Climate Audit

As noted before, IPCC 1995 (SAR) arrived at the following important new conclusion based only on Wigley and Kelly 1990, as stated in the Chapter 3 Executive Summary:

Based on the incomplete observations and paleoclimatic evidence available, it seems unlikely that global mean temperatures have increased by 1 deg C or more in a century at any time during the last 10,000 years.

Then in the chapter summary, the following statement is made.

In at least some regions, 20th century temperatures have been warmer than any other century for some thousands of years.

The importance of these findings for IPCC was that IPCC 1990 had said that as long as it was unable to explain past larger warmings, it was unable to allocate a proportion of modern warming to human versus natural causes. These findings in IPCC 1995 are their way of circumventing the findings of IPCC 1990. However, I can find no support in the underlying reference (Wigley and Kelly 1990) for these conclusions. I’ve posted up a pdf for others to look at.

The running text in IPCC 1995 which supposedly supports the above comments is as follows (see my prior post on IPCC 1995 for a more extended context):

Alpine glacier advance and retreat chronologies (Wigley and Kelly, 1990) suggest that in at least alpine areas, global 20th century temperatures may be warmer than any century since 1000 AD and perhaps as warm as any extended period (of several centuries) in the past 10,0000 years.

It seems unlikely, given the smaller regional changes, that global mean temperatures have varied by 1 degree C or more in a century at any time during the last 10.000 years (Wigley and Kelly, 1990).

This latter sentence using the words “may be…perhaps as warm” was inflated in the summary to the more assertive “have been warmer”. There was some controversy at the time of IPCC SAR about text changes, although this was before my having a specific interest in these matters and I do not know whether the inflation of language shown here has any connection to the earlier controversies.

Here I want to look at the different issue of whether Wigley and Kelly, 1990 authorizes the above two claims, which were among the most positive conclusions of this chapter of IPCC 1995. I can find no support for the statements in the running text and, in fact, some statements seem directly opposed. I have at least temporarily posted up a pdf of Wigley and Kelly 1990 here.

It’s pretty surprising to read what they actually say. For example:

Although they have been far less spectacular than the longer ice-age timescale changes, significant changes in climate have occurred during the Holocene. The evidence for these changes comes from a variety of forms of indirect or proxy evidence.

Fossil pollen data indicate that, during the early Holocene, summer temperatures in many parts of the globe were noticeably warmer than today, by up to 2 deg C (where “‹Å“today’ refers to the early twentieth century, and lake level data suggest that the early Holocene was considerably wetter than today in low latitudes…

Because most available proxy data tend to reflect summer climate conditions (Williams and Wigley 1983), it is not known with certainty whether any noticeable changes in annual-mean temperature occurred on the 1000-year timescale… Model results however together with the zonal annual-mean insolation data which show a net increase north or south of 43 degrees N or S suggest that annual-mean temperatures were greater in the early Holocene, at least in high latitudes (Mitchell et al 1988)

On shorter timescales, it is clear that substantial global-scale annual-mean temperature changes did occur during the Holocene. The main evidence for this comes from glacial advance and retreat chronologies.

Superimposed on this [Milankowitch-scale changes] there have been a number of century-timescale Little Ice Age events, apparently occurring at random. …what is the cause of the series of Little Ice Ages? Grove [1988] has reviewed the possibilities. Here we concentrate on just one of these, solar variability.

They then carry out an analysis of C14 fluctuations, demonstrating that these fluctuations have been material. The figure below shows their view of C14 fluctuations, with a notable anomaly in the recent Little Ice Age event (using this term in the same sense as Wigley and Kelly 1990).

Original caption: Figure 2. Atmospheric C14 anomalies and climate. The two curves show bandpass filtered values of the atmospheric C14 anomaly using (a) 1000- and 100- year filters and (b) 1000- and 200- year filters. In (b) cold intervals (from figure 1 converted to calendar years) are marked by screening or as vertical lines (for the three shortest duration events).

They then argue the following:

We now consider the implications of a C14-climate link and estimate the change in solar irradiance that is likely to be associated with C14-anomaly-producing solar events like the Maunder Minimum. To do this, it is first necessary to estimate the global-mean temperature deviation associated with the major glacial advances documented in figure 1. This can be done (albeit with some uncertainty) by reference to the most recent period of the record. Over the past 100-150 years, the globe has warmed by approximately 0.5 deg C (Jones et al 1986). This interval corresponds to that part of the most recent warming trend in figure 1 that lies above the zero reference line. From this it can be deduced that the glacial advance maxima correspond roughly to an annual global-mean cooling of 0.4 deg C below the reference line, or about 0.6 deg C below the average level of the warm intervals. We therefore assume that the cooling associated with C14 maxima is 0.4-0.6 deg C.

Here is their original figure 1 derived from Rothlisberger 1986.

Figure 1. Alpine glacier advance and retreat chronologies for the NorthernHemisphere, Southern Hemisphere and globe based on Rothlisberger 1986. W and C refer to warm (glaciers less advanced) and cold (glaciers more advanced).

They conclude:

The atmospheric C14 evidence of solar variability on the century timescale during the Holocene is indisputable. Because major C14 anomalies occur throughout the Holocene (17-21 in 9500 years) spanning the whole of available data, this type of solar variability must be considered to be a permanent feature of the Sun’s behaviour. Global-scale climate fluctuations of the Little Ice Age type appear to be associated with these C14 anomalies. This implies significant irradiance variations occur in parallel with the solar fluctuations responsible for the C14 anomalies.

They argue that some estimates of change in solar irradiance in the Maunder Minimum must be under-estimates.

On the assumption that the Little Ice Age events of the Holocene are the result of solar irradiance changes, we have used an appropriate time-dependent climate model to estimate the magnitude of these changes. The results depend on the cooling assumed to occur during the Little Ice Ages which we have taken to be 0.4-0.6 degree C and on the assumed climate sensitivity. For sensitivities in the range 0.33-1.0 deg C per wm-2 (which span the known range of uncertainty for this parameter) the implied solar irradiance changes averaged over the 200-year interval range between 0.55 and 0.22 %.

I don’t see that Wigley and Kelly, 1990 affirm clearly either of the following two claims made in the running text. If anyone sees otherwise, I’d appreciate the feedback.

Alpine glacier advance and retreat chronologies (Wigley and Kelly, 1990) suggest that in at least alpine areas, global 20th century temperatures may be warmer than any century since 1000 AD and perhaps as warm as any extended period (of several centuries) in the past 10,0000 years.

It seems unlikely, given the smaller regional changes, that global mean temperatures have varied by 1 degree C or more in a century at any time during the last 10.000 years (Wigley and Kelly, 1990).

It’s also ironic to see the acceptance by Wigley and Kelly 1990 of the existence of Little Ice Age events. It’s certainly an interesting contortion for IPCC 1995 to discuss the recent Little Ice Age event entirely in the context of Bradley and Jones without at least noting up the opposing view of Wigley and Kelly 1990.