If there is someplace I can drive and take pictures and notes for you in the area, let me know, I would be more than happy. I am also curious about Montreal, since those temps should track pretty closely too, and they are under a different regimen of adjustment, I assume.

Yes, the website you linked to was in fact where I found the paper last year. Thanks for reminding me where I found it!!

#29 yorick

I have looked at Plattsburgh, but only superficially. I did notice that it seemed to represent a dampened version of Burlington. Right now I am trying to understand how its temperature was homogenized for UHI effect.

#27 Murray Duffin
YEAR ANN
1837 5.45
1838 6.42
1839 6.98
1840 7.68
1841 6.64
1842 7.59
1843 5.79
1844 6.86
1845 7.52
1846 7.4
1847 6.54
1848 7.11
1849 7.01
1850 7.29
1851 6.76
1852 6.35
1853 7.48
1854 6.9
1855 6.98
1856 6.81
1857 5.46
1858 6.04
1859 5.98
1860 5.92
1861 5.62
1862 5.49
1863 5.53
1864 5.93
1865 7.03
1866 6.19
1867 6.05
1868 4.7
1869 5.82
1870 7.35
1871 6.56
1872 5.63
1873 4.53
1874 6.41
1875 4.68
1876 6.99
1877 7.1
1878 8.27
1879 6.58
1880 7.86

Have you seen this website yet. It contains a depository of station histories that seem to have all been done complementary with “THE HISTORY OF SURFACE WEATHER OBSERVING IN BURLINGTON, VERMONT, 1832-1973″ that you began this discussion with.

http://mrcc.sws.uiuc.edu/FORTS/histories/

http://data.giss.nasa.gov/cgi-bin/gistemp/gistemp_station.py?id=425726170050&data_set=1&num_neighbors=1

Plattsburgh Air Force Base was an old Strategic Air Command base. It ramped up with several thousand new soldiers and camp followers (for lack of a better term) in the ’60s, then gradually declined in population.

Here is a map that shows relative population growth since the 1960, note that Plattsburgh has experienced negative population growth while Burlington’s growth is strongly positive.

http://data.giss.nasa.gov/cgi-bin/gistemp/gistemp_station.py?id=425726170050&data_set=1&num_neighbors=1

You may also note that Plattburgh’s temp history shows little if any warming since 1946, when it began. The Difference between it and Burlington varies over time, with Burlington warming since the 60s, but there is the fact that Burlington was also warmer than Plattsburgh in the 40′ and 50′s. Plattsburgh seemed to catch up right about the time the Air Force base was ramped up in the cold war.

I did a calculation in excel using eyeballed endpoints from the graph and I get the same .2 value (7.5 to 7.7 deg at the end). My next question is what would the resulting trend line be if you readjusted all the points to the corrected trend lines and then did an overall trend calculation?

I would say the Burlington record is important if only because it illuminates difficulties in maintaining an accurate record over a long period of time, that is fairly comparable to other records over such time period.

I would restate the latter question slightly. Is it possible to obtain from this record and similar records evidence of change in global temperature over time to a level of accuracy that justifies an action being requested? That begs the next question, if not, how do we fix it?

I’m not looking for a conspiracy. I am most concerned about the level of certainty that we may derive from the data we have so far, and how do we make it better. In that regard I find the contributions to this blog fascinating. Thank you John Goetz, Anthony Watts and others and of course Steve McIntyre.

However, as to the point at hand, I would like to ask a hypothetical question. If all of the current data tweaking and prodding were to be thrown away, and we could start with just the raw data, how would you proceed? What criteria would you use for tweaking and prodding the data into something useable?

I’d suggest that one would try to identify the stations most likely to have had a constant microclimate in locations of the most interest/geographical spacing and use the daily raw data as a start. Annual means are hardly a good descriptor of climate for most earthly locations. Perhaps 25 for the US.

I’d also note that the USGS seem to have water temperature measurements at a number of its river flow measurement stations. Perhaps if those data are of better provenance than the air temperatures, continental climate change could be better measured.

But the history file tells another story. Notably, the station was located at a power plant from 1925 to 1982, then at a radio station for 10 years before finally being sent to the farm. If the station’s history were properly assessed it might not be deemed “rural” and used to adjust other stations.

Similar for Tekamah, NE where it was at a newspaper building in town, various homes in town and now in a rural backyard uphill and upwind from town.

The surfacestations.org database should likely have a station rating plot by year (monthly would be better) rather than just an “as visited” rating. (Note: the Norwalk, Ohio station was in the middle of a significant move the day it was visited. The instrument head was still wired to the data logger while it was sitting out of the way of the ongoing construction. With the rate of station moves these days, it’s not too unlikely to have the equivalent happen again.)