Feb 20 2008
Recent Temperature Increases at Lampasas, TX: A Real Signal?
The Lampasas, Texas surface station has been under discussion lately. In 2000, this station moved near a parking lot. It has been argued that the recent sharp increase in temperature was due to this move. At the time, I had access to data only through 2005, and argued that since this increase was not seen in the raw data, any temperature changes were not the result of a station move, but by an adjustment made to the data. I now have the data from this station updated through 2007. The last two years of the raw data do indeed make it look like a “hockey stick”.
However, a station move should manifest itself as a near step function over a maximum of two years (in the yearly data). This is because unless the move occured on January 1st (or whenever the yearly data chooses its first date), the first year will have data values from both the old and new locations. It is not apparent to me why there would be a sudden increase in temperatures due to a station move after more than 5 years.
Steve McIntyre writes:
Recently Anthony Watts noted that the Lampasas TX station was relocated in 2000 to an extremely poor location and attributed a hockey-sticking of the Lampasas series to this re-location. In a comparison that I made with nearby Blanco TX (which is the sort of comparison that USHCN says that they do), it seemed plausible that the move could have added over 1 deg C.
One thing is certain, if the observed temperature increase is due to the station move, it would not be seen in the raw temperature time-series from nearby stations. I show below the time series of Lampasas, TX and its nearest 13 surface stations. These are supposed to be within 75km of Lampasas, but it seems odd that there would be such a high station density. It’s possible I misplaced a decimal point.
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
Many of the stations surrounding Lampasas, TX show a similar warming since the year 2000. These stations may or may not have moved during that period; I have not looked. However, in order for the station temperatures to track as similar as they do since 2000, it leads one to believe that any changes due to the station move are insignificant compared to other factors.
(No Ratings Yet)
Loading ...Related Posts:
11 Responses to “Recent Temperature Increases at Lampasas, TX: A Real Signal?”
To reduce spam, comments are automatically closed 30 days after the last comment. If you would like to comment on any closed thread, please use the contact form at the top of this page.
Thanks for the graphs! Interesting findings. However, would be VERY helpful to look at the data using a different time scale. Since the question was regarding changes that took place ~ 2000, would be very informative to look at the data over the past 20-30 years. From just eyeballing the graphs above, it looks like there is a major difference between station readings in the period post 2000 for the above stations.
Bruce
Just eyeballing the graphs you so kindly provided…would it be fair to say that Lampasas temp generally tracked or was slight below the others until the move it 2000. Subsequently, Lampasas generally recorded temps slightly above the others, overall trend notwithstanding? Hummm?
Well, they all popped up similarly in the last couple of years. Imagine that.
Here’s a comparison that I did with Blanco TX which was a rural station in the GISS neighbor listing for Lampasas and the first comparandum that I looked at
http://www.climateaudit.org/?p=2731#comment-212113 . There was certainly a noticeable difference here.
What were the stations that you used for your comparison?
OK I see the list of stations in your legends.
The above plots show the yearly averaged maximum temperature anomalies. I’m looking at monthly anomalies (min, max, and mean) right now.
The last time we saw a situation like this, in Detroit Lakes, MN, it resulted in a correction to GISS due to a dataset merge fault that was discovered. I’m not saying thats absolutely the case here, but the scenario looked similar.
We’ll see how it plays out. I could be wrong, or I could have opened a door to discover something else.
Lots of folks are now working hard to solve the issue, and that’s the best thing that could happen.
I agree that station moves often look like step functions, but in some cases, other things that occur near the measurement environment, both natural and man made, can mimic that.
Of course if the USHCN network wasn’t so problematic in the first place, we wouldn’t be having these arguments.
[Reply: Looking at the monthly data, it's clear that the station move is visible in the raw version. I haven't looked at any other versions to see if the corrections worked. Today's post will have some pretty pictures.
Added: I should be a little more careful. It's clear something happened at about the same time as the last station move.]
I zoomed the plots above and appearantly there was a 1°C +/-0,5°C temperature shift due to the move (comparing relative values before and after the move).
Mr Hansens correction is clearly not working in this case.
go look at the daily data from ushcn. i havent, but it might show
something
# Steve Bloom on 21 Feb 2008 at 2:10 pm
“Well, they all popped up similarly in the last couple of years. Imagine that”
Except that Lampasas “popped up” more relative to the other sites after the 2000 move. No one has to imagine that (or were you just being sarcastic?). I think that’s the basis for what appears to be a reasonable line of questions?
The (obvious) point is that based on the above plots it doesn’t look wildly divergent. Another one is that I don’t think NOAA or GISS would disagree that there a fair number of problem records. The upshot is that they don’t have a meaningful effect on the big-picture results.
That said, everyone would love it if we could just have a brand new station network that avoided all of these problems… oh, wait, we’ve been planning one for over ten years and the last stations are just being finished.
[Reply: While the above plots don't show the effect of the move, it's clearly visible in the monthly data when compared to 5 nearly stations.]