TABLE 1 Snowfall Events* in New Brunswick, N.J. by Decade, 1860-1989

 

Number of Snowfall Events*

Decade

=5" to <10"

10"

=5"

1860s**

18

4

22

1870s

13

5

18

1880s

13

3

16

1890s

12

4

16

1900s

15

6

21

1910s

12

6

18

1920s

14

1

15

1930s

10

2

12

1940s

10

5

15

1950s

7

2

9

1960s

11

7

18

1970s

13

2

15

1980s

19

4

23

Total

167

51

218

* Snowfall event may be longer than I day, but may not be interrupted by a day without snowfall.

** Decade of the 1860s = winters 1859-60—1868-69.

tionship between temperature and snowfall. When examining Northern Hemisphere snow and temperature together, as shown in Figure 3, we have found that both are 12-month running means. The snow, which is from the NOAA

FIGURE 3

Twelve-month running means of snow cover, in millions of km2 and surface air temperature, in degrees Celsius, over Northern Hemisphere lands for January 1973 through August 1990. Temperatures are expressed as departures from the reference period, 1951-1970. Temperature data from P.D. Jones. From Robinson et al. (1991).

hemispheric snow charts, is a weekly product that we have averaged into monthly values. The temperature is Phil Jones's data set. Notice the striking relationship between hemispheric temperature—and that is from pole to equator—and Northern Hemisphere snow cover. We are beginning to examine snow and temperature by region and by season. So we see the same general association at this point, at least in a hand-waving sense, as Dr. Groisman found with the snowfall.

Discussion

RIND: You mentioned the relationship between precipitation over North America and global mean temperature. I wanted to add that the warming patterns for the 1930s and the 1980s were very different, just as you said the precipitation patterns were. The 1930s had very large warming at high latitudes with little effect at low latitudes, like an ocean heat-transport change pattern. The 1980s had very uniform warming as a function of latitude, more like what the models are projecting for CO2 and trace-gas forcing. They may be good examples of two different types of processes.

WALSH: That increase of snowfall in northern Canada is interesting in relation to the borehole measurements mentioned yesterday. If snowfall increases, it tends to insulate the ground during the coldest part of the year, which can complicate interpretation of borehole measurements. Also, perhaps we should take a closer look at how subsurface temperatures relate to snowfall to see what the implications might be of a continued increase in snowfall in permafrost areas.

TRENBERTH: The summer and winter correlations between precipitation and temperature tend to be very different.

GROISMAN: That's true. We simply wanted to create annual precipitation time series we could relate to some of the early data.



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement