Digital Diatribes

A presentation of data on climate and other stuff

January 2009 Update on Global Temperature – NCDC

Posted by The Diatribe Guy on January 28, 2009

I haven’t touched the NCDC data in a couple months, so I decided to take another look at that. I look at this particular data set with a fairly skeptical eye, but it’s out there, so let’s take a look…


December Data Point
The December anomaly is 0.4825 (heretofore all anomalies will be stated in 0.01 C, i.e. 48.25). This was 13.68 lower than the November 2008 anomaly, but 5.52 warmer than December 2007.

The December anomaly ranks as the 8th highest December anomaly in 129 observations. It ranks 90th of 1,548 total observations.

No major streaks of consecutive warming or cooling year-over-year anomalies are present at the moment. In the last 10 months, 5 are cooler than the prior year and 5 are warmer than the prior year. However 4 of the last 5 months show a warmer anomaly.

12-month Average Temp
The 2008 12-month average anomaly was 48.69. This was the third consecutive increase in the 12-month average calculation, and the current average is the highest since the period ending January 2008. Since it is pleasing to the eye to look at individual calendar years, here is a little chart of the calendar year anomalies since the year 1998.


The overall since inception slope value is 0.04298, which is a per-month average increase rate in terms of 0.01 degrees Celsius. This equates to 0.5157 degrees average warming per Century. The chart is at the beginning of this post.

A current flat/cooling trend line can be extended back to January 2001, which now means that the NCDC data shows no warming for 8 years. This is substantially less than the results of the satellite data, which show no warming for nearly 12 years. It should be noted that the HadCrut data, which is a surface temperature measurement similar to NCDC/GISS is more consistent with the satellite data in this regard. Therefore, the differential is not as simple as saying it’s surface vs. tropospheric measurements. I suspect it has more to do with their applied adjustments to the data. HadCrut is a black box, so we don’t know what their adjustments are.


60-month trend line: The current slope value is -0.1138. This is the third consecutive increase, and the highest slope (least negative) since May 2008. You can see the longer term trend in how slope values have changed in the chart below:


120-month trend line: +0.1241, which has been up and down over the last few months. The current value is the lowest since June 2008.

180-month trend line: +.1404, shown below:


This is the lowest 180-month slope value since February 2002. It continues a decline since the peak slope value for the period ending March 2007, which is shown here:


240-month trend line: +0.1543 is the current value, and it is the lowest since the period ending February 2002. For a long-term view of how these slopes cycle, I present the long-term running chart of slopes here. I have added red lines at zero-anomaly averages to get an idea of typical cycle lengths.


300-month trend line: At +0.1616, it is the steepest of the trend line measures I evaluate. This value has fluctuated between 0.1600 and 0.1620 since March 2008. The current value represents a warming rate of 1.94 degrees per Century. I point this out not because I think it is sensible to extrapolate this out, but to point out the insanity of predictions of anywhere for 3 to 10 degree warming predictions over the next Century. Anyway, even thought the slope values have fluctuated in a channel recently, there is still a small negative trend if you fit a line to them:


360-month trend line: More of the same. It has been on a fairly long and steady decline (since November 2003), and the current slope value is +0.13626, which is the lowest value since the period ending February 2002.

Projected Anomalies
A few months back I offered my projections of future NCDC anomalies based on a best-fit weighting method. The suddenly high anomaly in October was not predicted. My model did, in fact, predict a jump in November, but since October was so high, the extent of the jump was underestimated in September. The model predicted a drop again in December, which occurred, but again the extent of the anomaly was lower than actual. So, there’s good and bad with the predictions. It seemed that it properly assessed some movements in direction, but the rubber meets the road on magnitude, and it lacked somewhat on that.

Here were the predictions:
October: 45.6
November: 54.9
December: 42.0

Here are the actuals:
October: 62.3
November: 61.9
December: 48.3

Average projected for the quarter = 47.5
Average, actual = 57.5

Undeterred, I once again present projected anomalies based on a re-weighting of my factors:
January 2009: 55.50
February 2009: 53.84
March 2009: 54.60
April 2009: 50.29
May 2009: 53.79
June 2009: 51.37
July 2009: 47.68
August 2009: 45.70
September 2009: 47.81
October 2009: 43.59
November 2009: 37.67
December 2009: 40.90

I haven’t scrutinized my model for a while, but I’ll try to take a fresh look at it one of these days.

The long-term projections are shown in this chart:


About these ads

3 Responses to “January 2009 Update on Global Temperature – NCDC”

  1. Jeff Id said

    I know you have a lot of history in your work but I wonder if you would consider reporting your trends as change per year or decade rather than monthly as most of climatology seems to work on that scale. The transition might be just in the addition of the decadal trend below the typical ls fit.

    I think it would help in the interpretation of the scale of the trend. Either way I know now how much work goes into a blog like yours and have plenty of my own requests for more work from my own readers. I spent four hours last night debugging, graphing and interpreting the frequency response of the lower troposphere (after my son went to bed). Odd hobby I’ve developed, I wonder if there’s a cure for it.

  2. The Diatribe Guy said

    Jeff, I’ll take that under consideration. I admit that showing a slope in terms of monthly rate of change per 0.01 C isn’t quite as comprehensible as rate of change per year/decade/century. Due to the volume of graphs generated from the data, I’d want to figure out a way for that number to appear automatically, and not make it a manual addition.

    I know what you mean about this being an odd hobby. As opposed to Watts, this is neither my field of work nor is it any way related to what generates my income. Truth be told, it’s not even my #1 hobby of interest. I’m struggling betwen figuring out how to properly divvy up my time between this, music, investing interests, exercise… and that’s just the time I have outside of work and family.

    Sometimes I kick myself for even starting this, but I can’t help but believe it serves some value, and it certainly helps sharpen some skills and forces me into awareness of what’s going on around me. Those benefits are what keep me going.

  3. Jeff Id said

    Sharpening the skills and taking the brain out for a walk. That is something I definitely understand.

    The thing which got me about your blog, more than even the data was your writing style. I’ve read nearly every post. Sometimes my eyes blur with the piles of numbers but I’m an engineer so I force myself to try and understand the meaning. If more people could do that, I think we would be in a better political situation. I’m having a hard time accepting our new circumstances. It’s like the government trains/schools our kids to be slow, then enslaves them.

    Sorry for the OT but it’s hard to watch.

Leave a Reply

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

You are commenting using your 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


Get every new post delivered to your Inbox.

%d bloggers like this: