Skip to content

Cooling The Past In Nebraska

March 25, 2014

By Paul Homewood

 

multigraph

http://www.ncdc.noaa.gov/cag/

 

 

According to NOAA, July 2012, the hottest month in Nebraska in recent years was still not as hot as either July 1934 or 1936. Comparison with 1934, for instance shows:

 

  Mean
July 1934 81.4F
July 2012 80.0F

 

However, the original State Climatological Report for Nebraska tells us that the mean temperature in July 1934 was 82.5F, so what is going on?

 

Scan

http://www1.ncdc.noaa.gov/pub/orders/IPS/IPS-6A84C3F3-F1B7-4012-88CB-C82793FF966F.pdf

 

 

Of course, it is always possible that the mix of stations has changed, but this is easy to check. There are 34 USHCN stations in Nebraska, which have records for both years, and these are split into eight climatological divisions. [These are labelled 1-9, but for some reason Division 4 does not exist!]. Remember that USHCN stations are regarded as “high quality” ones.

Monthly temperature data for each station is available from the State Climatological Reports,  for 1934 & 2012.

By averaging temperature data for the two years within each division, and then weighting the divisional results on the basis of area, we can overcome any issue concerning mix.

The results of this exercise are shown in the Appendices. But to summarise:

 

  Per State Climatological
Reports
Per NOAA
July 1934 82.2F 81.4F
July 2012 79.8F 80.0F
Difference 2.4F 1.4F

 

So, in relative terms, NOAA’s current version makes 2012 a full degree warmer, essentially by cooling the past. Remember too that, if anything, older temperatures should be adjusted UPWARDS to allow for UHI.

Of course, this does not simply apply to Nebraska. As previous exercises have shown, the same sort of cooling the past has taken place on a nationwide scale. I have challenged NOAA to tell me exactly how their current calculations for 1934 are arrived at, and why they are so different to the original figures. Other than giving me a bit of waffle, they were totally unable to.

There are two possible answers:

1) Their results are heavily skewed by unreliable urban/airport sites, which add a warming bias.

2) They have adjusted past temperatures downwards.

 

Nationally, over the last century, temperatures have risen at a rate of 0.14F/decade, so if the Nebraska “adjustments” are applied across the CONUS, two thirds of this increase would be accounted for by adjustments.

 

multigraph

 

 

APPENDIX A

DIV
1934 2012
1 ALLIANCE 77.4 77.4
1 BRIDGEPORT 76.6 80.8
1 HARRISON 74.8 75.0
1 HAY 75.8 76.7
1 KIMBALL 77.1 75.1
1 LODGEPOLE 78.2 78.8
76.7 77.3
2 PURDUM 81.6 80.1
2 STAPLETON 81.4 78.2
81.5 79.2
3 OAKDALE 81.5 80.2
3 TEKEMAH 83.0 81.4
3 WAKEFIELD 80.4 80.5
81.6 80.7
5 BROKEN BOW 83.7 78.4
5 GOTHENBERG 83.2 79.4
83.5 78.9
6 ASHLAND 86.2 83.5
6 DAVID 83.2 81.3
6 GENOA 84.0 80.8
6 SEWARD 85.1 79.8
6 YORK 86.0 80.2
84.9 81.1
7 CURTIS 82.7 82.5
7 IMPERIAL 83.0 78.9
7 MCCOOK 83.6 84.0
83.1 81.8
8 BEAVER CITY 87.1 82.2
8 FRANKLIN 86.5 81.8
8 HADTINGS 86.2 81.0
8 HOLDREGE 85.8 79.3
8 MINDEN 87.0 81.2
8 RED CLOUD 87.6 82.3
86.7 81.3
9 AUBURN 87.2 82.2
9 CRETE 87.8 82.6
9 FAIRBURY 87.2 81.6
9 GENEVA 86.8 80.9
9 HEBRON 88.6 82.0
9 PAWNEE 88.2 82.0
87.6 81.9

 

APPENDIX B


WEIGHTED TEMPERATURE
DIV SQ MILES 1934 2012
1 14225 14.1 14.2
2 18828 19.9 19.3
3 7645 8.1 8.0
5 7367 8.0 7.5
6 8395 9.2 8.8
7 9476 10.2 10.0
8 4549 5.1 4.8
9 6742 7.7 7.1
TOTAL 77227 82.2 79.8
14 Comments
  1. Don permalink
    March 25, 2014 4:58 pm

    “…. mean temperature in July 1934 was 82.5C”.

    Toasty.

    Though you meant F, not C.

    Cheers!

  2. William Abbott permalink
    March 25, 2014 5:07 pm

    Paul,
    That is so cool. What a lot of work. Your exactly right, extrapolating the Nebraska distortions and applying it nationally and you have causally identified the upward trend. So the NOAA waffles a bit, but are totally unable to answer your question, “why aren’t you using the original temperatures?” “why and how did you adjust them?” They get paid to answer questions like that. I should show my congressman your work and ask him to ask them.

    • March 26, 2014 7:10 pm

      Proper procedure for making adjustments:
      1. Give full data as currently recorded.
      1.a. Identify data that need to be corrected, and explain why.
      2. Show proposed alterations, justifying each.
      3. Give adjusted data set, demonstrating its increased “fitness”.

      None of these are done for the current covert “corrections”. The new stuff just shows up, willy-nilly, make of it what you will.

  3. March 25, 2014 7:29 pm

    Thanks, Paul. Outstanding work.
    Since Anthony Watts et al (2012) we know.
    See http://wattsupwiththat.com/2012/07/29/press-release-2/

  4. mkelly permalink
    March 25, 2014 7:51 pm

    Paul could you do a tutorial on how to get the state log and the NOAA information? The link you have under the historical climatolog gives 404 error. Thanks.

  5. William Abbott permalink
    March 25, 2014 9:19 pm

    PDF link worked for me

  6. J Martin permalink
    March 25, 2014 9:34 pm

    “They have adjusted past temperatures downwards.” They have (missing adjective) adjusted past temperatures downwards.

  7. March 26, 2014 9:08 am

    Not only Nebraska but around the world.
    I have started taking display some series from each country.

    http://knuta.no/Dagens_justering-13972s.html
    (Sorry norwegian but i have planned to translate to english,
    but all the graphs are in english.)

  8. David permalink
    March 26, 2014 11:16 am

    Paul,

    I’m sure you’re aware of the papers that inform NOAA’s USHCN dataset adjustments, and that probably the the most pronounced bias corrected for isn’t station moves or instrument changes but ‘time of observation bias’ (TOB). This occurs where min/max thermometers are used (they were in common use in old US voluntary stations). Once basic max/min thermometers are re-set they immediately start recording again. This can lead to quite large errors. As an example (exaggerated for clarity):

    A volunteer observer reads a max/min thermometer around noon daily. Today the max is 10C and the low is 5C. She records this, calculates a mean of 7.5 C for 26th, then resets the thermometer. But it’s been a cloudy morning, and in the afternoon the sun comes out, raising the daily high temperature to 12C. Since a temperature for the 26th is already recorded and the thermometer reset, this new high is carried over into the 27th.

    Say the 27th is unusually cool, with actual temperatures never rising above 5C before noon. Yet when the observer arrives to take the 27th reading she records a high of 12C. If no allowance or adjustment is made for this, then a false warm bias is inadvertently introduced into the temperature record.

    The old US climatology record was found to suffer considerably from this sort of TOB (though not as exaggerated), since many observations occurred at inappropriate times of the day. This is the basis for much, if not most, of the adjustments we are considering here. So the situation is a lot less simple than it may appear.

    David.

    • March 26, 2014 1:53 pm

      Certainly, TOBS is a factor, but according to NOAA, this would account for only about 0.3F.

      http://www.ncdc.noaa.gov/oa/climate/research/ushcn/ushcn.html

      At many of the sites, that make up the NCDC averages for the CONUS, UHI must be at least this, and probably much more.

      Ronan Connolly has an interesting paper on how NCDC, GISS etx fail to allow adequately for UHI. I’ll be summarising shortly.

      Paul

  9. Andy DC permalink
    March 26, 2014 11:08 pm

    I don’t think anyone in 1934 would be so idiotic to take their daily observation at 12 noon. Maybe midnight or 7 AM, but not in the middle of the day. All quirks should even out in the long run.

Trackbacks

  1. More Games by the AGW Crowd - Defending The Truth Political Forum

Comments are closed.

%d bloggers like this: