Below is a graph showing the annual mean temperature in Cleveland from 1856 to the present. The data for 1856 to 1870 inclusive are from records taken by Gustavus Hyde, who resided near downtown Cleveland. The records from 1871 to 1940 inclusive are from official Weather Bureau readings taken at a rooftop exposure in downtown Cleveland. The readings from 1941 to the present are from Cleveland Hopkins International Airport (originally Municipal Airport). The data set has not been greatly affected by these discontinuities. There was a 35-year period (1871-1905) of overlap between Hyde’s records and those taken by the U.S. Signal Service Corps and Weather Bureau employees. The Hyde annual mean temperature averaged just 0.1F higher than the Weather Bureau annual mean temperature. The airport averaged about 0.5F cooler than the downtown rooftop station during a 13-year period of overlap from 1941 to 1954. This may seem surprising, given the fact that the airport is several miles from downtown and less influenced by Lake Erie being further inland. However, the difference was minimal on yearly average mean temperature. There were seasonal variations, though. The downtown site was a couple degrees warmer during the winter and fall, while the airport was somewhat warmer during the spring and summer months.

There has been a general upward trend of +.13F per decade since the records began, although the data also exhibit ups and downs. In particular, the graph shows a warmer period during the 30s, 40s, and 50s, which was followed by a sharp cooling into the 60s and 70s. This, in turn, was followed by a warming trend to the present. In fact, 2012 (which is not shown) is on track to become the warmest year on record. It may become the first year to average 54F or higher in Cleveland’s weather history. The site discontinuities do not explain the warming trend, because each move introduced a slight cooling bias to the data set (not a warming bias). The site has also been influenced by urbanization over the entire period of record. During the early years, records were taken from instruments mounted on the rooftop of various tall buildings downtown. Thus, it is unlikely that urban effects are artificially increasing the trend. Moreover, the trend is consistent with nearby rural sites with long period of records, as well as the warming of adjacent Lake Erie.

In the future, I will provide additional information about the past locations of the official weather station. I will also break down the data by month to examine whether the general trend is consistent in all seasons & months, or whether there are certain months that are warming faster than others. I may also break down individually each of the three sites used to create the data set below. The Hyde record began in May 1855 and was continuous until December 1906. The downtown Weather Bureau office station was continuous from January 1871 through December 1954, although it sporadically reported until November 1957 before being permanently discontinued. The airport site has been in continuous use since at least January 1941.

Figure 1. Graph of the annual mean air temperature in Cleveland, Ohio for the period 1856 to 2011. Notable years have been identified for easier reference. There has not been a year with a mean temperature at or below 49F since 1996, nor a year with a mean temperature at or below 48F since 1980.

Edit 8/24: I’ve attached a new graphic below that displays the 10-year running mean. This has been done to allow an easier comparison to the Berkeley Earth Surface Temperature reconstruction for Cleveland, Ohio. The BEST study was led by Dr. Richard Muller, a  physicist at the University of California, Berkeley. Dr. Muller had expressed skepticism of existing temperature records and this led him to assemble a team of scientists and statisticians to create a new temperature record. The BEST reconstruction used data from many more stations than existing records (nearly 8,000 in the U.S. alone), and employed a novel technique to minimize the impact of location and instrument changes. The “scalpel” technique used by the BEST team treated each such change as creating an entirely new station. The trend depicted for a given point location then is not derived from any one station, but from many stations in the general region. Inhomogeneities, such as may result from urban heat island effects or undocumented site or location changes, were eliminated by regression to nearby sites.

40.99 N, 80.95 W Mean Temperature

Figure 2. BEST temperature reconstruction for Cleveland, Ohio. The first recorded temperatures at Cleveland were taken by Gustavus Hyde in May 1855. Records prior to that date are interpolated from data for other locations in the region.

The BEST study found a warming trend of 0.84C per century for the Cleveland, Ohio area since 1860. The official NWS records, with the additional Hyde data, show a warming trend of 0.71C per century since 1856. The slightly smaller warmer trend is likely due to the uncorrected site changes. The downtown station was generally between  0.5 and 1F warmer than the airport site. Additionally, the airport station was originally mounted on a rooftop of the Administrative Building, and later the Terminal. In 1976, the station was moved to the ground. As discussed in this NWS report, rooftop temperatures can be significantly warmer than nearby ground-based records. This site move thus may have introduced an additional cooling bias in the records. As you can see from Figure 3 below, the official records closely resemble those produced by BEST. The BEST reconstruction, however, is slightly cooler during the 1930s-1950s than shown by the official records. Nevertheless, the two records closely mirror in each other in terms of overall temperature changes during the period of overlap (1856 to the present).

Figure 3. Annual mean temperature at Cleveland, Ohio from 1856 to the present. The graph has been updated to display the 10-year running mean to allow for easier comparison to the BEST reconstruction displayed in Figure 2.

Advertisements