• Bai, A., P. Zhai, and X. Liu, 2007: Climatology and trends of wet spells in China. Theor. Appl. Climatol., 88, 139148, https://doi.org/10.1007/s00704-006-0235-7.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Chen, H., 2013: Projected change in extreme rainfall events in China by the end of the 21st century using CMIP5 models. Chin. Sci. Bull., 58, 14621472, https://doi.org/10.1007/s11434-012-5612-2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Chen, H., and J. Sun, 2009: How the “best” models project the future precipitation change in China. Adv. Atmos. Sci., 26, 773782, https://doi.org/10.1007/s00376-009-8211-7.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Chen, H., J. Sun, X. Chen, and W. Zhou, 2012: CGCM projections of heavy rainfall events in China. Int. J. Climatol., 32, 441450, https://doi.org/10.1002/joc.2278.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Ding, Y. H., 2013: China’s Climate (in Chinese). Science Press, 557 pp.

  • Ding, Y. H., Z. Y. Wang, Y. F. Song, and J. Zhang, 2008: Causes of the unprecedented freezing disaster in January 2008 and its possible association with the global warming. Acta Meteor. Sin., 66, 808825.

    • Search Google Scholar
    • Export Citation
  • Feng, S., S. Nadarajah, and Q. Hu, 2007: Modeling annual extreme precipitation in China using the generalized extreme value distribution. J. Meteor. Soc. Japan, 85, 599613, https://doi.org/10.2151/jmsj.85.599.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Gao, H., 2009: China’s snow disaster in 2008, who is the principal player? Int. J. Climatol., 29, 21912196, https://doi.org/10.1002/joc.1859.

  • Gao, X., Y. Shi, D. Zhang, J. Wu, F. Giorgi, Z. Ji, and Y. Wang, 2012: Uncertainties in monsoon precipitation projections over China: Results from two high-resolution RCM simulations. Climate Res., 52, 213226, https://doi.org/10.3354/cr01084.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Gao, X., Y. Shi, Z. Han, M. Wang, J. Wu, D. Zhang, Y. Xu, and F. Giorgi, 2017: Performance of RegCM4 over major river basins in China. Adv. Atmos. Sci., 34, 441455, https://doi.org/10.1007/s00376-016-6179-7.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Giorgi, F., C. Jones, and G. R. Asrar, 2009: Addressing climate information needs at the regional level: The CORDEX framework. WMO Bull., 58, 175183.

    • Search Google Scholar
    • Export Citation
  • Giorgi, F., and Coauthors, 2012: RegCM4: Model description and preliminary tests over multiple CORDEX domains. Climate Res., 52, 729, https://doi.org/10.3354/cr01018.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Han, Z., B. Zhou, Y. Xu, J. Wu, and Y. Shi, 2017: Projected changes in haze pollution potential in China: An ensemble of regional climate model simulations. Atmos. Chem. Phys., 17, 10 10910 123, https://doi.org/10.5194/acp-17-10109-2017.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • IPCC, 2013: Climate Change 2013: The Physical Science Basis. Cambridge University Press, 1535 pp.

  • Jiang, D., and Z. Tian, 2013: East Asian monsoon change for the 21st century: Results of CMIP3 and CMIP5 models. Chin. Sci. Bull., 58, 14271435, https://doi.org/10.1007/s11434-012-5533-0.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Jiang, D., Z. Tian, and X. Lang, 2016: Reliability of climate models for China through the IPCC Third to Fifth Assessment Reports. Int. J. Climatol., 36, 11141133, https://doi.org/10.1002/joc.4406.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Liu, B., M. Xu, M. Henderson, and Y. Qi, 2005: Observed trends of precipitation amount, frequency, and intensity in China, 1960–2000. J. Geophys. Res., 110, D08103, https://doi.org/10.1029/2004JD004864.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Liu, N., J. Liu, Z. Zhang, H. Chen, and M. Song, 2012: Is extreme Arctic sea ice anomaly in 2007 a key contributor to severe January 2008 snowstorm in China? Int. J. Climatol., 32, 20812087, https://doi.org/10.1002/joc.2400.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Liu, Y., G. Ren, and H. Yu, 2012: Climatology of snow in China (in Chinese). Dili Kexue, 32, 11761185.

  • Liu, Y., G. Ren, H. Yu, and H. Kang, 2013: Climatic characteristics of intense snowfall in China with its variation (in Chinese). J. Appl. Meteor. Sci., 24, 304313.

    • Search Google Scholar
    • Export Citation
  • Moss, R. H., and Coauthors, 2010: The next generation of scenarios for climate change research and assessment. Nature, 463, 747756, https://doi.org/10.1038/nature08823.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Qin, D. H., W. J. Dong, and Y. Luo, 2012: Climate and Environmental Change in China: The Physical Science Basis (in Chinese). China Meteorological Press, 432 pp.

  • Song, R., X. Gao, Y. Shi, D. Zhang, and X. Zhang, 2008: Simulation of changes in cold events in southern China under global warming (in Chinese). Adv. Climate Change Res., 4, 352356.

    • Search Google Scholar
    • Export Citation
  • Sun, B., and H. Wang, 2013: Water vapor transport paths and accumulation during widespread snowfall events in northeastern China. J. Climate, 26, 45504566, https://doi.org/10.1175/JCLI-D-12-00300.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Sun, J., and J. Ao, 2013: Changes in precipitation and extreme precipitation in a warming environment in China. Chin. Sci. Bull., 58, 13951401, https://doi.org/10.1007/s11434-012-5542-z.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Sun, J., H. Wang, W. Yuan, and H. Chen, 2010: Spatial-temporal features of intense snowfall events in China and their possible change. J. Geophys. Res., 115, D16110, https://doi.org/10.1029/2009JD013541.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Sun, X. Z., Y. Luo, X. Zhang, and Y. X. Gao, 2010: Analysis on snowfall change characteristic of China in recent 46 years (in Chinese). Plateau Meteor., 29, 15941601.

    • Search Google Scholar
    • Export Citation
  • Taylor, K. E., R. J. Stouffer, and G. A. Meehl, 2012: An overview of CMIP5 and the experiment design. Bull. Amer. Meteor. Soc., 93, 485498, https://doi.org/10.1175/BAMS-D-11-00094.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wang, H., and S. He, 2013: The increase of snowfall in northeast China after the mid-1980s. Chin. Sci. Bull., 58, 13501354, https://doi.org/10.1007/s11434-012-5508-1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wang, H., E. Yu, and S. Yang, 2011: An exceptionally heavy snowfall in northeast China: Large-scale circulation anomalies and hindcast of the NCAR WRF Model. Meteor. Atmos. Phys., 113, 1125, https://doi.org/10.1007/s00703-011-0147-7.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wang, H., and Coauthors, 2012: Extreme climate in China: Facts, simulation and projection. Meteor. Z., 21, 279304, https://doi.org/10.1127/0941-2948/2012/0330.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wang, H., S. He, and J. Liu, 2013: Present and future relationship between the East Asian winter monsoon and ENSO: Results of CMIP5. J. Geophys. Res. Oceans, 118, 52225237, https://doi.org/10.1002/jgrc.20332.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wang, Y., B. Zhou, D. Qin, J. Wu, R. Gao, and L. Song, 2017: Changes in mean and extreme temperature and precipitation over the arid region of northwestern China: Observation and projection. Adv. Atmos. Sci., 34, 287305, https://doi.org/10.1007/s00376-016-6160-5.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wang, Z., S. Yang, and B. Zhou, 2017: Preceding features and relationship with possible affecting factors of persistent and extensive icing events in China. Int. J. Climatol., 37, 41054118, https://doi.org/10.1002/joc.5026.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wen, M., S. Yang, A. Kumar, and P. Zhang, 2009: An analysis of the large-scale climate anomalies associated with the snowstorms affecting China in January 2008. Mon. Wea. Rev., 137, 11111131, https://doi.org/10.1175/2008MWR2638.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wu, J., B. Zhou, and Y. Xu, 2015a: Response of precipitation and its extremes over China to warming: CMIP5 simulation and projection. Chin. J. Geophys., 58, 461473, https://doi.org/10.1002/cjg2.20187.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wu, J., X. Gao, Y. Xu, and J. Pan, 2015b: Regional climate change and uncertainty analysis based on four regional climate model simulations over China. Atmos. Oceanic Sci. Lett., 8, 147152, https://doi.org/10.3878/AOSL20150013.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Xu, C.-H., and Y. Xu, 2012: The projection of temperature and precipitation over China under RCP scenarios using a CMIP5 multi-model ensemble. Atmos. Oceanic Sci. Lett., 5, 527533, https://doi.org/10.1080/16742834.2012.11447042.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • You, Q., S. Kang, E. Aguilar, and Y. Yan, 2008: Changes in daily climate extremes in the eastern and central Tibetan Plateau during 1961–2005. J. Geophys. Res., 113, D07101, https://doi.org/10.1029/2007JD009389.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Zhai, P., X. Zhang, H. Wan, and X. Pan, 2005: Trends in total precipitation and frequency of daily precipitation extremes over China. J. Climate, 18, 10961108, https://doi.org/10.1175/JCLI-3318.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Zhang, Q., C.-Y. Xu, Z. Zhang, Y. D. Chen, and C.-L. Liu, 2009: Spatial and temporal variability of precipitation over China, 1951–2005. Theor. Appl. Climatol., 95, 5368, https://doi.org/10.1007/s00704-007-0375-4.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Zhang, R., M. Hong, K. Liu, and Y. Chen, 2012: Subtropical high circulation background and its variation characters in a serious cold rain-snow frost disaster in winter of 2007/2008. Trans. Atmos. Sci., 35, 19.

    • Search Google Scholar
    • Export Citation
  • Zhou, B. T., Q. H. Wen, Y. Xu, L. Song, and X. Zhang, 2014: Projected changes in temperature and precipitation extremes in China by the CMIP5 multimodel ensembles. J. Climate, 27, 65916611, https://doi.org/10.1175/JCLI-D-13-00761.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Zhou, B. T., Y. Xu, J. Wu, S. Dong, and Y. Shi, 2016: Changes in temperature and precipitation extreme indices over China: Analysis of a high-resolution grid dataset. Int. J. Climatol., 36, 10511066, https://doi.org/10.1002/joc.4400.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Zhou, B. T., Z. Wang, and Y. Shi, 2017: Possible role of Hadley circulation strengthening in interdecadal intensification of snowfalls over northeastern China under climate change. J. Geophys. Res. Atmos., 122, 11 63811 650, https://doi.org/10.1002/2017JD027574.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Zhou, B. Z., and Coauthors, 2011: The great 2008 Chinese ice storm: Its socioeconomic–ecological impact and sustainability lessons learned. Bull. Amer. Meteor. Soc., 92, 4760, https://doi.org/10.1175/2010BAMS2857.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • View in gallery
    Fig. 1.

    Map showing the topography (shading; unit: m) of China, distribution of stations (dots), and domains of the four subregions (gray lines): northwestern China (NWC; west of 90°E; north of 40°N, including 34 stations), northeastern China (NEC; east of 120°E; north of 40°N, including 81 stations), the eastern Tibetan Plateau (ETP; 87.5°–105°E; 28°–35°N, including 62 stations), and southeast China (SEC; east of 110°E; 28°–38°N, including 118 stations).

  • View in gallery
    Fig. 2.

    (a) Climatology (mm) and (b) standard deviation (mm) of the snowfall amount (based on snow water equivalent) during the wintertime of 1961–2013 in the observations.

  • View in gallery
    Fig. 3.

    Climatology of (a) the number of snow days (days yr−1) and (b) the mean snowfall intensity (mm day−1) during the wintertime of 1961–2013 in the observations.

  • View in gallery
    Fig. 4.

    Linear trends of (a) snowfall amounts (mm decade−1, based on snow water equivalent), (b) snow days (days decade−1), and (c) mean snowfall intensities (mm day−1 decade−1) during the wintertime of 1961–2013 in the observations. Only the trends significant at the 95% level are shown.

  • View in gallery
    Fig. 5.

    (a) Ratio (%) of the four categories of snowfall events to the total snowfall amount (based on snow water equivalent) during the wintertime of 1961–2013 in the observations. (b) Standard deviation (mm) of the amount (based on snow water equivalent) of the four categories of snowfall events during the wintertime of 1961–2013 in the observations.

  • View in gallery
    Fig. 6.

    Trends (bars) of (a) amounts (% decade−1, based on snow water equivalent), (b) days (% decade−1), and (c) mean intensities (mm day−1 decade−1) of different categories of snowfall events in the four subregions during the wintertime of 1961–2013 in the observations. Asterisks indicate the trends significant at the 95% level.

  • View in gallery
    Fig. 7.

    Trends of the ratio (% per decade) of different categories of snowfall events to (a) the total snowfall amount (based on snow water equivalent) and (b) the total number of snow days in the four subregions during the wintertime of 1961–2013 in the observations. Asterisks indicate the trends significant at the 95% level.

  • View in gallery
    Fig. 8.

    Spatial distribution of (a),(b) snowfall amounts (%), (c),(d) snow days (%), and (e),(f) mean intensities (mm day−1) during the wintertime of 1986–2005, for (left) observations and (right) the ensemble simulation. Considering RegCM4’s wet bias in modeling precipitation and systematic bias caused by different measures between the simulations and the observations to identify snowfall, the snowfall amounts and snow days in each station are expressed as the percentage divided by their sums across all of China.

  • View in gallery
    Fig. 9.

    The ensemble projected percentage changes (relative to 1986–2005) of (a) amounts, (b) days, and (c) mean intensities of the total snowfall during the wintertime of 2080–99 under RCP4.5. Regions where all the ensemble members agree on the sign of change are hatched.

  • View in gallery
    Fig. 10.

    As in Fig. 9, but for the amount of different categories of snowfall events.

  • View in gallery
    Fig. 11.

    As in Fig. 9, but for the number of days of different categories of snowfall events.

  • View in gallery
    Fig. 12.

    As in Fig. 9, but for the ratio of different categories of snowfall events to the total snowfall amount (left-hand panels) and to the total number of snow days (right-hand panels).

  • View in gallery
    Fig. 13.

    Scatterplots of the observed (a) surface air temperature (°C; abscissa) vs snow days (days; ordinate) and (b) specific humidity (g kg−1; abscissa) vs snowfall intensity (mm day−1; ordinate) over stations during the wintertime of 1961–2013. Each dot represents the corresponding values for one station in each winter. The snow days are calculated as the total number of the days with snowfall occurring in the wintertime, and the surface air temperature, specific humidity, and snowfall intensity are calculated as their respective averages for the days with snowfall occurring in the wintertime. The red line indicates the best linear fit.

  • View in gallery
    Fig. 14.

    As in Fig. 13, but for scatterplots of the ensemble projected (a) surface air temperature (°C; abscissa) vs snow days (days; ordinate) and (b) specific humidity (g kg−1; abscissa) vs snowfall intensity (mm day−1; ordinate) over stations during the wintertime of 2080–99 under RCP4.5 (relative to 1986–2005).

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 768 362 30
PDF Downloads 675 285 30

Historical and Future Changes of Snowfall Events in China under a Warming Background

Botao ZhouCollaborative Innovation Center on Forecast and Evaluation of Meteorological Disasters/Key Laboratory of Meteorological Disaster, Ministry of Education, and School of Atmospheric Sciences, Nanjing University of Information Science and Technology, Nanjing, and National Climate Center, China Meteorological Administration, Beijing, China

Search for other papers by Botao Zhou in
Current site
Google Scholar
PubMed
Close
,
Zunya WangNational Climate Center, China Meteorological Administration, Beijing, China

Search for other papers by Zunya Wang in
Current site
Google Scholar
PubMed
Close
,
Ying ShiNational Climate Center, China Meteorological Administration, Beijing, China

Search for other papers by Ying Shi in
Current site
Google Scholar
PubMed
Close
,
Ying XuNational Climate Center, China Meteorological Administration, Beijing, China

Search for other papers by Ying Xu in
Current site
Google Scholar
PubMed
Close
, and
Zhenyu HanNational Climate Center, China Meteorological Administration, Beijing, China

Search for other papers by Zhenyu Han in
Current site
Google Scholar
PubMed
Close
Open access

Abstract

Using station data and Regional Climate Model version 4 (RegCM4) simulations under the representative concentration pathway 4.5 (RCP4.5) scenario, this article addresses historical and future changes of the wintertime snowfall over China. The observational results generally show a decrease in the frequency and an increase in the mean intensity of snowfalls in northwestern China (NWC), northeastern China (NEC), the eastern Tibetan Plateau (ETP), and southeastern China (SEC) since the 1960s. The total amount of wintertime snowfall, however, has increased in NWC, NEC, and ETP but decreased in SEC. The decrease in snow days is primarily due to the reduction of light snowfall events. The increase in the total amount is primarily explained by increases in heavy snowfalls, and the corresponding decrease is the result of decreases in light-to-heavy snowfalls. The RegCM4 ensemble, which can well simulate the observed snowfall climatology, projects that the snow days will be further reduced by the end of the twenty-first century relative to 1986–2005, primarily owing to the decline of light snowfall events. The total amount is projected to increase in NWC but decrease in the other three subregions. The increase in the total amount in NWC is attributed to increases in heavy and large snowfalls. Decreases in light snowfalls play a leading role in the decrease of the total amount in NEC. In ETP and SEC, the decrease in the total amount is the result of overall decreases in light-to-heavy snowfalls. The mechanism for such changes is an interesting topic to study in the future.

Denotes content that is immediately available upon publication as open access.

© 2018 American Meteorological Society. For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses).

Corresponding author: Zunya Wang, wangzy@cma.gov.cn

Abstract

Using station data and Regional Climate Model version 4 (RegCM4) simulations under the representative concentration pathway 4.5 (RCP4.5) scenario, this article addresses historical and future changes of the wintertime snowfall over China. The observational results generally show a decrease in the frequency and an increase in the mean intensity of snowfalls in northwestern China (NWC), northeastern China (NEC), the eastern Tibetan Plateau (ETP), and southeastern China (SEC) since the 1960s. The total amount of wintertime snowfall, however, has increased in NWC, NEC, and ETP but decreased in SEC. The decrease in snow days is primarily due to the reduction of light snowfall events. The increase in the total amount is primarily explained by increases in heavy snowfalls, and the corresponding decrease is the result of decreases in light-to-heavy snowfalls. The RegCM4 ensemble, which can well simulate the observed snowfall climatology, projects that the snow days will be further reduced by the end of the twenty-first century relative to 1986–2005, primarily owing to the decline of light snowfall events. The total amount is projected to increase in NWC but decrease in the other three subregions. The increase in the total amount in NWC is attributed to increases in heavy and large snowfalls. Decreases in light snowfalls play a leading role in the decrease of the total amount in NEC. In ETP and SEC, the decrease in the total amount is the result of overall decreases in light-to-heavy snowfalls. The mechanism for such changes is an interesting topic to study in the future.

Denotes content that is immediately available upon publication as open access.

© 2018 American Meteorological Society. For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses).

Corresponding author: Zunya Wang, wangzy@cma.gov.cn

1. Introduction

The global mean surface temperature is known to have risen since the second half of the twentieth century. Accompanying this warming, both the mean and extreme precipitation events have exhibited notable changes over a wide range of regions across the world and have exerted profound impacts on natural and human systems (IPCC 2013). Therefore, the past and future changes in precipitation and its related extremes are of particular concern.

In recent years, a great number of studies have been devoted to this issue in China and huge progress has been made (e.g., Liu et al. 2005; Zhai et al. 2005; Bai et al. 2007; Feng et al. 2007; You et al. 2008; Zhang et al. 2009; Chen et al. 2012; Wang et al. 2012; Xu and Xu 2012; Chen 2013; Sun and Ao 2013; Zhou et al. 2014, 2016; Wu et al. 2015a; Jiang et al. 2016; Y. Wang et al. 2017). For example, it is well documented that the extreme precipitation in northwestern and eastern China has been increasing and that the extreme precipitation in the southwest–northeast belt from southwestern China to northeastern China has been decreasing (Zhai et al. 2005; Bai et al. 2007; Feng et al. 2007; Wang et al. 2012; Zhou et al. 2016). In addition, annual precipitation over southwestern, northwestern, and eastern China have increased over the past decades, with corresponding decreases over central, northern, and northeastern China (Liu et al. 2005; Chen and Sun 2009; Wang et al. 2012; Y. Wang et al. 2017). Based on the simulations of global climate models (GCMs) within phase 5 of the Coupled Model Intercomparison Project (CMIP5) under different representative concentration pathways (RCPs), an increase in the mean precipitation and more frequent and intense precipitation extremes are anticipated for China by the end of this century (Xu and Xu 2012; Chen 2013; Wu et al. 2015a; Zhou et al. 2014; Y. Wang et al. 2017). These findings have greatly improved our knowledge of climate change in China.

However, the changes in snowfall have yet to be systematically addressed. In the context of climate warming, intense snowfalls have hit China frequently in recent winters and have caused severe damages to the sustainability of the society. For instance, a severe snowfall event that occurred in southern China in early 2008 led to an economic loss of above 20 billion U.S. dollars and affected more than 100 million people (Zhou et al. 2011). Given the serious impacts of this snowfall event, investigating present and future changes in snowfall is an urgent necessity for disaster prevention and mitigation.

Until now, studies concerning snowfall variations in China have mostly focused on the influential factors and associated physical processes (e.g., Ding et al. 2008; Gao 2009; Wen et al. 2009; Wang et al. 2011; Zhang et al. 2012; Sun and Wang 2013; Z. Wang et al. 2017). There are relatively few works on the secular changes. Wang and He (2013) indicated an increase in the snowfall over northeastern China after the mid-1980s. J. Sun et al. (2010) revealed that intense events with daily snowfall of no less than 5 mm increased in northern Xinjiang and in the eastern Tibetan Plateau but decreased in eastern China during 1962–2000. However, some issues remain unknown. For instance, how have different types of snowfall events (i.e., light, moderate, large, and heavy snowfall) changed in the past decades? How large are their respective contributions to the change of the total snowfall?

Another issue is the future changes of the snowfall in China in warmer scenarios. Based on four CMIP3 GCMs’ simulations under the A1B and A2 scenarios, the frequencies of intense snowfall events are projected to decrease over southern China but increase initially and then decrease over northern China during the twenty-first century (J. Sun et al. 2010). The projection of the snowfall change over southern China from a high-resolution regional climate model (RCM) shows a similar feature but with more local small-scale information (Song et al. 2008). Compared with CMIP3, CMIP5 is characterized by more reasonable RCP scenarios (Moss et al. 2010; IPCC 2013) and substantial model improvements for climate projections (Taylor et al. 2012; Xu and Xu 2012; IPCC 2013; Jiang et al. 2016). In addition, RCMs with higher resolutions are demonstrated to outperform GCMs on a regional scale (Wu et al. 2015b; Gao et al. 2012, 2017). The question arises as to how the snowfall in China will change under the RCP scenarios in the RCM projection.

Thus, with the aim of addressing these gaps, this study attempts to investigate the observed changes in snowfall in China, including its amount, frequency, and intensity as well as the contribution from different types of events over the past decades. Their future changes at the end of this century under RCP4.5 are also projected based on an RCM ensemble simulation. We expect the findings of this study to deepen our understanding of the snowfall changes in China and to provide more comprehensive and detailed information for disaster prevention and mitigation.

2. Observational data, RCM simulations, and methods

The observed daily snowfall data for the period of 1961–2014 from 836 meteorological stations in China, compiled by the China Meteorological Administration (CMA) after quality control, are used in this study. Snowfall is identified using the weather phenomena record. Because of some missing data in certain stations, we eliminated the stations with records of less than 30 years, leaving a total of 610 stations (Fig. 1) for use in the analysis.

Fig. 1.
Fig. 1.

Map showing the topography (shading; unit: m) of China, distribution of stations (dots), and domains of the four subregions (gray lines): northwestern China (NWC; west of 90°E; north of 40°N, including 34 stations), northeastern China (NEC; east of 120°E; north of 40°N, including 81 stations), the eastern Tibetan Plateau (ETP; 87.5°–105°E; 28°–35°N, including 62 stations), and southeast China (SEC; east of 110°E; 28°–38°N, including 118 stations).

Citation: Journal of Climate 31, 15; 10.1175/JCLI-D-17-0428.1

The RCM applied is the RegCM4, which was developed by the International Centre for Theoretical Physics (ICTP) (Giorgi et al. 2012). This model has a horizontal resolution of 25 km and 18 vertical sigma layers with the top at 50 hPa. The RegCM4 simulations are driven by the historical (1979–2005) and RCP4.5 (2006–99) simulations at 6-h time steps from four CMIP5 GCMs, namely the Commonwealth Scientific and Industrial Research Organization Mark 3.6.0 (CSIRO-Mk3.6.0), the European Centre Earth System Model (EC-EARTH), the Hadley Centre Global Environment Model, version 2 with Earth System configuration (HadGEM2-ES), and the Max Planck Institute Earth System Model, medium resolution (MPI-ESM-MR). To drive RCM modeling, the ratio of the resolution between GCMs and RCMs should not exceed 6–8; that is, only those GCMs with a resolution of 1°–2° or higher can be used to drive the RegCM4 despite the fact that about 20 GCMs in the CMIP5 provide the 6-h outputs for dynamical downscaling (Han et al. 2017). The reduction in the potential pool of GCMs to downscale is based on the assumption of no inner nesting. Because of the availability of CMIP5 GCMs and considering huge volume of outputs for ~120-yr RegCM4 simulations, we just used these four GCMs for this study. Note that the set of 18 vertical sigma layers is the standard configuration of the RegCM4 and has been widely used for RegCM4 simulations. Although increasing vertical resolution could improve model performance, the impact from the increase of vertical layers is not considered in this study, given the huge volume of outputs for the long-term integration. The historical simulation represents the past climate, and the RCP4.5 indicates a medium-low emission scenario with the radiative forcing peaking at 4.5 W m−2 by 2100 (Moss et al. 2010; Taylor et al. 2012). The domain in the RegCM4 simulations covers China and adjacent regions, as recommended by the Coordinated Regional Climate Downscaling Experiment (CORDEX)-East Asia phase II (Giorgi et al. 2009).

Since the RegCM4 simulations do not output snowfall separately, we identified snowfall events with the following criteria as applied in J. Sun et al. (2010): 1) daily precipitation no less than 0.1 mm; 2) daily mean surface air temperature below 0°C; 3) ground temperature below 0°C. The last two criteria ensure that the precipitation falls in the form of snow. As the observed weather phenomena record is on the daily scale, to match the observations we use daily mean temperature and precipitation as the threshold for snowfall in the simulations, although the RegCM4 simulations output subdaily data at a 6-h time step.

For the projection, we focus on the ensemble results at the end of the twenty-first century (2080–99). The ensemble mean is simply taken as an arithmetic average of the four simulations. The historical simulation during 1986–2005, commonly adopted for the CMIP5 evaluation and projection, is employed to evaluate the fidelity of the RegCM4 and to act as a reference for the projection.

The snowfall events are divided into four categories according to the CMA’s classification based on snow water equivalent: light snowfall (0.1–2.5 mm day−1), moderate snowfall (2.5–5 mm day−1), large snowfall (5–10 mm day−1), and heavy snowfall (>10 mm day−1), in order to examine the changes in their amounts, frequencies, intensities, ratios, and contributions. The amount is defined as the sum of the snowfall falling on all days affiliated in each category. The frequency is represented by the number of snow days involved in each category. The intensity is taken as the amount divided by the frequency for each category. The ratio is calculated as the snowfall of each category divided by the total snowfall. The contribution is measured as the percentage of the snowfall change in each category to the total snowfall change. In this study, the wintertime refers to the period from October to April of the following year, during which snowfall mainly occurs in China. Trends are calculated with the least squares method and the correlation factors test for their significance (p < 0.05) with the Student’s t test.

3. Observations

Figure 2a shows the spatial distribution of the snowfall amount averaged over the winters of 1961–2013. The result clearly indicates that snowfall amounts greater than 30 mm are mainly located in northwestern China, northeastern China, and the eastern Tibetan Plateau. These amounts may largely result from the frequent occurrences of snowfall. As shown in Fig. 3a, the greatest number of snow days is also observed in these regions. Our results are consistent with previous studies (Y. Liu et al. 2012; Qin et al. 2012; Ding 2013). The high snowfall in the three regions may reflect the influence of high latitudes and high elevations. As the northwestern and northeastern regions of China are located in the high latitudes (Fig. 1), cold air intrusions frequently occur there, thereby resulting in frequent occurrences and great amounts of snowfall. The eastern Tibetan Plateau is located at the high elevations (Fig. 1). Local low temperatures and relatively high moisture may cause high snowfall there (Y. Liu et al. 2012; Qin et al. 2012; Ding 2013).

Fig. 2.
Fig. 2.

(a) Climatology (mm) and (b) standard deviation (mm) of the snowfall amount (based on snow water equivalent) during the wintertime of 1961–2013 in the observations.

Citation: Journal of Climate 31, 15; 10.1175/JCLI-D-17-0428.1

Fig. 3.
Fig. 3.

Climatology of (a) the number of snow days (days yr−1) and (b) the mean snowfall intensity (mm day−1) during the wintertime of 1961–2013 in the observations.

Citation: Journal of Climate 31, 15; 10.1175/JCLI-D-17-0428.1

The distribution of the standard deviation of the snowfall amount is displayed in Fig. 2b. It is noted that the standard deviation pattern shows a similar feature to that of the climatology; that is, a strong variability appears over northwestern China, northeastern China, and the eastern Tibetan Plateau. In addition, another center of large variability is observed in southeastern China (Fig. 2b) where the climatological snowfall amount is somewhat smaller than that in the aforementioned three regions (Fig. 2a). The snowfall intensity in southeastern China is also much stronger despite relatively fewer snow days (Fig. 3). The fewer snow days in southeastern China are mainly due to higher local temperature climatology. Relatively high temperature in the wintertime tends to prevent precipitation from falling in the form of snow. Nevertheless, the moisture is abundant in southeastern China. Once intense cold air masses break out southward and help lower the local temperature below 0°C, it is easy to induce relatively intense snowfall (Y. Liu et al. 2012; Qin et al. 2012; Ding 2013).

Figure 4 plots the linear trends of the snowfall amounts, snow days, and mean intensities during the wintertime of 1961–2013 for each station in China. For the snowfall amount, there are 50 stations showing significant increasing trends. These stations are mainly distributed in northwestern China, northeastern China, and the eastern Tibetan Plateau. The significant decreasing trends are observed over 45 stations that are mainly scattered around the Yellow River valley and the Yangtze River valley, in the southern part of the Tibetan Plateau, and along the coast of northeastern China (Fig. 4a). X. Sun et al. (2010) analyzed the change of annual mean snowfall during 1960–2005 and found that the annual snowfall amount also increased in northeastern and northwestern China but decreased in the Yangtze River valley. The number of snow days shows an overall decreasing trend across China, particularly in northern China and western China. In total, there are 284 stations with significant decreasing trends. Only one station shows a significant increasing trend (Fig. 4b). In addition, significant increasing trends in the mean snowfall intensity have been observed over 178 stations (Fig. 4c). These results imply that snowfall has generally occurred less frequently but with enhanced intensity over the past decades under the background of global warming.

Fig. 4.
Fig. 4.

Linear trends of (a) snowfall amounts (mm decade−1, based on snow water equivalent), (b) snow days (days decade−1), and (c) mean snowfall intensities (mm day−1 decade−1) during the wintertime of 1961–2013 in the observations. Only the trends significant at the 95% level are shown.

Citation: Journal of Climate 31, 15; 10.1175/JCLI-D-17-0428.1

To conduct a more detailed and quantitative examination of the regional changes in the different types of snowfall events in China, based on the above analyses we defined four typical climatological subregions (Fig. 1): northwestern China (NWC; west of 90°E and north of 40°N), northeastern China (NEC; east of 120°E and north of 40°N), the eastern Tibetan Plateau (ETP; 87.5°–105°E; 28°–35°N), and southeastern China (SEC; east of 110°E and 28°–38°N). On the regional average, the light snowfall events account for almost half of the total amount of snowfall in NWC, NEC, and ETP, followed by the contributions from moderate, large, and heavy snowfall sequentially. In contrast, in SEC, the ratios to the total snowfall amount are comparable for the four categories (Fig. 5a). In addition, among the four categories, relatively larger standard deviation is noted for the light snowfall amount in NWC, NEC, and ETP, and for the heavy snowfall amount in SEC (Fig. 5b).

Fig. 5.
Fig. 5.

(a) Ratio (%) of the four categories of snowfall events to the total snowfall amount (based on snow water equivalent) during the wintertime of 1961–2013 in the observations. (b) Standard deviation (mm) of the amount (based on snow water equivalent) of the four categories of snowfall events during the wintertime of 1961–2013 in the observations.

Citation: Journal of Climate 31, 15; 10.1175/JCLI-D-17-0428.1

Figure 6 shows the linear trends of the amounts, days, and mean intensities of different categories of snowfall events in the four subregions. All indices except the mean intensity are expressed as percentage anomalies relative to the climatologies of 1961–2013. As shown in Fig. 6, some notable differences can be seen among the subregions. Over the past 53 years, the total snowfall amounts have increased in NWC and NEC at rates of 9.7% and 3.6% decade−1 (p < 0.05), respectively. The ETP also experiences a tendency toward increasing but, due to the decreases observed over some stations in its southern flank, the increasing rate (0.5% decade−1) averaged over the whole region is not significant. Contrary to the changes in NWC, NEC, and ETP, the total snowfall amount in SEC has decreased at a rate of 2.7% decade−1. The total frequency and mean intensity show uniform decreases and increases, respectively, in the four subregions. Therefore, changes in the snowfall intensity may play a major role in the increase of the total snowfall amount in NEC, NWC, and ETP, while changes in the snowfall frequency may play a prominent role in the decrease of the total snowfall amount in SEC.

Fig. 6.
Fig. 6.

Trends (bars) of (a) amounts (% decade−1, based on snow water equivalent), (b) days (% decade−1), and (c) mean intensities (mm day−1 decade−1) of different categories of snowfall events in the four subregions during the wintertime of 1961–2013 in the observations. Asterisks indicate the trends significant at the 95% level.

Citation: Journal of Climate 31, 15; 10.1175/JCLI-D-17-0428.1

As for the secular changes in the four categories of snowfall events, overall increases in the amounts, frequencies, and intensities are detected in NWC, except for a decrease in the frequency of light snowfall events. During 1961–2013, the increasing rates related to the moderate, large, and heavy snowfall events are 11.0%, 17.1%, and 29.8% decade−1 for their amounts, 10.6%, 17.3%, and 29.4% decade−1 for their frequencies, and 0.08, 0.26, and 0.34 mm day−1 decade−1 for their intensities, respectively. These are all significant at p < 0.05. Thus, changes in both their frequencies and intensities are responsible for the enhancement of the moderate-to-heavy snowfall amounts. For the light snowfall events, an increase in the amount concurrent with a decrease in the frequency and an increase in the intensity implies that the enhancement of the amount is mainly due to the strengthening of the intensity across the snowfall distribution. Moreover, through a comparison of the changes between the total snowfall and the four categories of snowfall events, the changes in the light-to-heavy snowfall amounts can be concluded to positively contribute to the augmentation of the total snowfall amount; the largest contribution is from the heavy snowfall amount. The decline in the total number of snow days mainly results from the reduction of light snowfall events. In addition, the ratio of the light snowfall to the total snowfall amount and frequency in the wintertime of each year has declined at a rate of 3.3% and 1.2% decade−1 (p < 0.05), respectively. In contrast, the ratios of the large and heavy snowfall events have increased significantly (Fig. 7).

Fig. 7.
Fig. 7.

Trends of the ratio (% per decade) of different categories of snowfall events to (a) the total snowfall amount (based on snow water equivalent) and (b) the total number of snow days in the four subregions during the wintertime of 1961–2013 in the observations. Asterisks indicate the trends significant at the 95% level.

Citation: Journal of Climate 31, 15; 10.1175/JCLI-D-17-0428.1

Similar changes are observed for the moderate-to-heavy snowfall events in NEC but with a smaller amplitude (Fig. 6). From 1961 to 2013, the cumulative amounts of moderate, large, and heavy snowfall have increased by 33.9%, 39.8%, and 43.5%, respectively, consequently resulting in an enhancement of the total amount. The corresponding numbers of snow days have also increased by 33.9%, 39.2%, and 34.5%. Although changes of the light snowfall events resemble that in NWC—that is, with a decrease in the frequency (7.9% decade−1, p < 0.05) and an increase in the intensity (0.03 mm day−1 decade−1, p < 0.05)—the light snowfall amount has been reduced by 1.6% in NEC since the 1960s. However, this reduction is not significant. Additionally, Fig. 7 also indicates increasingly higher proportions of the moderate-to-heavy snowfall events and increasingly lower proportions of the light snowfall events in NEC. The situation in ETP bears a general resemblance to that in NEC.

In contrast with the changes in the three subregions above, consistent decreases in the amounts and frequencies are noticed for all four types of snowfall events in SEC (Fig. 5). The percentage change of the light snowfall events is the most pronounced, with a decreasing rate of 6.9% (12.4%) decade−1 (p < 0.05) for the amount (the frequency). The smallest decrease in the amount and frequency is seen for the heavy snowfall events. Their respective trends are −0.2% and −0.1% decade−1, and are not significant. Owing to the greater reduction of light snowfall events, the ratios of the large and heavy snowfall events to the total snowfall events have increased (Fig. 7).

4. Model evaluations and projections

a. Performance of the RegCM4 simulations

The fidelity of the RegCM4 in simulating the snowfall in China is briefly evaluated through the comparison with the observations. Figure 8 presents the observed and the ensemble simulated patterns of the snowfall amounts, snow days, and mean intensities during the wintertime of 1986–2005. To facilitate the comparison, the gridded simulation data are interpolated to the same stations as in the observations. Considering RegCM4’s wet bias in modeling the precipitation in China (Gao et al. 2017) and that different measures between the simulations and the observations to identify snowfall may also produce systematic bias, the snowfall amount and the number of snow days in each station are expressed as the percentage divided by their sums across all of China.

Fig. 8.
Fig. 8.

Spatial distribution of (a),(b) snowfall amounts (%), (c),(d) snow days (%), and (e),(f) mean intensities (mm day−1) during the wintertime of 1986–2005, for (left) observations and (right) the ensemble simulation. Considering RegCM4’s wet bias in modeling precipitation and systematic bias caused by different measures between the simulations and the observations to identify snowfall, the snowfall amounts and snow days in each station are expressed as the percentage divided by their sums across all of China.

Citation: Journal of Climate 31, 15; 10.1175/JCLI-D-17-0428.1

As shown in the figure, the ensemble simulation is noted as well capturing the observed spatial features with the greatest snowfall amount and the largest number of snow days in northwestern China, northeastern China, and the eastern Tibetan Plateau. The strong snowfall intensity over southeastern China from the observations can also be reproduced although it is overestimated in the ensemble simulation. The spatial correlation coefficients between the ensemble simulation and the observations are 0.50, 0.79, and 0.49 (p < 0.05), respectively, for the amount, frequency, and mean intensity of snowfall (Table 1). The RegCM4 simulation driven by the ERA-Interim reanalysis also captures the observed spatial features well. Their spatial correlation coefficients are 0.44, 0.79, and 0.50 (p < 0.05) for the snowfall amount, frequency, and intensity, respectively. For the simulations of the ensemble members, they all resemble the observations with the spatial correlation coefficients ranging from 0.51 to 0.55 (p < 0.05) for the snowfall amounts and 0.77 to 0.83 (p < 0.05) for the snow days. However, the spatial correlation coefficients are relatively lower for the snowfall intensities, which are in the range of 0.11 to 0.23 (Table 1). The ratios of different categories of snowfall events to the total snowfall are in general in agreement with the observations, although the heavy snowfall is overestimated (figure not shown).

Table 1.

Statistical results of spatial correlation coefficients between the observations and the historical simulations of the ensemble mean and its members for the snowfall amount, snow days and snowfall intensity during the period of 1986–2005. Asterisks indicate the spatial correlation coefficients significant at the 95% level.

Table 1.

In addition, the ensemble simulation can yield a large interannual standard deviation of the snowfall amount over northwestern China, northeastern China, the eastern Tibetan Plateau, and southeastern China as shown in the observations, although the standard deviation center is slightly westward in southeastern China (figure not shown). The spatial correlation coefficient between the ensemble simulation and the observations is 0.58 (p < 0.05) and that between the individual simulations and the observations varies from 0.49 to 0.64 (p < 0.05) (Table 1).

These results suggest that the RegCM4 ensemble simulation has a good performance in simulating the climatology and standard deviation of the snowfall in China. However, it shows relatively poor skills in modeling snowfall trends, particularly modeling the snowfall intensity trend during 1986–2005 (Table 1). The inconsistency between the historical simulations and the observations in the trends indicate that the observed trends may not be driven by external forcings. Compared with the performance on the snowfall climatology, the RegCM4 simulation driven by the ERA-Interim reanalysis also shows relatively lower skills in capturing snowfall trends (figure not shown).

b. Projected changes in snowfall

Figure 9 shows the ensemble projected percentage changes in the amount, frequency, and mean intensity of the total snowfall at the end of the twenty-first century under RCP4.5 relative to 1986–2005. The total amount is projected to increase in northwestern China and decrease in almost all the remaining regions (Fig. 9a). The area-averaged percentage changes are 3.6% in NWC, −3.9% in NEC, −15.0% in ETP (p < 0.05), and −33.0% in SEC (p < 0.05) (Table 2). In addition, the projections exhibit an overall decrease in the snowfall frequency (Fig. 9b) and a general increase in the snowfall intensity (Fig. 9c). By the end of the twenty-first century, the snow days (the mean intensity) tend to decrease (increase) by 8.9% (13.7%), 10.2% (7.0%), 17.7% (3.3%), and 35.9% (4.4%) in NWC, NEC, ETP, and SEC, respectively (Tables 3 and 4). These changes are all significant at p < 0.05. Therefore, the increase in the total amount may be the consequence of the increase in the mean intensity, while the decreases in the amount are highly related to the decreases in the frequency.

Fig. 9.
Fig. 9.

The ensemble projected percentage changes (relative to 1986–2005) of (a) amounts, (b) days, and (c) mean intensities of the total snowfall during the wintertime of 2080–99 under RCP4.5. Regions where all the ensemble members agree on the sign of change are hatched.

Citation: Journal of Climate 31, 15; 10.1175/JCLI-D-17-0428.1

Table 2.

The ensemble projected percent changes of the snowfall amount at the end of the twenty-first century under RCP4.5 relative to 1986–2005 in the four subregions, with the spread of projections from the ensemble members shown in square brackets. Values in the parentheses indicate the contributions of snowfall changes of a certain category to the total snowfall change. Asterisks indicate the ensemble projected changes significant at the 95% level. Unit: %.

Table 2.
Table 3.

As in Table 2, but for percent changes of the snowfall frequency.

Table 3.
Table 4.

As in Table 2, but for percent changes of the snowfall intensity.

Table 4.

The four ensemble members agree on the signs of their changes in most of the regions. However, some inconsistencies are also noticed, for instance for the total snowfall amounts in parts of northeastern China, northwestern China, and the southwestern Tibetan Plateau, and for the snowfall intensity in the southern Tibetan Plateau. The former inconstancies are due to the differences of the RegCM4 projection driven by EC-EARTH, by MPI-ESM-MRC, and by CSIRO-Mk-3.6.0 and EC-EARTH from the ensemble projection, respectively. The latter inconsistencies are mainly due to the differences of the RegCM4 projection driven by EC-EARTH (figures not shown). For the area-averaged in NWC, NEC, ETP, and SEC, the percentage changes projected by the individual members are in the range of −5.3% to 9.4%, −12.2% to 4.6%, −27.4% to −8.3%, and −47.2% to −22.6% for the total snowfall amount (Table 2); −11.3% to −4.9%, −13.1% to −2.7%, −26.1% to −11.0%, and −45.9% to −26.5% for the number of snow days (Table 3); and 6.8% to 19.3%, 0.6% to 13.2%, −2.8% to 12.1%, and −2.4 to 8.7% for the snowfall intensity (Table 4), respectively. The across-ensemble robustness of the projected changes in the frequency is higher than that in the amount and intensity.

The light snowfall amount shows a declining tendency across the entire country by the end of the twenty-first century (Fig. 10a), which may be mainly caused by the reduction of its frequency (Fig. 11a). The ensemble projected decreases for the light snowfall amounts (days) in NWC, NEC, ETP, and SEC are 11.5% (13.3%), 12.1% (11.9%), 17.9% (18.4%), and 35.7% (36.0%), respectively, all of which are statistically significant at p < 0.05 and consistently projected by the ensemble members (Tables 2 and 3). The declines in the light snowfall amounts yield positive contributions of 90.7%, 34.2%, and 13.0% to the decreases in the total amounts in NEC, ETP, and SEC, respectively, and a negative contribution of 76.9% to the increase of total amount in NWC (Table 2). The reductions of light snowfall events account for 106.5%, 91.1%, 78.2%, and 60.7% of the decreases in the total snow days in NWC, NEC, ETP, and SEC, respectively (Table 3).

Fig. 10.
Fig. 10.

As in Fig. 9, but for the amount of different categories of snowfall events.

Citation: Journal of Climate 31, 15; 10.1175/JCLI-D-17-0428.1

Fig. 11.
Fig. 11.

As in Fig. 9, but for the number of days of different categories of snowfall events.

Citation: Journal of Climate 31, 15; 10.1175/JCLI-D-17-0428.1

Compared with the changes in the light snowfall events, the main discrepancies of the moderate snowfall events are that both their amounts and frequencies are projected to increase in most parts of western China (Figs. 10b and 11b). However, when considering the target subregion NWC as a whole, the amount and the frequency still reduce by 4.2% (not significant at p < 0.05) and 4.6% (p < 0.05), accounting for −25.0% and 7.6% of the total snowfall change, respectively. In NEC, ETP, and SEC, the moderate snowfall amounts (days) are anticipated to significantly decrease by 6.8% (7.1%), 16.4% (16.4%), and 36.8% (36.8%), respectively. Their respective contributions to the total snowfall change are 34.3% (8.0%), 21.8% (12.3%), and 15.0% (14.9%). The individual members show consistent projections in all the subregions except in NEC (Tables 2 and 3).

For the large snowfall events, changes in the amount and frequency resemble each other, with an increase (a decrease) to the north (south) of a southwest–northeast-oriented belt (Figs. 10c and 11c). The increase in NWC (5.4%) and decreases in ETP (−16.1%) and SEC (−37.5%) for the amount and the decreases in ETP (−16.2%) and SEC (−37.6%) for the frequency are significant. Their respective contributions to the local changes in the total snowfall amounts are 40.1%, 22.9%, and 27.8%, and are 6.5% and 14.1% for the total snow days. The ensemble members only agree on the signs of changes in ETP and SEC (Tables 2 and 3).

For the heavy snowfall events, their amounts and frequencies both increase in northern China and decrease in southern China by the end of the twenty-first century (Figs. 10d and 11d). The projected changes of the intense snowfall events are generally consistent with the result of J. Sun et al. (2010), which was based on the ensemble projection of four CMIP3 GCMs under the A1B scenario. Regionally averaged, the significant changes in the heavy snowfall amount (frequency) include an increase of 22.3% (19.6%) in NWC and a decrease of 10.4% (12.1%) and 29.0% (31.7%) in ETP and SEC, respectively. The increase in the heavy snowfall is the most pronounced contributor to the increase of the total snowfall amount in NWC (with a contribution of 160.7%). The decreases in the heavy snowfall in ETP and SEC account for 20.1% and 43.0% of the decreases in the total amounts, respectively. The individual members agree on the signs of changes in the amounts in NWC and SEC and changes in the frequencies in NWC, ETP, and SEC. The projected changes in NEC are not robust in sign (Tables 2 and 3). In addition, an overall intensification of heavy snowfall events is projected by the ensemble. Such changes in sign are robust across the ensemble members (Table 4). Thus, the increases in the heavy snowfall amounts in northern China are the result of increases in both the frequency and intensity, while the decreases in the heavy snowfall amounts over southern China are mainly due to the decreases in the frequency.

To further explore the future behavior of the four categories of snowfall events within the wintertime snowfall, the ensemble projected ratio changes are presented in Fig. 12. A common feature is a decrease in the proportion of the light snowfall and an increase in the proportion of the heavy snowfall. By the end of the twenty-first century, the ensemble projected ratios of the light (heavy) snowfall in NWC, NEC, ETP, and SEC tend to decrease (increase) by 3.5% (4.7%), 2.5% (2.7%), 1.0% (1.6%), and 0.5% (3.0%) in terms of the amounts (Table 5) and by 3.4% (1.3%), 1.5% (0.5%), 0.6% (0.3%), and 0.1% (0.7%) in terms of the frequencies (Table 6), respectively. The individual members have better agreement in the signs of projected changes in northern China than in southern China. These results hint at a higher fraction of intense snowfall in the future as the climate warms.

Fig. 12.
Fig. 12.

As in Fig. 9, but for the ratio of different categories of snowfall events to the total snowfall amount (left-hand panels) and to the total number of snow days (right-hand panels).

Citation: Journal of Climate 31, 15; 10.1175/JCLI-D-17-0428.1

Table 5.

As in Table 2, but for changes in the ratios of the amount of a certain category of snowfall to the total snowfall amount.

Table 5.
Table 6.

As in Table 2, but for changes in the ratios of the frequency of a certain category of snowfall to the total snowfall frequency.

Table 6.

5. Discussion

The observation and projection results generally indicate a decrease in the total snowfall frequency and an increase in the mean snowfall intensity under a warming background. One question is how the climate warming causes such changes in snowfall. Naturally, low temperatures and high moisture are essential for snowfall occurrences (Qin et al. 2012; Ding 2013; Sun and Wang 2013). Thus, the two factors are investigated to explore possible explanation.

Figures 13a and 14a show the scatterplots of the observed and the RegCM4 ensemble projected surface air temperature and snow days over stations, respectively. An inverse relationship between the temperature and the snowfall frequency is evident, with relatively higher surface air temperature corresponding to fewer snow days. From 1961 to 2013, the wintertime surface air temperature has significantly increased across China. The increasing rates in NWC, NEC, ETP, and SEC are 0.39°, 0.35°, 0.33°, and 0.33°C per decade (p < 0.05), respectively. Such increases in the surface air temperature may reduce the number of cold days, hence decreasing local snow days. By the end of the twenty-first century relative to 1986–2005, the wintertime surface air temperature in NWC, NEC, ETP, and SEC is projected by the RegCM4 ensemble to increase by 2.6°, 2.2°, 3.1°, and 2.1° (p < 0.05), respectively, which may result in further decreases of snow days. The scatterplots of the specific humidity and the snowfall intensity (Figs. 13b and 14b) indicate an in-phase relationship. Relatively higher specific humidity generally corresponds to an enhancement of snowfall intensity. In line with the Clausius–Clapeyron relationship, the climate warming in China may lead to an increase in atmospheric moisture (Sun and Ao 2013; Wu et al. 2015a), which favors the increase of snowfall intensity. Changes in the total snowfall amount are the combined results of changes in the snowfall frequency and intensity.

Fig. 13.
Fig. 13.

Scatterplots of the observed (a) surface air temperature (°C; abscissa) vs snow days (days; ordinate) and (b) specific humidity (g kg−1; abscissa) vs snowfall intensity (mm day−1; ordinate) over stations during the wintertime of 1961–2013. Each dot represents the corresponding values for one station in each winter. The snow days are calculated as the total number of the days with snowfall occurring in the wintertime, and the surface air temperature, specific humidity, and snowfall intensity are calculated as their respective averages for the days with snowfall occurring in the wintertime. The red line indicates the best linear fit.

Citation: Journal of Climate 31, 15; 10.1175/JCLI-D-17-0428.1

Fig. 14.
Fig. 14.

As in Fig. 13, but for scatterplots of the ensemble projected (a) surface air temperature (°C; abscissa) vs snow days (days; ordinate) and (b) specific humidity (g kg−1; abscissa) vs snowfall intensity (mm day−1; ordinate) over stations during the wintertime of 2080–99 under RCP4.5 (relative to 1986–2005).

Citation: Journal of Climate 31, 15; 10.1175/JCLI-D-17-0428.1

Certainly, the above physical explanation is just from the perspective of changes in temperature and moisture conditions. In fact, the physical mechanisms for the snowfall changes in different regions are complicated, since snowfall is influenced by multiple factors such as large-scale atmospheric circulations, air–sea interactions, and sea ice (e.g., Ding et al. 2008; Gao 2009; Wen et al. 2009; Wang et al. 2011; N. Liu et al. 2012, Zhang et al. 2012; Wang and He 2013; Zhou et al. 2017). For example, recent studies documented that the interdecadal weakening of the East Asian winter monsoon in the 1980s may induce an increase in the snowfall intensity in NEC, through increasing local air temperature and moisture (Wang and He 2013; Zhou et al. 2017). Projections from CMIP5 models reveal that the East Asian winter monsoon tends to be weakened by the end of the twenty-first century (Jiang and Tian 2013; Wang et al. 2013). This weakening is expected to exert an effect on the snowfall changes. Thus, the detailed mechanisms responsible for the snowfall changes (including changes in different types of snowfall events) in different subregions of China deserve further in-depth analyses.

In addition, the snowfall in China also exhibits intraseasonal variability (J. Sun et al. 2010; Liu et al. 2013; Qin et al. 2012). The change in different types of snowfall events at the intraseasonal time scale is an interesting topic.

6. Summary

In this study, we comprehensively examined the observed changes of the snowfall in China during the past 53 years, including the amounts, frequencies, and intensities of the total and the four types of snowfall events. Their future changes by the end of the twenty-first century under RCP4.5 are also projected based on an ensemble of simulations using RegCM4. The main findings are summarized below:

  1. Climatologically, there are four key regions with a great amount and/or large variability of wintertime snowfall in China: NWC, NEC, ETP, and SEC. The notable amounts in NWC, NEC, and ETP are mainly determined by the frequencies of the local snowfall events, while that in SEC mainly results from the intensities of snowfall events. Since the 1960s, the wintertime snowfall has generally occurred less frequently but with a strengthening intensity over all the subregions. Owing to the increase in the mean intensity, the total snowfall amount has increased in NWC, NEC, and ETP. In contrast, because of the decrease in the snowfall frequency, the total snowfall amount has decreased in SEC.

  2. The snowfall amounts of the four types of events in NWC show consistent increasing trends from 1961 to 2013. The increases in the moderate-to-heavy snowfall amounts, which are the consequence of the increases in the frequencies and intensities of these events, play dominant roles in the augmentation of the total snowfall amount. The increase in the light snowfall amount, mainly caused by the strengthening of the light snowfall intensity, plays a slightly positive role. Similar changes are found in NEC and ETP, except here there is a slightly negative contribution to the total amount from the decrease in the light snowfall amount, which mainly results from the decrease in the light snowfall frequency. In SEC, the consistent decreases in light-to-heavy snowfalls, with the largest change in the light snowfall and the smallest change in the heavy snowfall, play positive roles in the change of the total snowfall. The decreases in snow days in the four subregions are mainly due to the reductions of light snowfall events. Given the greater reductions of the light snowfall events, the ratio of light snowfall to the total snowfall is reduced, suggesting a higher probability of intense snowfall.

  3. Relative to 1986–2005, the number of wintertime snow days is projected to decline and the mean snowfall intensity is projected to strengthen in the four subregions by the end of the twenty-first century under RCP4.5. The projected wintertime snowfall amount tends to increase in NWC but to decrease in the other three subregions. The former change may be due to the increase in the mean intensity, and the latter changes may result from the decreases in the snowfall frequency.

  4. The amounts and the frequencies of the light and moderate snowfalls are projected to decrease in the four subregions; the large snowfall amount and frequency are projected to increase in NWC but decrease in the other three subregions; and the heavy snowfall amount and frequency are projected to increase in NWC and NEC but decrease in ETP and SEC by the end of the twenty-first century under RCP4.5. The projected decreases in the total number of snow days in the four subregions are primarily due to the decreases of light snowfall events. The projected increase in the total amount of snowfall in NWC is attributed to the increases of heavy and large snowfalls. In NEC, the decrease in light snowfall plays a leading role in the decrease of the total snowfall amount. In ETP and SEC, the projected decrease in the total snowfall amount is a consequence of an overall decrease in light-to-heavy snowfalls. In addition, a lower ratio of the light snowfall and a higher ratio of the heavy snowfall are projected in the four subregions.

Acknowledgments

This research was jointly supported by the National Key Research and Development Program of China (2016YFA0600701), the National Natural Science Foundation of China (41675069), the Climate Change Specific Fund of China (CCSF201731), and the National Program for Support of Top-Notch Young Professionals.

REFERENCES

  • Bai, A., P. Zhai, and X. Liu, 2007: Climatology and trends of wet spells in China. Theor. Appl. Climatol., 88, 139148, https://doi.org/10.1007/s00704-006-0235-7.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Chen, H., 2013: Projected change in extreme rainfall events in China by the end of the 21st century using CMIP5 models. Chin. Sci. Bull., 58, 14621472, https://doi.org/10.1007/s11434-012-5612-2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Chen, H., and J. Sun, 2009: How the “best” models project the future precipitation change in China. Adv. Atmos. Sci., 26, 773782, https://doi.org/10.1007/s00376-009-8211-7.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Chen, H., J. Sun, X. Chen, and W. Zhou, 2012: CGCM projections of heavy rainfall events in China. Int. J. Climatol., 32, 441450, https://doi.org/10.1002/joc.2278.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Ding, Y. H., 2013: China’s Climate (in Chinese). Science Press, 557 pp.

  • Ding, Y. H., Z. Y. Wang, Y. F. Song, and J. Zhang, 2008: Causes of the unprecedented freezing disaster in January 2008 and its possible association with the global warming. Acta Meteor. Sin., 66, 808825.

    • Search Google Scholar
    • Export Citation
  • Feng, S., S. Nadarajah, and Q. Hu, 2007: Modeling annual extreme precipitation in China using the generalized extreme value distribution. J. Meteor. Soc. Japan, 85, 599613, https://doi.org/10.2151/jmsj.85.599.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Gao, H., 2009: China’s snow disaster in 2008, who is the principal player? Int. J. Climatol., 29, 21912196, https://doi.org/10.1002/joc.1859.

  • Gao, X., Y. Shi, D. Zhang, J. Wu, F. Giorgi, Z. Ji, and Y. Wang, 2012: Uncertainties in monsoon precipitation projections over China: Results from two high-resolution RCM simulations. Climate Res., 52, 213226, https://doi.org/10.3354/cr01084.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Gao, X., Y. Shi, Z. Han, M. Wang, J. Wu, D. Zhang, Y. Xu, and F. Giorgi, 2017: Performance of RegCM4 over major river basins in China. Adv. Atmos. Sci., 34, 441455, https://doi.org/10.1007/s00376-016-6179-7.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Giorgi, F., C. Jones, and G. R. Asrar, 2009: Addressing climate information needs at the regional level: The CORDEX framework. WMO Bull., 58, 175183.

    • Search Google Scholar
    • Export Citation
  • Giorgi, F., and Coauthors, 2012: RegCM4: Model description and preliminary tests over multiple CORDEX domains. Climate Res., 52, 729, https://doi.org/10.3354/cr01018.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Han, Z., B. Zhou, Y. Xu, J. Wu, and Y. Shi, 2017: Projected changes in haze pollution potential in China: An ensemble of regional climate model simulations. Atmos. Chem. Phys., 17, 10 10910 123, https://doi.org/10.5194/acp-17-10109-2017.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • IPCC, 2013: Climate Change 2013: The Physical Science Basis. Cambridge University Press, 1535 pp.

  • Jiang, D., and Z. Tian, 2013: East Asian monsoon change for the 21st century: Results of CMIP3 and CMIP5 models. Chin. Sci. Bull., 58, 14271435, https://doi.org/10.1007/s11434-012-5533-0.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Jiang, D., Z. Tian, and X. Lang, 2016: Reliability of climate models for China through the IPCC Third to Fifth Assessment Reports. Int. J. Climatol., 36, 11141133, https://doi.org/10.1002/joc.4406.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Liu, B., M. Xu, M. Henderson, and Y. Qi, 2005: Observed trends of precipitation amount, frequency, and intensity in China, 1960–2000. J. Geophys. Res., 110, D08103, https://doi.org/10.1029/2004JD004864.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Liu, N., J. Liu, Z. Zhang, H. Chen, and M. Song, 2012: Is extreme Arctic sea ice anomaly in 2007 a key contributor to severe January 2008 snowstorm in China? Int. J. Climatol., 32, 20812087, https://doi.org/10.1002/joc.2400.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Liu, Y., G. Ren, and H. Yu, 2012: Climatology of snow in China (in Chinese). Dili Kexue, 32, 11761185.

  • Liu, Y., G. Ren, H. Yu, and H. Kang, 2013: Climatic characteristics of intense snowfall in China with its variation (in Chinese). J. Appl. Meteor. Sci., 24, 304313.

    • Search Google Scholar
    • Export Citation
  • Moss, R. H., and Coauthors, 2010: The next generation of scenarios for climate change research and assessment. Nature, 463, 747756, https://doi.org/10.1038/nature08823.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Qin, D. H., W. J. Dong, and Y. Luo, 2012: Climate and Environmental Change in China: The Physical Science Basis (in Chinese). China Meteorological Press, 432 pp.

  • Song, R., X. Gao, Y. Shi, D. Zhang, and X. Zhang, 2008: Simulation of changes in cold events in southern China under global warming (in Chinese). Adv. Climate Change Res., 4, 352356.

    • Search Google Scholar
    • Export Citation
  • Sun, B., and H. Wang, 2013: Water vapor transport paths and accumulation during widespread snowfall events in northeastern China. J. Climate, 26, 45504566, https://doi.org/10.1175/JCLI-D-12-00300.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Sun, J., and J. Ao, 2013: Changes in precipitation and extreme precipitation in a warming environment in China. Chin. Sci. Bull., 58, 13951401, https://doi.org/10.1007/s11434-012-5542-z.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Sun, J., H. Wang, W. Yuan, and H. Chen, 2010: Spatial-temporal features of intense snowfall events in China and their possible change. J. Geophys. Res., 115, D16110, https://doi.org/10.1029/2009JD013541.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Sun, X. Z., Y. Luo, X. Zhang, and Y. X. Gao, 2010: Analysis on snowfall change characteristic of China in recent 46 years (in Chinese). Plateau Meteor., 29, 15941601.

    • Search Google Scholar
    • Export Citation
  • Taylor, K. E., R. J. Stouffer, and G. A. Meehl, 2012: An overview of CMIP5 and the experiment design. Bull. Amer. Meteor. Soc., 93, 485498, https://doi.org/10.1175/BAMS-D-11-00094.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wang, H., and S. He, 2013: The increase of snowfall in northeast China after the mid-1980s. Chin. Sci. Bull., 58, 13501354, https://doi.org/10.1007/s11434-012-5508-1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wang, H., E. Yu, and S. Yang, 2011: An exceptionally heavy snowfall in northeast China: Large-scale circulation anomalies and hindcast of the NCAR WRF Model. Meteor. Atmos. Phys., 113, 1125, https://doi.org/10.1007/s00703-011-0147-7.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wang, H., and Coauthors, 2012: Extreme climate in China: Facts, simulation and projection. Meteor. Z., 21, 279304, https://doi.org/10.1127/0941-2948/2012/0330.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wang, H., S. He, and J. Liu, 2013: Present and future relationship between the East Asian winter monsoon and ENSO: Results of CMIP5. J. Geophys. Res. Oceans, 118, 52225237, https://doi.org/10.1002/jgrc.20332.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wang, Y., B. Zhou, D. Qin, J. Wu, R. Gao, and L. Song, 2017: Changes in mean and extreme temperature and precipitation over the arid region of northwestern China: Observation and projection. Adv. Atmos. Sci., 34, 287305, https://doi.org/10.1007/s00376-016-6160-5.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wang, Z., S. Yang, and B. Zhou, 2017: Preceding features and relationship with possible affecting factors of persistent and extensive icing events in China. Int. J. Climatol., 37, 41054118, https://doi.org/10.1002/joc.5026.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wen, M., S. Yang, A. Kumar, and P. Zhang, 2009: An analysis of the large-scale climate anomalies associated with the snowstorms affecting China in January 2008. Mon. Wea. Rev., 137, 11111131, https://doi.org/10.1175/2008MWR2638.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wu, J., B. Zhou, and Y. Xu, 2015a: Response of precipitation and its extremes over China to warming: CMIP5 simulation and projection. Chin. J. Geophys., 58, 461473, https://doi.org/10.1002/cjg2.20187.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wu, J., X. Gao, Y. Xu, and J. Pan, 2015b: Regional climate change and uncertainty analysis based on four regional climate model simulations over China. Atmos. Oceanic Sci. Lett., 8, 147152, https://doi.org/10.3878/AOSL20150013.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Xu, C.-H., and Y. Xu, 2012: The projection of temperature and precipitation over China under RCP scenarios using a CMIP5 multi-model ensemble. Atmos. Oceanic Sci. Lett., 5, 527533, https://doi.org/10.1080/16742834.2012.11447042.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • You, Q., S. Kang, E. Aguilar, and Y. Yan, 2008: Changes in daily climate extremes in the eastern and central Tibetan Plateau during 1961–2005. J. Geophys. Res., 113, D07101, https://doi.org/10.1029/2007JD009389.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Zhai, P., X. Zhang, H. Wan, and X. Pan, 2005: Trends in total precipitation and frequency of daily precipitation extremes over China. J. Climate, 18, 10961108, https://doi.org/10.1175/JCLI-3318.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Zhang, Q., C.-Y. Xu, Z. Zhang, Y. D. Chen, and C.-L. Liu, 2009: Spatial and temporal variability of precipitation over China, 1951–2005. Theor. Appl. Climatol., 95, 5368, https://doi.org/10.1007/s00704-007-0375-4.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Zhang, R., M. Hong, K. Liu, and Y. Chen, 2012: Subtropical high circulation background and its variation characters in a serious cold rain-snow frost disaster in winter of 2007/2008. Trans. Atmos. Sci., 35, 19.

    • Search Google Scholar
    • Export Citation
  • Zhou, B. T., Q. H. Wen, Y. Xu, L. Song, and X. Zhang, 2014: Projected changes in temperature and precipitation extremes in China by the CMIP5 multimodel ensembles. J. Climate, 27, 65916611, https://doi.org/10.1175/JCLI-D-13-00761.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Zhou, B. T., Y. Xu, J. Wu, S. Dong, and Y. Shi, 2016: Changes in temperature and precipitation extreme indices over China: Analysis of a high-resolution grid dataset. Int. J. Climatol., 36, 10511066, https://doi.org/10.1002/joc.4400.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Zhou, B. T., Z. Wang, and Y. Shi, 2017: Possible role of Hadley circulation strengthening in interdecadal intensification of snowfalls over northeastern China under climate change. J. Geophys. Res. Atmos., 122, 11 63811 650, https://doi.org/10.1002/2017JD027574.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Zhou, B. Z., and Coauthors, 2011: The great 2008 Chinese ice storm: Its socioeconomic–ecological impact and sustainability lessons learned. Bull. Amer. Meteor. Soc., 92, 4760, https://doi.org/10.1175/2010BAMS2857.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
Save