June 2008 Update on Global Temperature – GISS
Posted by The Diatribe Guy on June 9, 2008
It’s always interesting to see the GISS global temperature anomalies, especially since the satellite temperature measures typically come out first. Both of those had negative anomalies. Now, understanding that GISS uses a different baseline, it is expected that the anomalies should come in higher. At some point I plan to do an analysis of the GISS versus the satellite data similar to what I did versus the NCDC data last month. Anecdotally, it seems like GISS is deviating more and more from the satellite information. There may be more to the sotry once we get into the data, but at a glance, every month seems to turn the NASA skeptics nothing but more skeptical as to whether or not the data is purposely being estimated and smoothed in such a way as to maximize the anomaly. Add to this the continual adjustments that are made that have been shown to lower the early period anomalies and raise the latter period anomalies (thus increasing the slope of the trend line) and red flags fly up all over the place. You would expect adjustments to act randomly enough so as to not continually shift the trend line in an upward direction, but it has been shown that this is NOT the case.
Anyway, all that is food for thought, but like a good soldier, I will present the numbers as they are given.
The May anomaly is 36 (0.36 degrees Celsius above the 1950-1980 baseline average temperature). While this is the coolest May since 2000, according to GISS, it is the 11th warmest on record, since 1880. This is a huge departure from the satellite data, which had May as the 4th and 7th coldest Mays in their 30 year histories. In other words, the satellites tell us there are more than twice as many warmer Mays in a mere 30 years than GISS tells us occurred in nearly 130 years. Something just doesn’t seem to add up there, but that is what the numbers say.
The latest 12-month average anomaly is 44.3. This continues to fall and is the lowest value since the 12-months ending October 2001.
Despite the variance from the satellite readings, there is one thing that is consistent: this is the 9th consecutive year-over-year decline in monthly anomaly. The last run of this length was the nine months ending July 2003. So, this isn’t necessarily a super-unusual occurrence, but the magnitude of the decline seems to be a little sharper this time around from the 2003 occurrence.
Although the overall trend line since inception doesn’t change in any noticeable way from month to month, I still like to re-present it to bring us back to reality. This is the trend since 1880, which has a slope of 0.0468. This is the equivalent of 0.5615 degrees Celsius warming per century. You can simply observe swings above and below this line over time, which should indicate that a swing below isn’t out of the realm of possibility, nor would it be unusual.
Also, I like to go back to see how long the latest cooling (or flat) period is extending. How long has global warming taken a hiatus? This month, we added another data point, and the trend line actually secured another data point on the front end, so our cooling period now extends back to February 2001. The current period of cooling is now 7 years and 4 months (88 months). This also isn’t unheard of during the last 30 years of increasing temperatures, so we can’t make any bold statements about it. However, it goes without saying that the longer this trend continues, the more it casts doubt on the severity of the global warming we’ve heard so much about. Here is the most recent cooling line, dating back to February 2001. (Yes, this is cherry picking… that’s actually the whole point)
I could post myriad graphs on the different trend lines I analyze, but I’ll just select a couple that I didn’t pick last time and just discuss the rest. Just to change things up, we’ll look at the last 15 and 20 year trend lines here on the charts.
But let’s start with the rolling 60-month trend. The current slope value of the last 60 anomalies = -0.0969. To put it on relative terms, that would be a 1.16 degree Celsius decline per Century. It is the lowest such 60-month slope since the period ending October 2001 (-0.1039). Based on my predicted anomalies, next month the slope would reach a level not seen since the period ending 1995. The predicted slope 12 months from now is -0.1806.
The 120-month slope is 0.1340. While this is down from the last couple months, it is worth pointing out that it is higher than the value four months ago (0.1274). Taking a look at the reasons for this, given the fact that anomalies have actually cooled over previous year, it is because the anomalies 10 years ago were fairly high, so it elevated the front of the line, lowering the slope. A couple of those old values dropped off, and the slope line increased to compensate. Based on predicted anomalies, I actually expect to see this trend line increase over the next few months, and then start a fairly rapid downward trend, as the 1998 values, which were higher, drop off. The predicted slope a year from now is at about current levels, 0.1316.
I’ll discuss the 180-month trend line movement and provide three charts as follows: (1) the current slope level of 0.1904 is presented (equivalent to 2.28 degrees warming per century). While this is a highly positive trend line, it has been trending down and is at its lowest value since the period ending March 2005. (2) The recent peak value for the 180-month trend was 0.2603 (3.12 degrees per century). (3) The change in the slope values have trended down fairly consistently from the peak value to the current value.
The 240-month trend line hasn’t seen a whole lot of movement, and actually isn’t expected to based on predicted anomalies. The slope is 0.1614 (1.94 degrees per century), which is down from the last couple months but up from 3 months ago. Similar to the 120-month trends, this line is being impacted by the values on the front end. Going forward, not much will likely change over the next year. The predicted slope in 12 months is 0.1556. The same three graphs are presented for the 240-month trend lines.
The 300-month slope is currently 0.1578. This trend line is expected to increase because the front end is going to come upon some low anomalies from 1984-85, coupled with some higher anomalies predicted near the end of this year in this particular model. The slope in 12 months is expected to be 0.1680.
The 360-month slope is currently 0.1397. Due to the length of this trend line, it changes pretty slowly over time. It is currently at the lowest value since the period ending December 2006, and is expected to decrease gradually over the next year to a value of 0.1338.
I will be outlining my updated predicted anomalies in a separate post, but those predicted anomalies are what drive the anticipated slope changes. This last month my predictions were all high, meaning the current temperature went below all the estimates based on historical predictive weights. Thus, future anomalies will be projecting lower in all cases.
The final chart provided is a graph that shows how the 60-month slope value changes over time. You can see that this is a periodic cycle, and the current series of trend lines seems to mirror, more or less, the cycles leading up to the period of the 1940s earlier in the 20th century, where there were a few cycles that barely dipped into negative trends before reversing.
I will also be presenting NCDC analysis, hopefully. I hope to get to the HadCrut data next. But you should know that we’re expecting a little one to arrive any time now, so I’m afraid that will take precedence. I know… I know… priorities! I’ll let everyone know when the little guy arrives. I’ll also be getting out my new predicted anomalies when I can. I’ve also got another segment of the Landscheidt paper on the way (which is basically a recap of his referenced Eddy paper[s]). The excitement never ends.