Things That Make You Go Hmmm

It’s been a good long while since I did a post where I myself took a look at some data. The reason being is that I lost some of my biggest data files from GISS (the data for each years anomaly maps starting at 1880 and going through 2009) and have been slowing rebuilding m database. While this was going on I had a thought that I would like to go back and check some things that Dr. Christy from UAH had brought up in a couple of video presentations. Basically what it dealt with is the San Joaquin Valley in California and how land use change and not CO2 is responsible for the rise in temperatures there. First the San Joaquin valley is the southern half of the California Central Valley with the Sacramento Valley being the northern one and it is where Dr. Christy grew up and he knows how irrigation had changed the valley. What he found was that Tmax (or Daytime temp) for both the irrigated Valley area and the still desert area in the foothills of the Sierra Nevada Mts. to the east of the Valley have the same trend, while the Tmin (night time temps) diverge with the valley Tmins rising and the Foothills not. This went against what the CO2 based AGW theory stated should have happened and the Tmin divergence increased each year as more and more irrigation turned more desert in the Valley into Farmland. You can see the video presentations here: 

http://www.youtube.com/watch?v=-WWpH0lmcxA 

http://www.youtube.com/watch?v=UcGgLoPpbBw 

The first link is to a lecture he gave at Auburn University and it lasts 1hr 5 mins, the second link is from a debate between Dr. Christy and Dr. Gavin Schmidt from NASA/GISS at Auburn University and is about 21  36 mins long. 

The paper that has the information, methodology and data sources is Christy et al 2006 and can be read online in PDF format: 

http://www.openmarket.org/wp-content/uploads/2009/08/2006_christynrg_ca.pdf 

Now what I’m going to do here is a lot simpler then what Dr. Christy and his team did. They went through hundreds of paper copies of CO-OP stations in the area listed in the paper for their work. They also broke everything down by seasons for each trend. What I wanted to do was see if there was a quick way to see if some recent changes affected the overall findings of this paper. What I wanted was to use paired valley/sierra stations in the USHCN database and see what if any effect the new method that Dr. Menne at NCDC came up with from his work on a non Qced 43% Surfacestations.org database to make adjustments and to extend it out from 2003 to 2009 if possible. 

First thing I found there is only 2 paired stations in both the Christy 2006 paper and in USHCN: Merced/Yosemite Valley and Hanford/Lemon Cove. You can get the data for yourself at this link: 

http://cdiac.ornl.gov/epubs/ndp/ushcn/ushcn_map_interface.html 

The rest of the stations are not in the USHCN database but as you will see when I get to the graphs these stations are used to infill missing raw data in the stations that are in the database. Example for Hanford there is years where there is no raw data in the raw data field, but those same years have data in the adjusted data field. USHCN uses nearby CO-OP and NWS first order stations to infill missing data as part of their adjustment process and those stations themselves do not have to be in USHCN. 

Then to make things even more fun besides using the “Raw” and USHCN adjusted data from NCDC I am getting the adjusted from GISS as well and lets see how it all shakes out. The reason to add the GISS adjusted to this is that GISS also has seasonal means and if all 3 metrics are in close enough agreement on the annual mean I might be able to do a more in depth look using Seasonal means from GISS. 

So I decided to start with the Hanford/Lemon Cove pair which can be seen in the Google Earth image below to show how close together they are (note notice that irrigation and farmland now seems to have reached out to Lemon Cove) 

 

Now the  first thing that I had to do was import the adjusted and raw temperatures from USHCN into a spreadsheet. These temperatures are in Fahrenheit. From there I made anomalies based on the Baseline of 1951-80. Then I imported in the adjusted Annual means from GISS, however unlike USHCN these temperatures are in Celsius. So I first had to convert those temperatures into Fahrenheit and then compute the anomalies on the same Baseline. 

So lets take a look at the Hanford annual mean temperature anomalies in Figure 1 

Figure 1

 

Now as shown the USHCN and GISS trends are very close at almost .7° to .8° Fahrenheit increase over the last 108 years. The USHCN Raw shows a trend of about .4° Fahrenheit increase over the same time period. So the adjustments, even though they double the trend, is not much of an increase over the Raw and all are showing an increase. 

NOTE: I couldn’t take the data all the way out to 2009 because GISS only has data up till 2008 so the rest of the graphs will only go that far as well. 

Here in Figure 2 we will look at the annual mean temperature anomalies for Lemon Cove 

Figure 2

 

Now in this graph you again see that the USHCN and GISS adjusted trends are very close at almost 1.5° Fahrenheit increase over the last 108 years. However unlike Hanford both adjusted trends are not even close to the USHCN Raw Trend, matter of fact the Raw shows a very slight decrease but is essentially flat over that time period. That much of a change is a big red flag and makes you wonder since Dr. Christy did his research using the Raw data from the paper copies. One thing this does is shoot the idea of using the GISS seasonal adjusted data for an in-depth look. However with that said it opens a new question: How is it that you only get a slight increase via adjustments for the dataset that already shows a very noticeable up trend, but in a dataset that is for all intents and purposes is a flat trend they somehow find a whopping 1.5° worth of adjustments to get a warming trend? 

So lets take a look at the Tmax and Tmin both adjusted and Raw for both locations. 

Lets take a look at Tmax in Figure 3 

Figure 3

 

Now isn’t that interesting. First the Tmax Raw trend for Hanford is -2.5° F but after adjustment the Tmax trend is 1° F. That is a huge shift of not just direction but of 3.5°. Lemon Cove is the same just not as bad. The Raw Trend is -1° F, while the adjusted is .75° F which gives a shift of 1.75° and from a cooling trend to warming one. 

Next it’s on to the Tmin in Figure 4 

Figure 4

 

Now on this graph you see that the adjustments for the Hanford Tmin go from a Raw trend of 3° F to an Adjusted trend of about .45° F. For Lemon Cove you see adjustments for Tmin go from a Raw trend of about 1° F to an Adjusted trend of almost 2.25° F. 

The adjustments for the Valley station (Hanford) and the Sierra station (Lemon Cove) ends up giving you the exact opposite results of what Dr. Christy found in his 2006 paper that used data up to 2003. In the Tmax Raw you see that the Valley station and the Sierra station both have very similar trends and both are not getting warmer just as Dr. Christy concluded in his paper, but the adjustments made by NCDC gives both a warming trend. In the Tmin the adjustments are even more egregious. The Valley station had a very significant warming trend in the Raw just as Dr. Christy found now it has been adjusted to be almost flat, while the Sierra station had a modest warming trend amplified to almost the same level as the Valley station’s Raw Tmin trend. Thus the adjustments and the adjustments only is the reason I have seen so far that cause the findings in Christy et al 2006 to not be only wrong but backwards. In Dr. Christy’s paper he states that the Valley’s Tmin has been warming at a rate much faster then the Sierra Tmin, while both Valley and Sierra’s Tmax showed no warming. Now I realize this is only on pair of stations but there is only one more pair in USHCN to check to see if the same thing happened in that pairs adjustments. That will be the next one I look at and if I find the same thing with that pair it will be time to start looking up paper copies online and see what they say for the other stations (hopefully there is more pairs that reach 2008 in there).

Advertisements

8 responses to “Things That Make You Go Hmmm

  1. E.M.Smith July 31, 2010 at 4:25 pm

    Fascinating. And well done.

    May I use one of your graphs in a “go read this” posting pointing at here?

    Sorry to hear about the database damage. I assume it’s recoverable, given what you said. FWIW, I’m doing a comparison of GHCN V1 vs V2, (V1 was up to 1990) but don’t have the GISS results from circa 1990 or prior. Do you happen to have those? I’d love to be able to do a V1 vs V2 “unadjusted data” compared with V1 vs V2 as GIStemp has changed them. But I don’t have the old GISS. So do you?

  2. boballab July 31, 2010 at 6:12 pm

    @1 You can grab a graph no prob. Anyone is free to grab something as long as they don’t change it and they give credit.

    As to the recovery, I got eveything back except the 1200km and 250km anomalies by year (its the data I used for my GISS infilling posts dealing with Costa Rica and Alert, each spreadsheet is 132 columns wide by over 16,000 rows down. So it’s taking time), so I’m forced to redownload each year’s anonaly after I have GISS make the map for that year. Thankfully GHCN v3 hasn’t come out yet like it was suppose to. I want to do a comparison of GISS using v2 to v3.

    As to your question on having GHCN v1 or the GISS analysis on that: I don’t have that either

  3. Ken McMurtrie July 31, 2010 at 7:42 pm

    Just to say – well done!
    Another seeker of truth is coming up with some scientific use and conclusions of data, instead of perverted science.

  4. Verity Jones August 1, 2010 at 5:58 pm

    I saw this last night but it was late and I was feeling too tired and too stupid to appreciate it. Nice piece of work! It frustrates the hell out of me that the adjustments seem to be about homogenisation rather than capturing and preserving the individual responses of microclimates. So now we not only have to worry about UHI and the shift to use of airports, but changing land use and no doubt more efficient technology (for irrigation etc) therein.

    Thanks for posting the Youtube links to Dr Christy’s presentation, although the links seem to be a bit mixed up – the first link you give seems to be another Christy presentation; the second link is the first one you meant to link; the debate with Gavin Schmit was easy enough to find.

  5. boballab August 1, 2010 at 6:37 pm

    @4
    Actually the links are in the right order. The first one is to a lecture he gave in 2007 at Auburn and it goes more in depth into the information he used in the debate in the second link (the debate is from 2010). At the 13:45 mark of the first video is where he talks about the Central Valley.

  6. Verity Jones August 2, 2010 at 5:28 pm

    Damn my impatience – I skipped though it too much. You’re right of course that is the best bit, which I’ve now watched.

  7. boballab August 2, 2010 at 7:51 pm

    @6
    Yeah sometimes that gets me too, but that lecture of his has much more info in it, the stuff from the debate is to me just the bullet points version of it. I found that first video last November at this site:

    http://www.greenworldtrust.org.uk/Science/Curious.htm

    It’s got more videos and the site operators (this person has actually done reconstruction work that later peer reviewed science confirmed: http://wattsupwiththat.com/2009/09/30/agu-presentation-backs-up-mcintyres-findings-that-there-is-no-hockey-stick-in-yamal/ ) own story of going from being a believer to a skeptic and much more.

  8. Verity Jones August 3, 2010 at 1:56 pm

    Lucy Skywalker’s site is very good. I had it on my blogroll for a while but it must have dropped off when I moved to WordPress. I think she was revamping it at the time or something and there was a broken link.

    I referred to her as The Oracle here: http://diggingintheclay.wordpress.com/2009/11/16/gistemp-reloaded/ and she said she was flattered.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: