November 1, 2004

Storm Scan

The specter of global climate change has people believing that the weather’s getting worse. But it isn’t (at least in Scandinavia).

With winter on the horizon, what better time than now to examine changes in storminess that might result in our future global greenhouse? Winter storms not only give the media a cheap thrill, but they also have real economic and social impacts.

But will future cyclones become more or less frequent and intense as greenhouse gases increase? That turns out to be a fairly tricky scientific question. Winter storms derive their energy from the jet stream. The projected pattern of temperature change (with the poles warming much more than the tropics) should produce a weaker winter jet. So we might expect weaker storms to be a harbinger of climate change. But research published in 2003 by Oliver Frauenfeld and Robert Davis showed a consistent upper-air response to warming that enhanced the upper-air circulation in some regions. Also, a moister atmosphere tends to have more rising air, which alone could produce more or stronger storms. What is clear is that the answer to the question is not at all clear.

A recent paper in Geophysical Research Letters by Lars Barring and Hans von Storch examining trends and variability in Scandinavian winter storms adds to the debate. The lead author has been analyzing a very rare kind of climatic data set—actual instrumental observations from the early 19th century. The barometer was invented by Torricelli in the mid-17th century, and some routine pressure observations can be found from the 18th century, although they typically are sporadic.

But Lars Barring has been working with two pressure data sets (from both Lund and Stockholm, Sweden) that have unusually high quality and consistent readings extending as far back as the late 1700s. Station pressure readings were taken at least three times daily at Lund since 1780 and at Stockholm from 1820. Barring examined these records for potential biases and inhomogeneities in prior publications and has developed a very long term time series of air pressure at these two sites.

In the current paper, the authors looked at three variables—the annual number of “storms” (station pressure less than 980 millibars), the annual number of observed pressure drops of more than 16 millibars in twelve hours, and extremes in the within-year distribution of 12-hour pressure changes. Each of these variables was then examined over the entire period of record to look for evidence of climate change.

Figure 1 shows the long-term records of each variable. These records are characterized by a remarkable tendency for flatness. The smoothed lines fitted through the data to better present long-term variations do show some minor undulations, but there is little evidence for long-term changes in variability, let alone much of a case to be made for many trends. The authors ran 32 trend analyses (four variables for two stations using two different statistical tests) and they uncovered only five significant trends in the 32 trials.

Scandinavian Storminess

Figure 1. Historical measures of “storminess” derived from the land-term surface pressure observations from two stations in Sweden: Lund (blue) and Stockholm (red). The top panel is the annual number of observations in which the pressure dropped below 980 mb; the next lower panel is the annual number of events in which the pressure dropped by more than 16mb in 12 hours; and the bottom two panels depict two measures of the extremes (the 95th and the 99th percentiles) in pressure differences between two consecutive observing periods (from Barring and Von Storch).

Of course, greenhouse gases have been increasing over this entire period (admittedly only very slowly back in 1780), but the rate of CO2 increase really shot up since the mid-20th century. If you only looked at records since World War II, you could easily be led to believe that storminess was increasing significantly, primarily driven by a rather stormy period in the 1980s. But in the context of this more complete long-term record, there was an even stormier period in the 1860s-1870s, when greenhouse gases were hardly a factor. Also, since 1990, the pressure readings appear to have settled back to values near their long-term average.

In the context of global warming, records from the North Atlantic should always be examined relative to the North Atlantic Oscillation, or the NAO. The NAO is invoked every time some so-called wise-guy climatologist remarks that long-term cooling has occurred over Greenland or northeastern North America. The NAO refers to the pressure difference between the Icelandic Low and the Azores High—when both of these semi-permanent pressure features are strong, the NAO is highly positive, the westerly winds across the Atlantic are strong, Europe has mild winters, and storms tend to track farther south than normal. But most importantly, the NAO is correlated with northern hemisphere winter temperatures. So a stronger-than normal Icelandic Low, and its associated counter-clockwise rotation, should generate stronger than normal northerly (i.e., cold) winds on its western flank—in other words, over Greenland and eastern Canada. Thus, the equation is simple: Global warming = regional cooling.

So do these Scandinavian pressure records support this NAO-based theory? Well, surprisingly, no—the correlations between the NAO and the storminess indices are low and not statistically significant.

In their introduction, Barring and von Storch directly address the disparity between public perceptions and climate reality:

The public and ecosystems in storm-prone areas…are well adjusted to the continuous stream of passing windstorms. However, every now and then extreme windstorms cause severe damage. Together with the perspective of anthropogenic climate change, such extreme events create the perception that the storm climate would change; that the storms lately have become more violent, a trend that may continue into the future. The question is, of course, whether this perception is essentially caused by certain deeply rooted cultural notions about the relationship between man and nature, or whether such changes are real.

The main conclusions related to climate change from this study:

1) “no significant robust long-term trends”;
2) “The conspicuous increase in [the frequency of large 12-hour pressure drops] in Stockholm in the 1980’s is evident but much less pronounced in the other storminess indices for Stockholm”;
3) “The 1860s-70s was a period when the storminess indices showed general higher values of comparable magnitude as during the 1980s-90s. However…it is also clear that the indices have returned to close to their long-term average”;
4) “The time series are remarkably stationary in their mean, with little variations on time scales of more than one or two decades.”

In conclusion, the authors write:

Thus the proxies support the notion of an amplified storminess in the 1980’s, but show no indication of a long-term robust change towards a more vigorous storm climate.

All climate records, particularly records prior to the 20th century, should be viewed with a healthy dose of skepticism. But these Scandinavian pressure histories do seem to question the extent to which we have a consistent story to tell about how global warming, storms, and the NAO are related over the North Atlantic. Until climatologists build an internally consistent picture from the evidence, it is likely that both increases and decreases in the intensity and number of storms will be used as evidence to both support and counter the notion of increasing greenhouse gas impacts.

References:

Barring, L., and H. Von Storch, 2004. Scandinavian storminess since about 1800. Geophysical Research Letters, 31, L20202, doi:10.1029/2004GL020441, 2004.

Frauenfeld, O.W., and R.E. Davis, 2003. Northern Hemisphere circumpolar vortex trends and climate change implications. Journal of Geophysical Research, 108, doi:10.1029/2002JD002958.




No Comments

No comments yet.

RSS feed for comments on this post. TrackBack URI

Sorry, the comment form is closed at this time.

Powered by WordPress