November 5, 2007

Greenland Climate: Now vs. Then, Part II. Record Greenland Melt Area?

Filed under: Arctic, Climate History, Polar

Recently the press was more ablaze than California with NASA proclamations that the surface area of Greenland had melted in 2007 at a record-high rate. This is true, if the record only extends back only 20 years or so—which is the case of the NASA dataset. If you could peer back a bit further into the past, say back into the 1950s, it is quite likely that the melt area in Greenland then was about the same as it is now, effectively rendering the 2007 melt area hardly newsworthy. Just another NASA climate-change exaggeration?

The NASA press release announcing all of this is rather cryptically headlined “NASA Finds Greenland Snow Melting Hit Record High in High Places” and then goes on to try to explain:

A new NASA-supported study reports that 2007 marked an overall rise in the melting trend over the entire Greenland ice sheet and, remarkably, melting in high-altitude areas was greater than ever at 150 percent more than average. In fact, the amount of snow that has melted this year over Greenland could cover the surface size of the U.S. more than twice.

A “record high in high places”? High altitude melting “greater than ever”? Amount of snow melt this year “could cover the surface size of the U.S. more than twice”? At what depth?

The NASA press release also included the following graphic (Figure 1).

Figure 1. The trend of the melting index in Greenland between 1988 and 2007 in relation to how many surface sizes of the U.S. match the melting index observed each year. (source: NASA press release)

This has to be one of the more bizarre graphs ever put out by NASA. It shows the annual “melt area index” of Greenland in relation to the size of the United States for each year from 1988 to 2007. The value for this year is a bit more than two times the size of the continental U.S. Now considering that the total area of Greenland is just more than one-quarter the area of the lower 48, you may wonder how an area of more than twice the size continental U.S. melted this year in Greenland. Good question.

The answer lies in what exactly the “melt area index” represents. Contrary to what it might seem—the area of Greenland that experience some surface snowmelt—it is in fact, the sum of the area of Greenland the experienced surface snowmelt across all days of the year that melting occurred. Thus if an area of Greenland equal to 1/365th the area of the U.S. experienced melting every day of the year, this would produce a “melt area index” for that location equal to the size of the entire contiguous United States. This certainly makes for a wacky comparison, although it certainly sounds impressive!

Rather than using a picture of the continental U.S. as a metric in their graph, even though it would have been less sensational, NASA should have simply plotted out the time history of the “melt area index” for Greenland and left it at that (we do this service for you in our Figure 2). That would have shown a general rise since NASA Greenland snow melt records began in 1988 up through 2007, but it would also have shown that all of the rise took place from 1988 to 1997. Since then, the “melt area index” has shown year-to-year variation, but no real overall change.

Figure 2. The same NASA data showing the “melt area index” of Greenland without using the area of the U.S. as a metric.

Then why all the news this year when Figure 2 shows that 6 of the past 10 years have had a larger “melt area index” than 2007? Well, if instead of calculating the “melt area index” for all of Greenland, if you limited your calculation only those regions that lie at elevations above 2,000 meters (in NASA parlance “high places”), voila! you find that the value for 2007 is the highest on record since 1988 (or, as NASA described it, “greater than ever”).

Figure 3. The “melt area index” from altitudes on the Greenland ice sheet that are above 2,000 meters (source: Tedesco, 2007).

But, as we have discussed in earlier installments of our series on Greenland’s climate history, Greenland was at least as warm as it is now, back in the early-to-mid 20th century and the coldest period of the past 100 years was probably the mid-to-late1980s. So, it should come as little surprise that the melt area of Greenland has been increasing as temperatures have warmed since the 1980s.

Inquiring minds might want to know what the melt area was like during the warm period 50 to 60 years ago. Obviously, we didn’t have satellites orbiting around taking measurements from space back then, but there was a good deal of climate research taking place on the ground across Greenland. In 1961, much of this work was summarized in an article by R.W. Gerdel in “A Climatological Study of the Greenland Ice Sheet” included in the Proceedings of the Symposium on the Physical Geography of Greenland of the XIX International Geographical Congress. Included among Gerdel’s discussions of temperature, precipitation, winds, fog, radiation, etc., is a section on “The Occurrence of Melting on the High Ice Sheet.”

Gerdel reports that there is evidence of summer melting occurring at least as high as 1700 meters above sea level on the interior of the ice sheet east of Thule at latitude 76ºN. Air temperature measurements from the Thule airbase (along the northwestern Greenland coast) coupled with those taken from elevations on the ice sheet indicate that the temperature lapse rate (decline in temperature with height) was found to be 0.6ºC/100m. Gerdel used this lapse rate to calculate the elevation on the ice sheet where the air temperature would reach the freezing point, extrapolated from the temperatures taken from coastal stations around Greenland (there were very few temperature measurements from locations on the interior ice sheet itself). From these data, Gerdel produced the following Table, which indicates the calculated average number of days during the summer (June, July, August) that the maximum air temperature on the ice sheet was at or above freezing for various latitudes and elevations (based upon observed sea level temperatures during the period 1946-1956).

Figure 4. Table containing the average number of days during the summer in which the maximum daily temperature was calculated to be at or above freezing for various elevations and latitudes across Greenland (source: Gerdel, 1961).

Gerdel described his technique and findings this way:

Assuming that the above relation [the lapse rate at Thule] was applicable to other sections of the Ice Sheet. The extent of melting in other areas was estimated from meteorological records reported from adjacent coastal stations. The number of days having maximum temperatures above freezing for latitudes of 61N, 67N and 7[6]N are shown in Table 5. The data indicate that, south of 67N latitude, considerable melting may be expected at elevations as high as 2700m.

Compare Gerdel’s calculations (based upon mid-20th century data) to the recently reported record melting announcements by NASA. While the latest NASA press release proclaiming the 2007 record melt area only reports anomalies, their press release from back in the spring of 2007 , touting the big melt in 2006 included a graphic illustrating the actual number of melt days.

As our Figure 1 illustrates, the “melt area index” of 2006 was very close to the 2007 value, a comparison of Gerdel’s number to the 2006 data will provide a fair assessment of how current conditions compare to mid-20th century ones.

Figure 5 shows the number of days with melting observed across Greenland in 2006 as reported by NASA (blue colors) along with the number of days of melting for the locations as calculated by Gerdel in 1961 (black numeric values). Notice that in every case, Gerdel calculated a greater number of days with melting than occurred in 2006 including in the “high places” on the ice sheet in the north, south central, and southern portions of Greenland. In other words, conditions were as favorable, if not more favorable, for snow melt to occur over the Greenland ice sheet (including NASA’s “high places”) in the mid-20th century compared with the present day.

Figure 5. Total number of days with melting in 2006 (colors) and total number of melt days expected per year calculated from temperature data collected in the mid-20th century (black numbers). (source: 2006 data from NASA press release and Gerdel, 1961).

So, all these announcements from NASA seem to be just another case of a record value being only as significant as the record length. The area of ice melt in Greenland in recent years (2007 included) is no doubt the highest in the past 20-30 years (since the end of the cold period there in the mid-1980s), but is probably very typical of the melting which took place during the warm period during the years around 1950 (or the decades prior).

We’ll let it up to each of you to determine how much you want to get worked up about all of this.


Gerdel, R.W., 1961. A climatological study of the Greenland ice sheet. Physical Geography of Greenland, 84-106

Tedesco, M., 2007. A new record for 2007 for melting in Greenland. Eos, Transactions of the American Geophysical Union, 88(39), 383.

No Comments

No comments yet.

RSS feed for comments on this post.

Sorry, the comment form is closed at this time.

Powered by WordPress