C. Derksen1, R. Brown2, K. Luojus3
1Climate Research Division, Environment Canada, Toronto, ON, Canada
2Climate Research Division, Environment Canada, Montreal, PQ, Canada
3Arctic Research Centre, Finnish Meteorological Institute, Sodankylä, Finland
November 15, 2013
- Northern Hemisphere spring snow cover extent (SCE) was lower than the historical mean (1967-2013) during 2013, with a new record low May SCE established for Eurasia. North American June SCE was the fourth lowest on record.
- The record-setting loss of Eurasian spring snow cover in May 2013, and the below normal June SCE in North America was driven by rapid snow melt, rather than anomalously low snow accumulation prior to melt onset.
- The rate of loss of June SCE between 1979 and 2013 (-19.9% per decade relative to the 1981-2010 mean) is greater than the loss of September sea ice extent (-13.7% per decade) over the same period.
Snow covers the Arctic land surface for up to 9 months of the year. Unlike other elements of the cryosphere (e.g., sea ice, glaciers) most terrestrial snow cover is seasonal, i.e., it melts and disappears completely each spring and summer. The timing of this melt has important implications for the energy budget through changes to surface albedo, for the water cycle through the release of stored water, and for geochemical cycles by influencing the ground thermal regime and the length of the growing season (Callaghan et al. 2011).
Because Arctic land areas are completely snow covered prior to the melt season, variability in spring snow cover extent (SCE) is controlled largely by surface temperatures (warmer temperatures induce earlier snowmelt onset). Arctic terrestrial snow cover is an important contributor to the cooling effect of the cryosphere, so recent reductions in Arctic spring snow cover have direct effects on the global climate system (Flanner et al. 2011). From a hydrological perspective, snow water equivalent (SWE) prior to melt onset is the key variable, as this represents the available store of freshwater. Variability in SWE is driven by the length of the accumulation season and the cumulative amount of cold season precipitation (modified by surface processes such as wind redistribution and sublimation), until the initiation of melt. Then, the rate of depletion is strongly influenced by air temperature (which is controlled largely by large scale atmospheric circulation and incoming solar radiation), and secondary influences such as incoming solar radiation and cloud feedbacks. Monitoring and understanding the interplay between SCE and SWE is vital to addressing the impacts of variability and change in Arctic terrestrial snow cover.
In spring 2013, Northern Hemisphere spring SCE anomalies computed from the weekly NOAA snow chart Climate Data Record (CDR) for months when snow cover is confined largely to the Arctic showed a continued reduction from the historical mean in May and June (Fig. 45). For Eurasia, a new record low May SCE was established, with June SCE tied for the second lowest since 1967. Across North America, April SCE was well above average (standardized anomaly of +2.0), May SCE near average (standardized anomaly of -0.4), and June SCE well below average (standardized anomaly of -1.8). The SCE changes are consistent with differences in continental air temperature anomalies (Overland et al. 2013), who illustrate positive anomalies over Eurasia and negative anomalies over North America (see Fig. 3c in the essay on Air Temperature). The shift to increasingly negative SCE anomalies as the melt season progresses is consistent with observations over the past decade (Fig. 45) and reflected in the monthly SCE trends, computed for 1967 through 2013 (Table 5. The rate of snow cover loss over Northern Hemisphere land areas in June between 1979 and 2013 is -19.9% per decade (relative to the 1981-2010 mean; updated from Derksen and Brown 2012). Interestingly, this exceeds the rate of September sea ice loss over the same time period (-13.7% per decade, Fig. 46; also see the essay on Sea Ice), which is widely used as evidence of the observed response of the cryosphere to rising Arctic temperatures.
|Snow Cover Extent (km2 x 106 x decade-1)|
The timing of snow cover onset in autumn is influenced by both temperature and precipitation. Snow cover duration (SCD) departures derived from the NOAA daily IMS snow cover product (Helfrich et al. 2007) show earlier than normal snow cover onset over Scandinavia (Fig. 47a), with no notable departures over other Arctic regions (earlier than normal snow onset was observed for a mid-latitude region of North America, and southeastern Eurasia). The negative SCE anomalies for May and June (Fig. 45) are reflected in earlier than normal snow melt across the Canadian tundra and eastern Siberia (Fig. 47b). Snow cover persisted longer than normal across northwestern Europe, which drove the positive Eurasian SCE anomalies for April (Fig. 45a). This region was climatologically snow free by May, so the positive spring SCD departures in this region had no impact on the record setting low SCE across Eurasia in May.
Mean monthly snow depth anomalies from the Canadian Meteorological Centre (CMC) daily gridded global snow depth analysis (Brasnett, 1999) for April, May, and June 2013 are shown in Fig. 48. In April (Fig. 48a), snow depth anomalies were positive over most of sub-Arctic Eurasia (mean anomaly of +16.9% relative to 1999-2010 average) and North America (mean anomaly of +29.4%). This is consistent with the negative winter season Arctic Oscillation (DJF mean of -1.12; weaker Arctic jet favourable to cold air outbreaks) which produced below average winter season temperatures over sub-Arctic Eurasia and North America (see Fig. 3b in the essay on Air Temperature). By May, however, the Eurasian snow depth anomalies were strongly negative (Fig. 48b; mean anomaly of -52.1%), illustrating the rapid response of snow conditions to positive surface temperature anomalies over most of Eurasia (see Fig. 3c in the essay on Air Temperature) concurrent with below normal cloud cover (as estimated by the ERA interim reanalysis; not shown).
The quick transition from above normal to below normal snow depth was also captured by the daily time series of Arctic SWE (land areas north of 60°N) derived from the CMC analysis (Fig. 49). Before melt onset, the total SWE was above the average for the data record (since 1998) over both Eurasia and North America. During a two week period in mid-May, the record high SWE over Eurasia plummeted to well below the dataset average (Fig. 49b). The decline in SWE was less dramatic for North America because regionally extensive positive temperature anomalies (also concurrent with below average cloud cover) did not set in until June (see Fig. 3 in the essay on Air Temperature). As was noted for 2012, this means the record setting loss of Eurasian spring snow cover in May 2013, and the below normal June 2013 SCE in North America, was driven by rapid snow melt, rather than anomalously low cold season snow accumulation.
Brasnett, B., 1999: A global analysis of snow depth for numerical weather prediction. J. Appl. Meteorol., 38, 726-740.
Brown, R., and D. Robinson, 2011: Northern Hemisphere spring snow cover variability and change over 1922-2010 including an assessment of uncertainty. Cryosphere, 5, 219-229.
Callaghan, T., and 20 others, 2011. The changing face of Arctic snow cover: A synthesis of observed and projected changes. Ambio, 40, 17-31.
Derksen, C., and R. Brown, 2012, Spring snow cover extent reductions in the 2008-2012 period exceeding climate model projections. Geophys. Res. Lett., 39, doi:10.1029/2012GL053387.
Helfrich, S., D. McNamara, B. Ramsay, T. Baldwin, and T. Kasheta, 2007, Enhancements to, and forthcoming developments in the Interactive Multisensor Snow and Ice Mapping System (IMS). Hydrolog. Process., 21, 1576-1586.
Flanner, M., K. Shell, M. Barlage, D. Perovich, and M. Tschudi, 2011: Radiative forcing and albedo feedback from the Northern Hemisphere cryosphere between 1979 and 2008. Nature Geosci., 4, 151-155, doi:10.1038/ngeo1062.