April 2012 HadCrut Anomaly

I will not be updating the charts in the wake of the latest anomaly released, but thought I’d throw out the quick note that it came in at 0.482, which breaks the stretch of lower anomalies over the last number of months. This is the highest the value has been since late 2010.

Random Slope Charts – HadCrut

I’ll put these charts up without much comment. I updated them, so I thought I’d post them. Each one represents the historical slope values for the different trend line lenghts (60, 120, …, 360). I just find it interesting to see how the best-fit trend lines adjust over time and compare where we’re at now to historical values.

60-hc-slopes-history

120-hc-slopes-history

180-hc-slopes-history

240-hc-slopes-history

300-hc-slopes-history

360-hc-slopes-history

Temperature Trends for the last 360 months – HadCrut

Here are the latest charts for the HadCrut Global Temperature Anomalies, for the 30 year period ending March 2012.

The Data

What is presented here is the trend over the last 360 months, the slope of which has been in consistent decline over the last eight years. The current slope itself is +.1279. This puts the longer-term trendline of 30 years at a higher slope level than the 20 or 25 year magnitude. If one were to extrapolate this value forward as an expectation of continued warming, it would represent a 1.53 Celsius temperature increase per century. This spurred the question to me as to what starting point in time to current represents the highest current slope. The answer to that is using a starting point of January 1974 (37 years, 3 months), which has a current slope value of 0.13735. Cherry-picking that data point would indicate warming at a rate of 1.65 degrees Celsius per Century. All trend lines shorter than that (starting at least 4 months ago) or longer than than have a smaller slope value.

Chart below:

Trend Line over Last 360 months – HadCrut

The next chart is simply a point of comparison. The highest slope value in recent years occurred with the 30-year period ending December 2003, at the height of global warming hysteria.

Peak 360 Month Trend Line

Finally, I show what has transpired with the slopes in recent years with the 30 year trend line. There has been a consistent decline in the slopes since November 2003 of the best-fitted lines. The decline itself is a very close linear fit though not quite as tight as some of the previous charts. There are periods of deceleration and acceleration in the trend, and even some periods where the slope value increases a bit for a time. Similar to the 25-year chart, in the most recent few months you can see it accelerate in its declining value. There is little indication that this decline will halt in the near term. I should explain how I arrive at that statement. I have a spreadsheet set up to emulate future periods, so by simply typing in prospective anomaly values, I can see what the level of the number needs to be in order to put an end to the declining slope values.

By way of exercise, here are the anomaly values needed to keep the slope the same from the current value to some period in the near future:
1) For the slope to remain level as of April 2012 month end, the next anomaly value has to be 52.
2) For the slope to remain level from April to May, the May anomaly needs to be 51.
3) For the slope to remain level from May to June, the June anomaly needs to be 60.
4) For the slope to remain level from June to July, the July anomaly needs to be 57.

None of these values are impossible. However, given the fact that last 20 anomaly readings are under 50, it would seem unlikely to expect such a series of values suddenly emerge.

Trend in recent 360-month slopes

Temperature Trends for the last 300 months – HadCrut

Here are the latest charts for the HadCrut Global Temperature Anomalies, for the 25 year period ending March 2012.

The Data

What is presented here is the trend over the last 300 months, the slope of which has been in very consistent decline over the last five years. The current slope itself is +.113 (actually a bit less than the 20 year slope). So, at this point, the 25-year trend is still positive

Chart below:

Trend Line over Last 300 months – HadCrut

The next chart is simply a point of comparison. The highest slope value in recent years occurred with the 25-year period ending February 2007.

Peak 300 Month Trend Line

Finally, I show what has transpired with the slopes in recent years with the 25 year trend line. There has been a consistent decline in the slopes of the best-fitted lines. The decline itself is a very close linear fit, though in the most recent periods you can see it accelerate in its declining value. There is little indication that this decline will halt in the near term. If I plug the latest 12-month average anomalies the values continue to decline.

Trend in recent 300-month slopes

Temperature Trends for the last 240 months – HadCrut

Here are the latest charts for the HadCrut Global Temperature Anomalies, for the 20 year period ending March 2012.

The Data

What is presented here is the trend over the last 240 months, the slope of which has been in very consistent decline over the last eight years. The current slope itself is +.1228. Sp, at this point, the 20-year trend is still positive, even though the last 15 year trend line is slightly negative. This is due to the first five years of the period (1992-97) exhibiting lower anomalies than current averages. That said, the current slope is the lowest value for a 20 year trend line fitted since the period ending June 1998.

Chart below:

Trend Line over Last 240 months – HadCrut

The next chart is simply a point of comparison. The highest slope value in recent years occurred with the 20-year period ending June 2004.

Peak 240 Month Trend Line

Finally, I show what has transpired with the slopes over the last number of years with the 20 year trend line. There has been a consistent decline in the slopes of the best-fitted lines. The decline itself is a very close linear fit. There is little indication that this decline will halt in the near term. If I plug the latest 12-month average anomalies into each of the next 6 months, we see this trend line dipping below the 0.10 mark in 6-7 months.

Trend in recent 240-month slopes

Temperature Trends for the last 180 months – HadCrut

Here are the latest charts for the HadCrut Global Temperature Anomalies, for the 15 year period ending March 2012.

The Data

What is presented here is the trend over the last 180 months, the slope of which has been in very consistent decline over the last five years. The current slope itself is -.0106, and it is the second month in which the trend line is negative. The current slope is the lowest value for a 15 year trend line fitted since February 1977 (back when people were concerned about another ice age).

Chart below:

Trend Line over Last 180 months – HadCrut

The next chart is simply a point of comparison. The highest slope value in recent years occurred with the 15-year period ending February 2007.

Peak 180 Month Trend Line

Finally, I show what has transpired with the slopes over the last number of years with the 15 year trend line. There has been a consistent decline in the slopes of the best-fitted lines. The decline itself is a very close linear fit. There is little indication that this decline will halt in the near term. If I plug the latest 12-month average anomalies into each of the next 6 months, we see this trend line dipping below the -0.03 mark.

Trend in recent 180-month slopes

Temperature Trends for the last 120 months – HadCrut

Here are the latest charts for the HadCrut Global Temperature Anomalies, for the 10 year period ending March 2012.

The Data

What is presented here is the trend over the last 120 months, the slope of which has been in very consistent decline over the last decade. The current slope itself is -.083722, which represents a cooling rate of 0.1 degrees per decade. The negative slope of -0.086789 that occurred as of the ten year period ending February 2012 was the steepest negative trend line fitted since May 1969, if my spreadsheet and eyeballs are working correctly!

Chart below:

Trend Line over Last 120 months – HadCrut

The next chart is simply a point of comparison. The highest slope value in recent years occurred with the 10-year period ending June 2002. The ten year slope does wobble, but not as much as the five year slope does.

Peak 120 Month Trend Line

Finally, I just wanted to graphically show what has transpired with the slopes over the last number of years with the ten year trend line. There has been a consistent decline in the best-fitted 10-year trend lines. The decline itself is a very close linear fit. There is little indication that this decline will halt in the near term. If I plug the latest 12-month average anomalies into each of the next 6 months, we see this trend line dipping below the -0.1 mark.

Trend in recent 120-month slopes

Temperature Trends for the last 60 months – HadCrut

While no conclusions regarding long-term trends can or should be suggested based on short-term trends, we can at the very least look at them to get insight on the current direction of temperature. It is a footnote of interest, but I’ll present it just the same.

The Data
Most pertinent points I wanted to make on the data were presented in the previous post, so please refer to that for the details.

What is presented here is the trend over the last 60 months, which had been an upward slope that was brought back down thanks to the cooler anomalies (on a year over year basis) of the last 16 months. For all practical and measurable purposes, it shows pure flatness over the five year period, meaning no significant indicated trend in the temperatures over that period one way or the other.

Chart below:

Trend Line over Last 60 months – HadCrut

The next chart is simply a point of comparison. The highest slope value in recent years occurred with the 5-year period ending April 2004. The five year slope does wobble a lot thansk to it’s shortness. Every data point has higher influence than in longer curves, so to see the change from 2004 is not necessarily a surprise, but it is worth noting that we have not seen a trend line reach those levels in 8 years.

Peak 60 Month Trend Line

Finally, I just wanted to graphically show what has transpired with the slopes over the last number of months in response to the lower anomalies. A longer view chart would have gone back to the 2004 peak to show the continual decline in slope measurements.

Trend in recent 60-month slopes

Overall HadCrut trends as of March month-end 2012

Time for an update on the HadCrut data set.

I plan on submitting separate posts over the next few days taking a look at the different time horizons on the charts.

Today’s submission is just the overall total trend, and the current longest period length of time backwards from current that demonstrates a flat or cooling trend in the data.

The Data
As can be expected, the overall chart from inception does not change much over the course of a year, but it is presented for completeness. However, I’ll speak to a few of the numbers as they’ve transpired since my last series of posts last year.

From December 2010 through December 2011 we had 13 consecutive months where the anomaly was lower than the previous year. As such, the trend lines have predictably declined in magnitude over that period of time. The January anomaly of 21.7 was slightly above the 2011 January value of 19.4, otherwise we’d be at a stretch of 16 consecutive cooler months on a year over year basis.

From inception (January 1850) through March 2012 the slope values are as follows:
Per month (degrees Celsius): +0.0003757
Per year: +0.0045085
Per Decade: +0.0450854
Per Century: +0.4508542

As demonstrated in the past, the temperature has not increased in anything resembling a straight line, but instead has seemed to follow a cyclical pattern that I have fitted relatively closely to a dual sine wave. However, the wave fits along the increasing line, so there does seem to be an upward trend in temperature that follows the straight line fairly well over that period. But it is less than a half-degree per Century rate.

The current anomaly of 30.50 is the lowest March anomaly since 1999.

Chart below:

Overall Trend in Global Temps – HadCrut

The other chart I’ve updated and am presenting here is the latest period of flat or cooling temperature anomalies. This is just using the latest month as an endpoint and seeing how far back we can draw a line where we can demonstrate straight from the data that temperature has not increased during that period. By implication, we must understand that drawing the line back to any prior point will show that there is a positive slope to average temperature. So, this is not attempting to demonstrate that temperatures have not increased at all over time, just that there has not been warming for some period of time. The period we can go back to under this measure is January 1997. In other words, we can draw a line back more than fifteen years such that the best fit trend line shows no warming in the global temperature anomalies.

Looking at the chart, At the time of my update last year there had been an increasing trend in the anomalies, elevating the trend line a bit. Since then, it’s been continually lower, and you can see the impact of the lower temps on the graph. Chart below:

Current Flat Trend Period – HadCrut

I’ll be updating the 5, 10, 15, 20, 25, and 30 year trend lines over the next couple weeks. I am also planning on doing a 60-year line, since that seems to be about the period of time that a full cycle takes place.

360-Month HadCrut trends as of April month-end 2011

Today I present the 360-month charts using the most recent HadCrut data.

The 360-month trend line is positive, with a slope of 0.1356, which translates to a temperature change of approximately 0.49 degrees Celsius over the course of 30 years.

Now that I’ve looked at all the different charts and see warming of 0.49 degrees over a 30 year period, here’s how it appears to break down by period:
May 1981 – Apr 1986: +0.10 degrees Celsius contribution to increase
May 1986 – Feb 1991: +0.05 degrees Celsius contribution to increase
May 1991 – Apr 1996: +0.24 degrees Celsius contribution to increase
May 1996 – Apr 2001: +0.17 degrees Celsius contribution to increase
May 2001 – Feb 2007: -0.18 degrees Celsius contribution to increase
Mar 2007 – Apr 2011: +0.11 degrees Celsius contribution to increase*
* – Due to an original error in the presented 60-month chart only being 50 months, the dates do not represent an equal 5-year span. The actual 5 year span was near zero. The previous 5 year period would adjust accordingly.

I’ve discussed the “step function” that seems to have taken place in 1996-97. While not presented here, a much better fit to describe the “trend” in global temperature, simply by observation, is to fit a line from 1981 to 1997-ish and then from that point forward. The best fits of those independent lines would show a very notable jump occurring at that point.

Chart below:

360-month Trend in Global Temps - HadCrut

SLOPE CHANGES
In the chart below, we see the latest series of 30-year slope values since 2003. There was a steady decline in the slope value into 2009 that met the .1300 line, but has since increased to the current levels. The trend line is responding to a “fulcrum” of sorts. What I mean by that is that, even though lower anomalies are dropping off the left side of the chart (which normally would imply a lowering of the trend line) this is outweighed by anomalies on the right side of the chart staying up around the last decade’s average, and this is because of that step function that is now reaching the mid-point of the chart. If future anomalies stay around where they are at now, then when that fulcrum point moves to the left side of the chart, we will see that trend line slope value decline. We wouold not expect to see the 30 year trend line go negative unless there is a prolonged stretch of declining anomalies. Even in that scenario, it will take years to move that line around.

Chart below:

Trend of 360-month Slope values in Global Temps - HadCrut

The entire history of 360-month slopes is shown in the chart below:

There has not been a negative trend line in the 30-year charts since the period ending November 1972. We won’t reasonably be expecting a negative trend line for another few years, even under a cooling scenario. What we can probably expect to see, however, is the slope in the trend line start to decrease soon. The slope will probably be fairly steady over the next 3-5 months, and then we’ll see some lower values emerge. In my estimation (assuming average 12-month anomaly values) we’ll see the slope on the trend line fall below 0.13 around march 2012, thus continuing the downtrend and slope values we’ve seen since 2003. We haven’t seen a slope under 0.13 in the 30-year line since April 1999.

Chart below:

History of 360-month Slope values in Global Temps - HadCrut