Which Was The Hottest Year In Ohio?
By Paul Homewood
According to NCDC, last year was the hottest on record in Ohio, tying with 1998 at 54.1F. The next hottest year was 1921 at 53.8F.
This is all very strange, because, according to the original state climatological reports, the annual temperature in 1921 averaged 54.6F.
http://www1.ncdc.noaa.gov/pub/orders/IPS-EAB33803-E82E-40F6-B969-0A98AB2A40C6.pdf
Of course, over the years, the mix of stations may change, and it could be that there more stations now in cooler locations. There is, however, a very easy way to check this, by looking at the temperature differences at individual stations.
There are 26 USHCN stations in Ohio, and of these there are 14 sites which have temperature data for both 1921 and 2012. These are listed in Appendix A.
Annual temperature data for both years is available from the State Climatological Reports, here and here.
As the list shows, only two stations had a higher temperature last year, (and one of these is only 0.01F higher). Most were substantially lower in 2012.
Do these 14 stations offer a reasonable cross section geographically? The list also shows the climatological division in which each station sits. There are 10 such divisions, and only division 10, (the Southeast) is not represented. Otherwise, the stations are well distributed, with no more than two in any division.
We can actually go one better. Appendix B gives the average temperature for each division, and these are weighted to give a statewide temperature, using the geographical area, as used by NOAA here. As Appendix C shows, the weighted average for 2012 is 0.6F lower than in 1921.
| 1921 | 2012 | Difference | |
| NCDC | 53.8 | 54.1 | 0.3 |
| USHCN | 54.5 | 53.9 | -0.6 |
So instead of last year being 0.3F hotter, the USHCN station records suggest that 2012 was actually 0.6F cooler than 1921. A difference of 0.9F.
Not for the first time, we discover that the past has been cooled.
Points To Note
1) Does the exclusion of Division 10 make a significant difference?
Out of the ten stations currently listed in the 2012, there is only one with a record for 1921, Marietta. This is not USHCN, but this shows 55.7F for 1921, and 55.4F for 2012.
So, the division seems to be following the same trend as the rest of the state.
2) It is interesting that the 1921 weighted average of USHCN stations, 54.5F, is so close to the original state report, which gave 54.6F. This would appear to indicate that the USHCN sites give a reasonable cross section.
3) Although the number of USHCN stations is relatively small, they are regarded as high quality sites. In contrast, the current state figures include city and airport sites, such as Toledo, Akron and Cleveland, which must create an upward bias.
4) The temperature data for both years is raw data. There is no adjustment for UHI, which would suggest that the underlying difference between 1921 and 2012 could have been greater still.
5) There is just one station that bucks the trend, Millport, where last year’s temperature was 2.2F higher. There appears to be no regional factors at play here.
Millport comes under the Northeast Hills Division, yet Warren in the neighbouring Northeast Division, and only 30 miles away, was 2.1F colder last year.
Moreover, there is one other station in Millport’s division with records in 1921. This is the non USHCN station at Dennison, where temperatures were 1.2F higher in 1921.
This would lead me to suspect that Millport is an outlier.
APPENDIX A
| OHIO USHCN ANNUAL MEAN TEMPS | ||||
| 1921 | 2012 | DIFF | DIVISION | |
| FINDLAY | 54.4 | 53.9 | 0.5 | 1 |
| WAUSEON | 52.2 | 52.3 | -0.1 | 1 |
| BUCYRUS | 53.7 | 51.9 | 1.8 | 2 |
| NORWALK | 53.8 | 53.6 | 0.2 | 2 |
| WARREN | 53.3 | 51.2 | 2.1 | 3 |
| GREENVILLE | 54.8 | 53.9 | 0.9 | 4 |
| URBANA | 54.3 | 53.4 | 0.9 | 4 |
| CIRCLEVILLE | 56.0 | 55.7 | 0.3 | 5 |
| MILLERSBERG | 54.0 | 52.4 | 1.6 | 6 |
| WOOSTER | 53.3 | 52.6 | 0.7 | 6 |
| MILLPORT | 51.5 | 53.7 | -2.2 | 7 |
| HILLSBORO | 55.5 | 54.4 | 1.1 | 8 |
| PORTSMOUTH | 58.7 | 56.9 | 1.8 | 9 |
| WAVERLY | 57.3 | 56.0 | 1.3 | 9 |
APPENDIX B
| DIVISIONAL TEMPERATURES | ||||
| 1921 | 2012 | DIFF | DIVISION | |
| FINDLAY | 54.4 | 53.9 | 0.5 | 1 |
| WAUSEON | 52.2 | 52.3 | -0.1 | 1 |
| 53.3 | 53.1 | 0.2 | 1 | |
| BUCYRUS | 53.7 | 51.9 | 1.8 | 2 |
| NORWALK | 53.8 | 53.6 | 0.2 | 2 |
| 53.8 | 52.8 | 1.0 | 2 | |
| WARREN | 53.3 | 51.2 | 2.1 | 3 |
| GREENVILLE | 54.8 | 53.9 | 0.9 | 4 |
| URBANA | 54.3 | 53.4 | 0.9 | 4 |
| 54.6 | 53.7 | 0.9 | 4 | |
| CIRCLEVILLE | 56.0 | 55.7 | 0.3 | 5 |
| MILLERSBERG | 54.0 | 52.4 | 1.6 | 6 |
| WOOSTER | 53.3 | 52.6 | 0.7 | 6 |
| 53.7 | 52.5 | 1.2 | 6 | |
| MILLPORT | 51.5 | 53.7 | -2.2 | 7 |
| HILLSBORO | 55.5 | 54.4 | 1.1 | 8 |
| PORTSMOUTH | 58.7 | 56.9 | 1.8 | 9 |
| WAVERLY | 57.3 | 56.0 | 1.3 | 9 |
| 58.0 | 56.5 | 1.6 | 9 |
APPENDIX C
| DIV TEMPS WEIGHTED BY AREA | |||
| DIVISION | 1921 | 2012 | SQ MILES |
| 1 | 53.3 | 53.1 | 4897 |
| 2 | 53.8 | 52.8 | 3310 |
| 3 | 53.3 | 51.2 | 3784 |
| 4 | 54.6 | 53.7 | 4064 |
| 5 | 56 | 55.7 | 4819 |
| 6 | 53.7 | 52.5 | 2972 |
| 7 | 51.5 | 53.7 | 3302 |
| 8 | 55.5 | 54.4 | 4539 |
| 9 | 58 | 56.5 | 4135 |
| WEIGHTED AV TEMP | 54.5 | 53.9 |
Presumably the previous record was adjusted downwards during homogeneity modifications made to the GHCN-M version 3.0.0 dataset, described here: http://www1.ncdc.noaa.gov/pub/data/ghcn/v3/techreports/Technical%20Report%20NCDC%20No12-02-3.2.0-29Aug12.pdf
Ah! Good ole Daniel Rothenberg!
He was the one who disappeared Iceland’s “Sea Ice Years”!
Is there a specific method that you object to?
I suggest you read about the temperature adjustments in Iceland. Try these three posts.
Then explain why the GHCN adjustments are correct.
Perhaps you might also like to tell Bryant Korzeniewski of GHCN, as he has been unable to justify them to me, despite promising an answer last October.
Amazing what can be achieved with a few ideologically compromised cat’s-paws in the right positions, with special access to the official records.