WunderBlog Archive » Dr. Ricky Rood's Climate Change Blog

Category 6 has moved! See the latest from Dr. Jeff Masters and Bob Henson here.

Attribution: Geographical Extent

By: Dr. Ricky Rood, 2:17 AM GMT on May 22, 2008

Attribution (3)

This is the third in a series on the attribution of climate change; that is, how do we determine to what extent the observed warming is caused by humans? The earlier entries are cataloged at the end. This one is complex, but I’m going to give it a try. If this is getting too technical, write me, let me know, and I will move on to another subject. I’m motivated by some of the comments written in the previous blogs.

Here is a picture used in the last blog from the National Climatic Data Center showing temperatures for March from 1880 to 2008.

Figure 1. Time series of anomalies of average temperature for March, 1880-2008. The anomalies are calculated as a difference from the 1961-1990 average. From the National Climatic Data Center .

A natural question arises about the length of the time period of observations used here as well as the averaging period. One of the papers that I think brings a useful and unique perspective to the determination of the recent warm periods is one by Tim Osborn and Keith Briffa in Science Magazine in 2006. The paper is The Spatial Extent of 20th-Century Warmth in the Context of the Past 1200 Years ( link to paper from Osborn’s web page). There has been a challenge and response to this paper. Links to all are given below, as is a link to the data used in the analysis.

In this paper Osborn and Briffa used 14 different time series that are proxies for temperature observations. The observations are spread across the Northern Hemisphere. The observations span the period from year 800 to 1995. Many will note, immediately, that the very warm period since 1995 (see Figure 1, above) is eliminated from the analysis. This elimination occurs because the number of proxy records since 1995 has decreased. The averaging period chosen is from the year 800-1995. The conclusion of the paper is that the recent warming, in the 20th century is unprecedented in the last approximate 1000 years and of broadest spatial extent.

Here is Figure 2 from the Osborn and Briffa paper.

Figure 2: From Osborn and Briffa in Science Magazine (2006). Times series from 800 – 1995 of the percentage of 14 data records that are either warmer (red) or cooler (blue) than average calculated over the entire length of the data records. Figure and data available at National Climatic Data Center. Light shading represents the data sets that are more than 1 standard deviation from the average. Dark shading is more than 2 standard deviations.

This is a complex figure. What is shown are the fraction of the data records that are above or below the average temperature as calculated between 800 and 1995. At any time, if you add the fraction of data records that are above the average to that below the average, the total comes to one (1); that is, all the data records. So, if 0.5 are above the average, then 0.5 are below the average. One thing to observe in the figure is that span of time from 1935-1950 is the time when most of the data sets are showing warm temperatures. Remember, this analysis does not use the observations after 1995.

The shading. The light red and light blue lines show all of the data sets. The lighter shading shows the number of data sets that are one standard deviation above or below average. The darker shading shows those two standard deviations above or below average. (Here is wikipedia definition of standard deviation.) If something is more than 2 standard deviations away from the average, it is a long way from the average. In the 20th century the number of temperature observations that are more than 2 standard deviations away from the average is the largest – by far. We are living at a time when it is not only warm, but very warm, and it is sustained.

There are other times when it was warm, in the late 800s, around 1000-1200, and around 1400. In none of these times, which includes the Medieval Warm Period is the warming as large as today. (At no time prior to the 20th century are more than 10% of the data records more than 2 standard deviations above average.) There are also sustained cold periods, in the early 900s, 1200-1350, near 1460, 1600, and 1800.

Briefly, around 960, almost all of the data sets showed warm anomalies, and there is no cool shading. In the 20th century there is, again, a sustained warming where almost no stations show a cool departure of more than 1 standard deviation. The light blue shading disappears.

To me, this paper, this figure is compelling. It shows the warm and cool variability in the Earth’s temperature, and it shows that the current warm period is without precedent in the last 1000 years. This establishes the extent and magnitude of the current warm period, but it still does not, by itself, support the attribution of the climate change to natural or human causes. Again, it is necessary, but not sufficient.


Osborn and Briffa, Science, 2006

Data used in Osborn and Briffa, Science, 2006

Technical comment on Osborn and Briffa by G. Burger, Science, 2007

Response to Burger’s comment by Osborn and Briffa, Science, 2007

WU blogs on Attribution of Climate Change to Human Activities:

WU Blog on Models and Attribution

Attribution (1)

Attribution (2)

The views of the author are his/her own and do not necessarily represent the position of The Weather Company or its parent, IBM.