June 2009 Update on Global Temperature – UAH
Posted by The Diatribe Guy on June 16, 2009
The UAH anomaly this month was 0.043. The link to the data can be found on the right of the page, under the Temperature Resources section.
The May data point
If you think that an anomaly near zero sounds about average, you’re right. In fact, it’s also around median (slightly below, actually). Go figure… The current anomaly ranks 16th of 31 May anomalies (48th percentile) and 202nd out of 366 overall anomalies (44th percentile).
The anomaly is stated in terms of degrees Celsius, but I will use the convention of 0.01 degrees Celsius, which translates to an anomaly of 4.3. When I say “units” I mean “0.01 degrees Celsius.” Units is more convenient.
The May anomaly was 4.7 units less than the April anomaly. However, due to a sharp negative reading last year, we were up 22.6 units this May over last May.
The current 12-month average is 14.0. Despite the lower anomaly this month, the annual average actually increased due to the fact that a cold anomaly dropped off from a year ago in the 12-month average calculation. The current average is the highest since the period ending April 2008. I’ve shown the plotted running 12-month average below. Note the easily seen single-step upward that looks to occur around 1997-98. I’ve noted that in the past.
For giggles, I looked at a couple other running averages, as well. The 36-month average is 18.85 ending May 2009. Interestingly, this average had trended downward every month since September 2007 until this month. It blipped up a bit this time around from the last month’s average of 18.70. (For those who like to accuse me of cherry-picking data, note that all the averages from 18-month average to 30-month average declined, and then every average from 42-months through 108-months declined from last month). Obviously, if you’ve thought about the math on this, it means that a low anomaly from three years ago dropped off. Going forward, it is probable that this average will continue to decline, barring some major warming.
Finally, I pulled the 120-month running average anomaly, which was 20.26. This is the fifth consecutive increase in the 120-month average. It is likely that we will see this average continue to increase over the next year+, unless we start seeing consistent single-digit or negative anomalies.
One caution on that chart: don’t read it as a trend. Looking at the raw data, or the 12-month running average, it is evident that the smooth increase over time has to do with autocorrelation of the data. As I discussed here, a single step in the data will give the appearance of a gradual trend when a plot is presented in terms of long-term averages, or a best-fit single linear trend. So keep all the charts in the right context.
Consecutive month streaks
We’ve now seen three consecutive anomalies lower than the previous month, but at the same time this is the seventh consecutive anomaly higher than the prior-year anomaly.
Keeping in mind the caveat about context and single steps, we live in a world that loves trend lines. So, here is that information:
The overall trend is 0.10464. This means that the trend line indicates that, each month, the anomaly is, on average, 0.0010464 degrees Celsius higher than the previous month. All those digits make it difficult to realize whether that’s big or small, so I let you know how that translates into the rate of warming per Century. The overall trend indicates 1.26 degrees Celsius warming per Century (heretofore presented in the format 1.26/C). The decadal trend is one-tenth that amount, or 0.126 degrees Celsius per Decade.
Of course, we know by simple observation of the chart, as well as knowledge of past trend lines, that a linear trend is not representative of reality. All it tells us is that the best-fit trend line is what it is. From there, we look at other things, like how slope values are changing, cycles in the data, and so on.
Another interesting trend line I keep track of asks the question “How far back can we go to see a non-warming trend?” This is, by design, an exercise in cherry-picking. But it serves a purpose. Obviously, the further back we can go, the higher the probability that the flat line actually means something. It’s valid to suggest that such periods can occur during warming trends. But it’s also valid to suggest that a long period could indicate an end to current warming. More likely, since we know that the long-term trend from surface measures – over 150+ years – indicates some warming, we can see this as an example of another part of the natural cycle taking place. It would be foolish to take the current non-warming trend and suggest that there will never be warming again. It seems every bit as foolish to be projecting runaway warming over the next Century in the midst of 12 years where no warming has occurred. My own analysis here, here, and here leads me to believe that cooling is on the way. But in each of those analyses, it recognizes cyclicality, and this cooling will once again be replaced by warming. And it has nothing to do with people, carbon dioxide, or plastic bags.
Anyway, here is the current non-warming trend:
Let’s take a quick look at the other 5-year-increment trend lines:
Value: -0.257894 (-3.09/C)
Comment: Trailing down after a slight move up from lows in early 2008.
Expectation: Using current average 12-month anomalies, will hit -0.47 by year end
Value: +0.080864 (+0.97/C)
Comment: Note that the longest negative trend is over 12 years, while the 10-year trend is positive. There is a dip in some front-end anomalies that move the slope about. Trends are negative with starting points to current month from May 1997 through April 1998, then they go positive for starting points of May 1998 through February 2000, at which point they go negative again.
Expectation: Using current average 12-month anomalies, the ten-year trend line will go negative by year-end.
Value: +0.083640 (+1.00/C)
Comment: Lowest value since the period ending December 1997.
Expectation: Using current average 12-month anomalies, expected to reach levels below 0.06 by year end.
Let’s also take a look at how the slope values have changed over time:
Value: +0.134764 (+1.62/C)
Comment: Lowest value since the period ending December 2001.
Expectation: Using current average 12-month anomalies, expected to reach levels below 0.123 by year end.
Here’s how recent slope values have trended:
Value: +0.141607 (+1.70/C)
Comment: Lowest value since the period ending January 2006.
Expectation: Using current average 12-month anomalies, expected to reach levels below 0.125 by year end.
Value: +0.103487 (+1.24/C)
Comment: Lowest value in the data set.
Expectation: Using current average 12-month anomalies, expected to reach a low of 0.1017 at August month end, before increasing to just under 0.103 by year end.
Here’s how the recent slope values have trended, although it’s only a handful of data points: