statistically significant and classified as strong. These categories were further classified into six classes according to the slopes of the statistical trends: negative strong, negative moderate, negative weak, positive weak, positive moderate, and positive strong.

A piecewise regression approach was used to detect the timing of the mean daily snow depth and the number of snow-cover days across 102 stations during the period of 1961 to 2010 [the Nielamu station (28.2°N, 86.0°E; 3810.0 m) was not included because of its location in a ravine]:

 
formula

where y is a snow variable (i.e., mean daily snow depth or the number of snow-cover days), t is the year, α is the estimated turning point (TP) of the time series, defining the timing of a trend change, β0, β1, and β2 are the regression coefficients, and ε is the residual of the fit. The snow variable linear trend is β1 before the TP, and β1 + β2 after it. Least squares linear regression was used to estimate α and the other coefficients. A p value ≤ 0.05 was considered significant.

3. Results

a. Spatial–temporal variations of snow depth and the number of snow-cover days

Large spatial heterogeneities were observed for both the snow depth and the number of snow-cover days. On average, the mean daily snow depth was less than 0.5 cm at 76.5% of all sites in the spring, summer, autumn, winter, and the snow-cover year averaged over the period 1961–2010 (Fig. 2). The mean daily snow depth from 1961 to 2010 was generally <1.5 cm in winter, with the exception of a site in a ravine in the southwestern region of the Qinghai–Tibetan Plateau (Fig. 2). The annual number of snow-cover days was 11.1 ± 1.0 days in winter, ranging from a minimum value of 0.1 to a maximum value of 41.8 days (Fig. 2n). Because of the high elevation, in the eastern Tanggula Mountains and the Himalayas, the snow cover was deep and persisted for many days (Fig. 2r).

Fig. 2.

Spatial patterns of mean daily snow depth and the number of snow-cover days from 1961 to 2010 and trends in mean daily snow depth and the number of snow-cover days across 103 stations during the period of 1961–2010. The inset graphs at bottom left show the frequency distributions of corresponding long-term means and trends.

Fig. 2.

Spatial patterns of mean daily snow depth and the number of snow-cover days from 1961 to 2010 and trends in mean daily snow depth and the number of snow-cover days across 103 stations during the period of 1961–2010. The inset graphs at bottom left show the frequency distributions of corresponding long-term means and trends.

Most sites had weak changing trends regarding the snow depth and the number of snow-cover days from 1961 to 2010 during the various seasons (Fig. 2). On average, the mean daily snow depth in the spring, summer, autumn, winter, and the snow-cover year showed weak trends at 53.4%, 20.4%, 54.4%, 57.3%, and 46.6% of all sites, respectively. Similarly, the number of snow-cover days during all seasons had a weak negative trend at 21.4%–40.8% of sites and a weak positive trend at 7.7%–35.9% of sites (Fig. 2). There were large spatial discrepancies in the trends of the mean daily snow depth and the number of snow-cover days among the four seasons. For example, in spring, 54.4% of sites with a positive mean daily snow depth trend were primarily located in the eastern Qinghai–Tibetan Plateau. In winter, however, 47.6% of sites with negative mean daily snow depth trends were located in the southeast of the plateau.

Our results indicated that the trends for both snow depth and the number of snow-cover days were not continuous from 1961 to 2010. The mean daily snow-cover depth of the snow-cover year increased nonsignificantly (R2 = 0.12; p = 0.13) during 1961–81 at a rate of 0.004 ± 0.003 cm yr−1, but a significant decreasing trend (−0.004 ± 0.002 cm yr−1; R2 = 0.16; p < 0.05) was observed during 1981–2010. Further analyses of mean daily snow depth and the number of snow-cover days indicated similar patterns, with different trends before and after the turning point in different seasons (Fig. 3). For example, most sites (56.9%) had positive trends of mean daily snow depth in spring before 1981, particularly in the eastern region of the Qinghai–Tibetan Plateau, after which negative trends were observed at 57.8% of sites (Figs. 3b,c).

Fig. 3.

Interannual variability and trends before and after the TP (i.e., turning year) of mean daily snow depth and the number of snow-cover days from 1961 to 2010. (left) Interannual variability: the arrow indicates the turning point, the dotted gray line denotes the trend over the period 1961–2010, the dotted red line denotes the trend before the turning year, and the solid blue line denotes the trend after the turning year. The trend (center) before and (right) after the TP; the inset graphs at bottom left show the frequency distributions of corresponding trends.

Fig. 3.

Interannual variability and trends before and after the TP (i.e., turning year) of mean daily snow depth and the number of snow-cover days from 1961 to 2010. (left) Interannual variability: the arrow indicates the turning point, the dotted gray line denotes the trend over the period 1961–2010, the dotted red line denotes the trend before the turning year, and the solid blue line denotes the trend after the turning year. The trend (center) before and (right) after the TP; the inset graphs at bottom left show the frequency distributions of corresponding trends.

Temperature and precipitation determined the changes in snow cover; however, these environmental variables played different roles in the various seasons. In spring, summer, and autumn, air temperature was the dominant variable regulating the mean daily snow depth and the number of snow-cover days (Figs. 4a–c; Table 1). For example, from 1961 to 2010, mean daily snow depth in spring increased before 1981 and then decreased (Fig. 3a), which coincided with changes in the air temperature (Fig. 4a; Table 1). On the contrary, in winter, precipitation determined the changes in mean daily snow depth and the number of snow-cover days (Fig. 4d; Table 1). From 1998 to 2003, a continuous 6-yr decrease in precipitation resulted in low mean daily snow depth (Fig. 4d).

Fig. 4.

Interannual variability of precipitation (black line), air temperature (red line), mean daily snow depth (blue line), and the number of snow-cover days (green line) across 102 stations over the Qinghai–Tibetan Plateau in (a) spring, (b) summer, (c) autumn, and (d) winter during the period of 1961–2010.

Fig. 4.

Interannual variability of precipitation (black line), air temperature (red line), mean daily snow depth (blue line), and the number of snow-cover days (green line) across 102 stations over the Qinghai–Tibetan Plateau in (a) spring, (b) summer, (c) autumn, and (d) winter during the period of 1961–2010.

Table 1.

Pearson correlation coefficients between mean daily snow depth, the number of snow-cover days, precipitation, and air temperature of different seasons in the Qinghai–Tibetan Plateau from 1961 to 2010; one and two asterisks denote significance at the 0.05 and 0.01 levels, respectively.

Pearson correlation coefficients between mean daily snow depth, the number of snow-cover days, precipitation, and air temperature of different seasons in the Qinghai–Tibetan Plateau from 1961 to 2010; one and two asterisks denote significance at the 0.05 and 0.01 levels, respectively.
Pearson correlation coefficients between mean daily snow depth, the number of snow-cover days, precipitation, and air temperature of different seasons in the Qinghai–Tibetan Plateau from 1961 to 2010; one and two asterisks denote significance at the 0.05 and 0.01 levels, respectively.

b. Responses of snow phenology to climate change

Figure 5 shows the spatial patterns and trends of the snow starting time, snow ending time, and snow-cover duration from 1961 to 2010. There were remarkable spatial variability in all three parameters (Figs. 5a,c,e). In terms of the snow starting time, a positive trend was found at 67.9% of sites (Fig. 5b), which indicated later snow cover starting dates at most sites. In contrast, a trend toward an earlier snow ending time was found at 61.2% of the sites (Fig. 5d), and 68.0% of sites had a negative trend in snow-cover duration, which means a shorter snow-cover duration (Fig. 5f). There were clear elevation gradients for snow cover starting date (earlier with higher elevation), snow cover ending date (later with higher elevation) and snow-cover duration (longer with higher elevation) (Figs. 6a,c,e; Table 2). Significant changes in snow phenology were found at all elevations, but particularly large changes occurred at high elevations (Table 2). The results showed a significant decreasing trend in snow-cover duration, which was jointly determined by later starting dates and earlier ending dates. For example, in areas with an elevation over 4000 m, the results indicated a later snow starting time (average: 1.6 ± 0.8 days decade−1), an earlier snow ending time (average: −2.0 ± 0.8 days decade−1), and a shorter snow-cover duration (average: −3.5 ± 1.2 days decade−1) over the period 1961–2010 for the Qinghai–Tibetan Plateau (Table 2).

Fig. 5.

Spatial patterns of mean annual (a) snow starting time, (c) snow ending time, and (e) snow-cover duration averaged over the period 1961–2010 and trends in (b) snow starting time, (d) snow ending time, and (f) snow-cover duration across 103 stations during the period of 1961–2010. The inset graphs at bottom left show the frequency distributions of corresponding long-term means in (a),(c),(e) and trends in (b),(d),(f).

Fig. 5.

Spatial patterns of mean annual (a) snow starting time, (c) snow ending time, and (e) snow-cover duration averaged over the period 1961–2010 and trends in (b) snow starting time, (d) snow ending time, and (f) snow-cover duration across 103 stations during the period of 1961–2010. The inset graphs at bottom left show the frequency distributions of corresponding long-term means in (a),(c),(e) and trends in (b),(d),(f).

Fig. 6.

The differences in annual (a) snow starting time, (c) snow ending time, and (e) snow-cover duration with elevation and the interannual variability in (b) snow starting time, (d) snow ending time, and (f) snow-cover duration during the period of 1961–2010 at different elevations (E1 < 3000 m, E2 = 3000–4000 m, E3 ≥ 4000 m, and E is all sites).

Fig. 6.

The differences in annual (a) snow starting time, (c) snow ending time, and (e) snow-cover duration with elevation and the interannual variability in (b) snow starting time, (d) snow ending time, and (f) snow-cover duration during the period of 1961–2010 at different elevations (E1 < 3000 m, E2 = 3000–4000 m, E3 ≥ 4000 m, and E is all sites).

Table 2.

Long-term change trends of annual snow starting time, snow ending time, and snow-cover duration from 1961 to 2010 at different elevations (E1 < 3000 m, E2 = 3000–4000 m, E3 ≥ 4000 m, and E is all sites).

Long-term change trends of annual snow starting time, snow ending time, and snow-cover duration from 1961 to 2010 at different elevations (E1 < 3000 m, E2 = 3000–4000 m, E3 ≥ 4000 m, and E is all sites).
Long-term change trends of annual snow starting time, snow ending time, and snow-cover duration from 1961 to 2010 at different elevations (E1 < 3000 m, E2 = 3000–4000 m, E3 ≥ 4000 m, and E is all sites).

Compared to the number of snow-cover days and the mean daily snow depth, the snow-cover duration exhibited a significant decreasing trend in recent decades. On average, the duration of snow cover was 188.9 ± 7.5 days from 2001 to 2010, which decreased 16.7 days compared with that from 1961 to 1970 (205.6 ± 9.0 days). Compared with the period of 1961–70, the snow starting time was delayed from 287.9 ± 7.5 to 297.8 ± 6.0 Julian day, while the snow ending time advanced from 127.5 ± 5.6 to 120.7 ± 3.9 Julian day in recent decades (Figs. 5 and 6).

Trends in temperature and precipitation are important in determining the duration of snow cover. The trends in mean annual air temperature were positively correlated with the trends in the snow starting time but negatively correlated with the trends in the snow ending time (Figs. 7a,b). There were no significant relationships found between the trends in precipitation and the trends in snow starting time or between the trends in precipitation and the trends in snow ending time during the year of the snow starting (ending) time across all sites (Figs. 7c,d).

Fig. 7.

Correlations between the trend in snow cover phenology (starting date DS and ending date DE) and environmental variables (annual mean air temperature T and precipitation P). The gray dashed line indicates the regression line.

Fig. 7.

Correlations between the trend in snow cover phenology (starting date DS and ending date DE) and environmental variables (annual mean air temperature T and precipitation P). The gray dashed line indicates the regression line.

4. Discussion

The Qinghai–Tibetan Plateau is the highest plateau globally, covering approximately 2.5 × 106 km2 at an average elevation of 4500 m. In the Qinghai–Tibetan Plateau, the weakening moisture-bearing monsoon and varying topography shape a semiarid climate with mean annual precipitation below 450 mm, as well as relatively little snowfall (Wang et al. 2009). The mean annual temperature ranges from −15° to 10°C and decreases from the edges of the plateau toward the center with the increasing altitudes (You et al. 2013). During the last 45 years, the Qinghai–Tibetan Plateau has experienced significant climate warming of approximately 0.265°C decade−1 (Lu and Liu 2010), and the rate of warming has accelerated in recent decades (Li et al. 2012), which has primarily resulted from increasing anthropogenic greenhouse gas emissions (Duan et al. 2006). Some studies reported a decreased precipitation trend as a result of the weakening Indian monsoon (Yao et al. 2012) and increasing temperature (Qin et al. 2009) at the high altitudes of the Qinghai–Tibetan Plateau. For example, Lin and Zhao (1996) found that the precipitation over the Qinghai–Tibetan Plateau decreased by 10–40 mm decade−1 from the 1950s to the 1990s.

Numerous studies have focused on the global changes in snow cover in response to climate warming during recent decades (Qin et al. 2006; Choi et al. 2010; Kapnick and Hall 2010; Brown and Robinson 2011; Gao et al. 2012; Guo and Li 2015; Ke et al. 2015). For example, Brown and Robinson (2011) highlighted that snow cover extent over the Northern Hemisphere significantly decreased over the past 90 years, with a very high level of confidence. Over the Northern Hemisphere, in situ data generally indicate a decrease in the number of snow-cover days and snow depth in response to climate warming, especially at lower elevations or higher average temperature locations (IPCC 2013). However, snow cover over the Qinghai–Tibetan Plateau displays unique characteristics. By blending in situ and satellite records, Qin et al. (2006) suggested increasing snow cover between 1951 and 1997, and western China did not experience a continual decrease in snow cover during the great warming period of the 1980s and 1990s. Our study, using a longer time series and extending the study period to 2010, found very complicated changes in the number of snow-cover days and snow depth for the Qinghai–Tibetan Plateau, which increased from 1961 to 1981 and then decreased after 1981 in the spring and snow-cover year (Fig. 3).

From a seasonal perspective, previous studies have indicated the largest decreases in snow cover in spring at high latitudes, resulting primarily from the rising spring temperature (Brown and Robinson 2011). Our results, however, indicated the strongest depletions in snow cover occurred in autumn over the Qinghai–Tibetan Plateau (Table 3), which is different from high latitudes. One possible cause is that air temperature increased at the largest rate in autumn (Table 3). Moreover, the variability in snow cover is controlled collectively by various climate factors rather than a single variable (i.e., precipitation or temperature). Our result showed that in spring, a decreasing trend in snow depth with an increasing trend in air temperature were found even though an increasing precipitation trend was observed, especially after 1981 (Figs. 3a and 4a). The major cause is the mean air temperature during spring is above the freezing point, and therefore, the increase in air temperature will lead to a decrease in snow depth, which neutralizes the impacts of increased precipitation (Ye et al. 1998; Peng et al. 2010; Gao et al. 2012). By contrast, although the air temperature still increased in winter, the snow depth decreased mainly as a result of the decrease in precipitation after 1997 (Fig. 4d). As noted in Peng et al. (2010), even if the temperature is increasing, the intensified precipitation will result in increased snow depth when the temperature is below the freezing point. Therefore, the competing effects of rising temperatures and changing precipitation resulted in the difference trend before and after the TP (Figs. 3 and 4). Note that an increasing trend in temperature and precipitation will continue in future decades based on the results from various general circulation models (GCMs) for the Qinghai–Tibetan Plateau (Su et al. 2013). Snow cover prediction is more complex because of the changes in precipitation regimes coupled with temperature variations under different topographic conditions.

Table 3.

Summary of changing trends in snow cover (i.e., mean daily snow depth and the number of snow-cover days) and environmental variables (i.e., precipitation and air temperature) from 1961 to 2010 over the Qinghai–Tibetan Plateau. Snow cover for year from August to the next July, for spring from March to May, for summer from June to August, for autumn from September to November, and for winter from December to February during the period of 1961–2010.

Summary of changing trends in snow cover (i.e., mean daily snow depth and the number of snow-cover days) and environmental variables (i.e., precipitation and air temperature) from 1961 to 2010 over the Qinghai–Tibetan Plateau. Snow cover for year from August to the next July, for spring from March to May, for summer from June to August, for autumn from September to November, and for winter from December to February during the period of 1961–2010.
Summary of changing trends in snow cover (i.e., mean daily snow depth and the number of snow-cover days) and environmental variables (i.e., precipitation and air temperature) from 1961 to 2010 over the Qinghai–Tibetan Plateau. Snow cover for year from August to the next July, for spring from March to May, for summer from June to August, for autumn from September to November, and for winter from December to February during the period of 1961–2010.

Our result showed the snow-cover duration significantly shortened in recent decades. Based on passive microwave satellite data collected since 1979, other studies support a significant trend toward a shortening of the snowmelt season over much of Eurasia (Takala et al. 2009) and the pan-Arctic region (Tedesco and Monaghan 2009). Snowmelt mainly occurs in spring when the temperature is closer to the freezing point; therefore, changes in air temperature are the most effective at increasing snowmelt. Our results also highlighted that snow cover phenology is more reliable, with variations in association with climate change rather than other snow-cover metrics. Previous study has indicated that snow depth over Eurasia significantly increases in winter, but the snow-cover period has significantly shortened (Bulygina et al. 2009).

This study identified larger differences in snow cover at higher elevations (Figs. 6a,c,e). This is the first study to show that the depletion rate of snow cover increases with respect to elevation below roughly 4000 m MSL, and this depletion rate disappears at higher elevations (>4000 m). Other evidence supports our conclusion that the warming rate increases from 3000 to 4800 m MSL in the Qinghai–Tibetan Plateau and then stabilizes, with a slight decline at the highest elevations (Qin et al. 2009). This altitudinal dependence of snow-cover changes has significant implications for the Qinghai–Tibetan Plateau because most snow surfaces significantly increase with elevation over the Qinghai–Tibetan Plateau (data not shown).

It is important to study the changes in snow depth and duration over the Qinghai–Tibetan Plateau. Snow cover is sensitive to the local temperature and precipitation, latitude and elevation gradient, and terrain roughness. Snow-cover variation is mainly attributable to large-scale atmospheric circulation or climatic forcing (Kripalani et al. 2001; Wu and Wang 2002; Xu et al. 2010; Birsan and Dumitrescu 2014), such as monsoons, the North Atlantic Oscillation, the Arctic Oscillation, and El Niño–Southern Oscillation. Our results identify a steep decline in the winter snow depth in 1998, after a rich snow cover occurred in 1997 (Fig. 4d). Chen (2001) suggested that the great floods of 1998 in the Yangtze River valley of China were related to the anomalous snow cover over the Qinghai–Tibetan Plateau, which resulted from the seasonal northward advance of the west Pacific subtropical high being delayed, with a subsequent enhanced summer rainfall in the Yangtze River valley. Moreover, the extent of snow cover and snow melting time can impact the carbon and nitrogen balances of terrestrial ecosystems (Monson et al. 2006; Jonas et al. 2008). Peng et al. (2010) indicated warmer soil conditions with a thicker snowpack may increase winter soil organic carbon decomposition and nitrogen mineralization owing to the enhancement of soil microbial activity.

It should be noticed that there are few stations over the depopulated western plateau (Fig. 1). The western region of the Qinghai–Tibetan Plateau is relatively dry and cold. Therefore, the current analyses based on the station observations are not sufficient to indicate the characteristics of the western plateau. Satellite-based snow cover products provide important datasets for investigating the characteristics of snow cover over regions without observations. Future studies should comprehensively integrate station and satellite-based observations to investigate the snow-cover changes over the Qinghai–Tibetan Plateau.

5. Conclusions

In this study, we investigated the spatial and temporal characteristics of mean daily snow depth and the number of snow-cover days over the Qinghai–Tibetan Plateau from 1961 to 2010. Weak negative trends for the mean daily snow depth and the number of snow-cover days were observed throughout the study in the snow-cover period, but two distinctly different trends were also identified for mean daily snow depth and snow-cover days. The long-term variation in the number of snow-cover days was in agreement with the snowfall record, but it was inconsistent with the warming trend. Our analyses of in situ observations indicated that snow cover over the Qinghai–Tibetan Plateau is controlled collectively by various climate factors rather than a single variable (i.e., precipitation or temperature). Furthermore, clear elevation gradients were found in the snow starting time (earlier with higher elevation), snow cover ending time (later with higher elevation), and snow-cover duration (longer with higher elevation). A later snow starting time (1.6 ± 0.8 days decade−1), an earlier snow ending time (−2.0 ± 0.8 days decade−1), and a shorter snow-cover duration (−3.5 ± 1.2 days decade−1) were observed in areas with an elevation over 4000 m.

Acknowledgments

This study was supported by Key Project of Chinese Academy of Sciences (CAS) (KJZDEW-G03-04); National Science Foundation for Excellent Young Scholars of China (41322005), One Hundred Person Project of CAS; and National Key Scientific Research Projects (2013CBA01802). Some of this work has been presented as a poster at the 2015 Fall Meeting by some of the coauthors.

REFERENCES

REFERENCES
Bamzai
,
A. S.
, and
J.
Shukla
,
1999
:
Relation between Eurasian snow cover, snow depth, and the Indian summer monsoon: An observational study
.
J. Climate
,
12
,
3117
3132
, doi:.
Barnett
,
T. P.
,
L.
Dümenil
,
U.
Schlese
, and
E.
Roeckner
,
1988
:
The effect of Eurasian snow cover on global climate
.
Science
,
239
,
504
507
, doi:.
Barnett
,
T. P.
,
J. C.
Adam
, and
D. P.
Lettenmaier
,
2005
:
Potential impacts of a warming climate on water availability in snow-dominated regions
.
Nature
,
438
,
303
309
, doi:.
Birsan
,
M. V.
, and
A.
Dumitrescu
,
2014
:
Snow variability in Romania in connection to large-scale atmospheric circulation
.
Int. J. Climatol.
,
34
,
134
144
, doi:.
Blanford
,
H. F.
,
1884
:
On the connexion of the Himalaya snowfall with dry winds and seasons of drought in India
.
Proc. Roy. Soc. London
,
37
,
3
22
, doi:.
Brown
,
R. D.
, and
D. A.
Robinson
,
2011
:
Northern Hemisphere spring snow cover variability and change over 1922–2010 including an assessment of uncertainty
.
Cryosphere
,
5
,
219
229
, doi:.
Bulygina
,
O. N.
,
V. N.
Razuvaev
, and
N. N.
Korshunova
,
2009
:
Changes in snow cover over Northern Eurasia in the last few decades
.
Environ. Res. Lett.
,
4
, 045026, doi:.
Chapin
,
F. S.
, and Coauthors
,
2005
:
Role of land-surface changes in Arctic summer warming
.
Science
,
310
,
657
660
, doi:.
Che
,
T.
,
X.
Li
,
R.
Jin
,
R.
Armstrong
, and
T. J.
Zhang
,
2008
:
Snow depth derived from passive microwave remote-sensing data in China
.
Ann. Glaciol.
,
49
,
145
154
, doi:.
Chen
,
L.
,
2001
:
The role of the anomalous snow cover over the Qinghai-Xizang Plateau and ENSO in the great floods of 1998 in the Changjiang River valley
(in Chinese).
Chin. J. Atmos. Sci.
,
25
,
184
192
.
Chen
,
Y.
, and Coauthors
,
2014
:
Comparison of satellite-based evapotranspiration models over terrestrial ecosystems in China
.
Remote Sens. Environ.
,
140
,
279
293
, doi:.
Choi
,
G.
,
D. A.
Robinson
, and
S.
Kang
,
2010
:
Changing Northern Hemisphere snow seasons
.
J. Climate
,
23
,
5305
5310
, doi:.
Dorrepaal
,
E.
,
S.
Toet
,
R. S. P.
van Logtestijn
,
E.
Swart
,
M. J.
van de Weg
,
T. V.
Callaghan
, and
R.
Aerts
,
2009
:
Carbon respiration from subsurface peat accelerated by climate warming in the subarctic
.
Nature
,
460
,
616
619
, doi:.
Duan
,
A.
,
G.
Wu
,
Q.
Zhang
, and
Y.
Liu
,
2006
:
New proofs of the recent climate warming over the Tibetan Plateau as a result of the increasing greenhouse gases emissions
.
Chin. Sci. Bull.
,
51
,
1396
1400
, doi:.
Euskirchen
,
E. S.
,
A. D.
McGuire
, and
F. S.
Chapin
,
2007
:
Energy feedbacks of northern high-latitude ecosystems to the climate system due to reduced snow cover during 20th century warming
.
Global Change Biol.
,
13
,
2425
2438
, doi:.
Gao
,
J.
,
M. W.
Williams
,
X.
Fu
,
G.
Wang
, and
T.
Gong
,
2012
:
Spatiotemporal distribution of snow in eastern Tibet and the response to climate change
.
Remote Sens. Environ.
,
121
,
1
9
, doi:.
Goulden
,
M. L.
, and Coauthors
,
1998
:
Sensitivity of boreal forest carbon balance to soil thaw
.
Science
,
279
,
214
217
, doi:.
Guo
,
L.
, and
L.
Li
,
2015
:
Variation of the proportion of precipitation occurring as snow in the Tian Shan Mountains, China
.
Int. J. Climatol.
,
35
,
1379
1393
, doi:.
IPCC
,
2013
: Climate Change 2013: The Physical Science Basis. Cambridge University Press, 1535 pp., doi:.
Jonas
,
T.
,
C.
Rixen
,
M.
Sturm
, and
M.
Stoeckli
,
2008
:
How alpine plant growth is linked to snow cover and climate variability
.
J. Geophys. Res.
,
113
, G03013, doi:.
Kapnick
,
S.
, and
A.
Hall
,
2010
:
Observed climate–snowpack relationships in California and their implications for the future
.
J. Climate
,
23
,
3446
3456
, doi:.
Ke
,
C.
,
X.
Li
,
H.
Xie
, and
C.
Kou
,
2015
:
Variability in snow cover phenology in China from 1952 to 2010
.
Hydrol. Earth Syst. Sci.
,
20
,
755
770
, doi:.
Kripalani
,
R. H.
,
A.
Kulkarni
, and
S. S.
Sabade
,
2001
:
El Niño Southern Oscillation, Eurasian snow cover and the Indian monsoon rainfall
. Proc. Indian Natl. Sci. Acad., 67, 361–368.
Li
,
R.
,
L.
Zhao
,
Y.
Ding
,
T.
Wu
,
Y.
Xiao
,
E.
Du
,
G.
Liu
, and
Y.
Qiao
,
2012
:
Temporal and spatial variations of the active layer along the Qinghai-Tibet Highway in a permafrost region
.
Chin. Sci. Bull.
,
57
,
4609
4616
, doi:.
Lin
,
Z.
, and
X.
Zhao
,
1996
:
Spatial characteristics of changes in temperature and precipitation of the Qinghai-Xizang (Tibet) Plateau
.
Sci. China
,
4
,
442
448
.
Lu
,
H.
, and
G.
Liu
,
2010
:
Trends in temperature and precipitation on the Tibetan Plateau, 1961–2005
.
Climate Res.
,
43
,
179
190
, doi:.
Ma
,
L.
, and
D.
Qin
,
2012
:
Temporal–spatial characteristics of observed key parameters of snow cover in China during 1957–2009
. Sci. Cold Arid Reg., 4, 384–393, doi:.
Marchane
,
A.
, and Coauthors
,
2015
:
Assessment of daily MODIS snow cover products to monitor snow cover dynamics over the Moroccan Atlas mountain range
.
Remote Sens. Environ.
,
160
,
72
86
, doi:.
Monson
,
R. K.
,
D. L.
Lipson
,
S. P.
Burns
,
A. A.
Turnipseed
,
A. C.
Delany
,
M. W.
Williams
, and
S. K.
Schmidt
,
2006
:
Winter forest soil respiration controlled by climate and microbial community composition
.
Nature
,
439
,
711
714
, doi:.
Peng
,
S.
,
S.
Piao
,
P.
Ciais
,
J.
Fang
, and
X.
Wang
,
2010
:
Change in winter snow depth and its impacts on vegetation in China
.
Global Change Biol.
,
16
,
3004
3013
.
Peng
,
S.
,
S.
Piao
,
P.
Ciais
,
P.
Friedlingstein
,
L.
Zhou
, and
T.
Wang
,
2013
:
Change in snow phenology and its potential feedback to temperature in the Northern Hemisphere over the last three decades
.
Environ. Res. Lett.
,
8
, 014008, doi:.
Qin
,
D.
,
S.
Liu
, and
P.
Li
,
2006
:
Snow cover distribution, variability, and response to climate change in western China
.
J. Climate
,
19
,
1820
1833
, doi:.
Qin
,
J.
,
K.
Yang
,
S.
Liang
, and
X.
Guo
,
2009
:
The altitudinal dependence of recent rapid warming over the Tibetan Plateau
.
Climatic Change
,
97
,
321
327
, doi:.
Rawlins
,
M. A.
,
C. J.
Willmott
,
A.
Shiklomanov
,
E.
Linder
,
S.
Frolking
,
R. B.
Lammers
, and
C. J.
Vörösmarty
,
2006
:
Evaluation of trends in derived snowfall and rainfall across Eurasia and linkages with discharge to the Arctic Ocean
.
Geophys. Res. Lett.
,
33
, L07403, doi:.
Robock
,
A.
,
M.
Mu
,
K.
Vinnikov
, and
D.
Robinson
,
2003
:
Land surface conditions over Eurasia and Indian summer monsoon rainfall
.
J. Geophys. Res.
,
108
, 4131, doi:.
Su
,
F.
,
X.
Duan
,
D.
Chen
,
Z.
Hao
, and
L.
Guo
,
2013
:
Evaluation of the global climate models in the CMIP5 over the Tibetan Plateau
.
J. Climate
,
26
,
3187
3208
, doi:.
Takala
,
M.
,
J.
Pulliainen
,
S. J.
Metsämäki
, and
J. T.
Koskinen
,
2009
:
Detection of snowmelt using spaceborne microwave radiometer data in Eurasia from 1979 to 2007
.
IEEE Trans. Geosci. Remote Sens.
,
47
,
2996
3007
, doi:.
Tedesco
,
M.
, and
A. J.
Monaghan
,
2009
:
An updated Antarctic melt record through 2009 and its linkages to high-latitude and tropical climate variability
.
Geophys. Res. Lett.
,
36
, L18502, doi:.
Wang
,
G.
,
H.
Hu
, and
T.
Li
,
2009
:
The influence of freeze–thaw cycles of active soil layer on surface runoff in a permafrost watershed
.
J. Hydrol.
,
375
,
438
449
, doi:.
Wu
,
B.
, and
J.
Wang
,
2002
:
Winter Arctic oscillation, Siberian high and East Asian winter monsoon
.
Geophys. Res. Lett.
,
29
, 1897, doi:.
Xu
,
L.
,
L.
Dong
, and
Z.
Hu
,
2010
:
Relationship between the snow cover day and monsoon index in Tibetan Plateau
(in Chinese).
Plateau Meteor.
,
29
,
1093
1101
.
Yang
,
F.
,
A.
Kumar
,
W.
Wang
,
H.-M. H.
Juang
, and
M.
Kanamitsu
,
2001
:
Snow-albedo feedback and seasonal climate variability over North America
.
J. Climate
,
14
,
4245
4248
, doi:.
Yao
,
T.
, and Coauthors
,
2012
:
Different glacier status with atmospheric circulations in Tibetan Plateau and surroundings
.
Nat. Climate Change
,
2
,
663
667
, doi:.
Yasunari
,
T.
,
A.
Kitoh
, and
T.
Tokioka
,
1991
:
Local and remote responses to excessive snow mass over Eurasia appearing in the northern spring and summer climate—A study with the MRI-GCM
.
J. Meteor. Soc. Japan
,
69
,
473
487
.
Ye
,
H.
,
H.-R.
Cho
, and
P. E.
Gustafson
,
1998
:
The changes in Russian winter snow accumulation during 1936–83 and its spatial patterns
.
J. Climate
,
11
,
856
863
, doi:.
You
,
Q.
,
K.
Fraedrich
,
G.
Ren
,
N.
Repin
, and
S.
Kang
,
2013
:
Variability of temperature in the Tibetan Plateau based on homogenized surface stations and reanalysis data
.
Int. J. Climatol.
,
33
,
1337
1347
, doi:.
Yuan
,
W
.
,
B.
Xu
,
Z.
Chen
,
J.
Xia
,
W.
Xu
,
Y.
Chen
,
K.
Wu
, and
Y.
Fu
,
2015
:
Validation of China-wide interpolated daily climate variables from 1960 to 2011
.
Theor. Appl. Climatol.
,
119
,
689
700
, doi:.
Yuan
,
W.
,
W.
Xu
,
M.
Ma
,
S.
Chen
,
W.
Liu
, and
L.
Cui
,
2016
:
Improved snow cover model in terrestrial ecosystem models over the Qinghai–Tibetan Plateau
.
Agric. For. Meteor.
,
218–219
,
161
170
, doi:.
Zhang
,
K.
,
J. S.
Kimball
,
Q.
Mu
,
L. A.
Jones
,
S. J.
Goetz
, and
S. W.
Running
,
2009
:
Satellite based analysis of northern ET trends and associated changes in the regional water balance from 1983 to 2005
.
J. Hydrol.
,
379
,
92
110
, doi:.
Zhang
,
Y.
,
T.
Li
, and
B.
Wang
,
2004
:
Decadal change of the spring snow depth over the Tibetan Plateau: The associated circulation and influence on the East Asian summer monsoon
.
J. Climate
,
17
,
2780
2793
, doi:.
Zhao
,
H.
, and
G.
Moore
,
2004
:
On the relationship between Tibetan snow cover, the Tibetan Plateau monsoon and the Indian summer monsoon
.
Geophys. Res. Lett.
,
31
, L14204, doi:.

Footnotes

© 2017 American Meteorological Society. For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses).