A couple of other projects have been eating my time this past week, and so I am going to continue the recent Saturday trend of looking at some of the smaller states – and will look this week at Connecticut, which has only four USHCN stations, but in contrast with other adjacent states, does, according to Chiefio’s list, also have a GISS station at Hartford.
Well the four USHCN stations don’t have any problems, and I downloaded those easily, but then I come to the GISS station and a problem that has cropped up before. The initial search for Hartford on the GISS station selector site does not give any locations, but if you put in Groton (one of the four USHCN sites) then look for adjacent sites, it does put up two candidates, Hartford Brainard Field with only a few data points between 1947 and 1972) and Hartford/Brad which has a continuous record from 1947 to date. I was just going to pick that one, but in looking at the list of stations adjacent to Groton there is also New Haven, which is given the designation USA, and has a continuous reading from 1880 to 1970, but nothing thereafter. I am going to use the Hartford/Brad data since it is more complete of the two Hartford stations, but out of curiosity will also include the New Haven data – recognizing that it has been discontinued. The change when the stations drop in and out can be seen quite clearly when one plots the difference between the average USHCN data and that of the GISS stations.
In all cases the GISS station data is higher than that of the USHCN data, even when homogenized, with the average difference over the years being 1.89 degF.
As I mentioned last time, since many of the plots are related to geography it is useful to know that for the state, and continuing to use the Netstate sites , the center of the state is at East Berlin, at 72.71 degW; 41.59N. The state is 110 miles long and 70 miles wide, with an average elevation of 152.4 m. (The four USHCN stations average 108 m elevation, Hartford and New Haven are 25 and 4 m above sea level).
Using the TOBS temperatures, there has been an average of 1.1 degF per century increase in the state temperature:
This is, if you remember from the Rhode Island data, less than the sea surface temperature rise, but that reflects the fact that more of the state is further from the sea.
So turning to the geographical impacts on temperature for the state, again the small number of stations make the following findings somewhat fragile, but for comprehensive coverage, here they are. (And no, since I write as I do this, I haven’t seen the plots when I write that). Given the small number of stations in the state I have included both GISS station averages in with the USHCN for the following plots.
Note that there is a correlation with latitude, consistent with earlier findings, but not much of one with longitude, also consistent.
However, when one uses elevation as a criteria (with the USHCN stations more evenly balanced than in some states):
Again there is a very strong correlation with elevation.
Interestingly this state, perhaps because it is one of the smallest, has more population in larger towns where the stations are, and that there are not a preponderance of small communities. That gives a broader scatter of the data, and a good correlation to the log normal relationship.
Which retains the consistency that is growing across this data review.
And, without further comment,