by Peter Schwartzman
Fortunately for us, for more than fifty years, diligent employees and volunteers have been painstakingly measuring meteorological variables in cities and towns across our country. This information, once collected and analyzed, provides us a means to determine what, if any, changes are occurring. While the past is not a definitive predictor of the future, much can still be learned by examining a locations (or regions) climatological trends and comparing them to predictions made by complex, computer-intensive climate models. Such a comparison helps to establish what has occurred and what the future may hold for a location. Both considerations are of great importance for economic, agricultural, and energy-related reasons.
Countries and communities the world over are reflecting on projected climate changes and are responding in novel and discernable ways. Just last month, the Kyoto Protocol became the worlds first active climate change treaty. Its entry into force, which required passage by a sufficient number of countries (who combined had to represent more than 55% of the worlds industrialized emissions), now requires developed nations to cut emissions of greenhouse gases by 5.2% below 1990 levels by 2008-2012. Many of the more than 140 nations that have signed and ratified the Kyoto Protocol have already begun reducing their emissions significantly in a variety of ways, including transitioning to renewable energy forms (e.g., wind, geothermal and solar). While the United States and Australia are two of a limited number of holdouts to the Protocol, many U.S. states and cities are adopting their own protocol-like commitments. With all of these discussions and policies under way, it is important for us to look at our region and determine where we stand on this issue as well. One of the first places to start is with the data.
So, are we experiencing climate change in Illinois? To figure this out, I decided to examine climatological records for ten cities throughout the state (shown in Figure I). The data that I analyzed, which can be obtained at the IL State Climatology Office (www.sws.uiuc.edu/atmos/statecli/), consisted of monthly precipitation and temperature measurements taken since 1950. This data was further broken into seasonal chunks, since, climatologically speaking, seasons are a more revealing unit of analysis than months. Five temperature variables were analyzedñhigh daily, low daily, mean daily, # of days where 90 degrees Fahrenheit was eclipsed, and # of days where sub-32 degree temperatures were observed. Two precipitation measures were analyzedñmonthly precipitation totals and monthly snowfall totals. Once the data was properly categorized (by variable, season and year), it was analyzed using linear regression analysis (LRA). LRA is a common method to discern changes in climatological variables, such as temperature. So, in total, more than 240 individual analyses where conducted.
Before exploring the results, a brief introduction to statistics is necessary for some, especially since so little attention is paid to explaining statistical results in the mainstream media, despite their ubiquitous reference and rhetorical power. A finding is said to be significant if is unlikely to have occurred by chance, where "unlikely" means "less than 5% chance." For the purposes of the following discussion, I will use the term "significant" in this specific way and the word "tendency" to refer to a changes that are directional but do not meet the statistical rigor to be considered significant. And while "significant" results often represent outcomes that require our attention, actually, the importance of a result lies in the rate of change observed (i.e., the slope of the "best-fit" line that fits the data). Thus, it is important to consider the rate of change as well as the statistical significance of any result.
Lastly, there are two ways to look at results, individually or collectively. Looking at them individually means determining if one line (drawn though the data for one variable, location and season) is significant (or important). For instance, an individual analysis answers the following question, "Is it snowing more in Galesburg in the winter?" Looking at results collectively requires seeing if individual results are behaving in a similar way to one another. For instance, a collective analysis allows us to answer regional questions such as, "Is it snowing more in IL during the winter?" It is through the combination of individual and collective analyses across temporal (time) and spatial scales that one can discern if climate change is occurring.
Okay, back to the results of my ten station study. First, lets look at temperature. High daily temperatures (HDTs) are not changing in a uniform direction; as many cooling tendencies are observed as warming ones. (Locally, Galesburgs HDT is significantly cooling in summer and fall; by about 2 degrees since 1950.) Interestingly, while HDTs are going up rapidly in Belleville, they are declining precipitously in Carbondale, which given the close proximity of these two locations suggests something local (rather than regional) is causing these changes. However, the number of days which go above 90ºF is declining in 28 out of 30 analyses (a significant result); yet, very few of these changes are significant on an individual basis. Notably, low daily temperatures (LDTs) are increasing across the board with five stations (including, locally, Galesburg, Peoria, and Quincy) exhibiting significant increases in spring (and two locations expressing significant changes in summer as well). And, not surprisingly, the number of days in which it cooled below 32 degrees is decreasing in nearly all seasons and all locations. More impressively, eight of the ten locations exhibit significant declines in sub-32 degree days in spring and three exhibit significant declines in winter as well. The mean daily temperatures (MDTs) observed mimic the changes found in the LDTs and so are generally increasing in most places and seasons, with significant increases in spring. The significant changes in low (& mean) temperature records are important for two reasons. One, in terms of the magnitude of the change, consider what is happening in Galesburg during spring. It is observing 9 fewer days in spring where the temperatures dips below freezing today than it did in the early 1950s. This obviously allows for longer growing seasons, which seems like a good thing. But, shifts in biological clocks of this magnitude in such a short period of time can cause many problems including the inability for migrating birds to find food in a timely fashion and the ability for agricultural pests to reproduce and disperse more readily. Two, most climate models suggest that increases in low daily temperatures is one of the key responses to be expected. Galesburgs increase in LDTs during spring (shown in figure II) and summer are also sizeable, +2.9ºF and +1.5ºF, respectively, since 1950. Changes of this order of magnitude suggest that global climate change may indeed be taking place in our neck of the woods and, as such, are definitely worth keeping an eye on.