IPCC 20th Century Simulations Get a Boost from Outdated Solar Forcings

Guest post by Bob Tisdale

IPCC 20th Century Simulations Get a Boost from Outdated Solar Forcings

Or The Sun Also Can’t Explain the Warming in the Early Part of the 20th Century

INTRODUCTION

In two previous posts, AGW Proponents Are Two-Faced When It Comes To Solar Irradiance As A Climate Forcing and Climate Modelers Reproduce Early 20th Century Warming With The Help Of Outdated Solar Forcings, I illustrated the basic errors that arise when GCMs use outdated TSI reconstructions while simulating 20th Century surface air temperatures. The problem results because the obsolete TSI reconstructions assumed that solar cycle minimums varied significantly, but the current understanding is that solar cycle minimums are, in fact, relatively flat. That is, minimum TSI level during the Dalton Minimum is no lower than the minimum TSI levels during late part of the 20th Century. This can be seen in the comparison chart available from Leif Svalgaard of Stanford University, Figure 1. The current understanding of TSI variability is identified as Svalgaard. Note in Figure 1 that the Preminger TSI dataset also does not have the large variation in solar cycle minimums.

This is discussed in Preminger and Walton (2005) “A New Model of Total Solar Irradiance Based on Sunspot Areas”.

http://www.agu.org/pubs/crossref/2005/2005GL022839.shtml

The other datasets with large variations in solar cycle minima are no longer considered valid.

http://s5.tinypic.com/mmuclk.jpg

Figure 1

And there are many more climate studies that use the erroneous TSI datasets, including those employed by the IPCC.

THE IPCC USED “QUESTIONABLE” TSI DATA FOR ITS 20th CENTURY CLIMATE SIMULATIONS IN AR4

In Chapter 2, “Changes in Atmospheric Constituents and in Radiative Forcing”, page 190 (page 62 of 106 of the following pdf file) of the IPCC’s AR4, the IPCC first describes the three assumptions or motivations for the existence of long-term variability in TSI, and in the next paragraph, they state, “Each of the above three assumptions for the existence of a significant long-term irradiance component is now questionable.” Refer to:

http://ipcc-wg1.ucar.edu/wg1/Report/AR4WG1_Print_Ch02.pdf

Then in their Supplementary Materials to Chapter 9, “Understanding and Attributing Climate Change”, the IPCC identifies the TSI reconstructions used by the modelers in their table “S9.1. Models used in chapter 9 to evaluate simulations of 20th century climate change with both anthropogenic and natural forcings and with natural forcings only”.

http://www.ipcc.ch/pdf/assessment-report/ar4/wg1/ar4-wg1-chapter9-supp-material.pdf

The IPCC Table S9.1 is shown as Figure 2. And what TSI reconstructions does the IPCC list for the 20th Century Climate Simulations? The ones they consider “questionable”, of course.

http://s5.tinypic.com/aouzpi.jpg

Figure 2

The key to the solar forcings follows.

Even GISS acknowledges the problems with the use of the Lean et al data in the Hansen et al (2007) paper “Climate simulations for 1880-2003 with GISS modelE”. They state, “Lean et al. (2002) call into question the long-term solar irradiance changes, such as those of Lean (2000), which have been used in many climate model studies including our present simulations. The basis for questioning the previously inferred long-term changes is the realization that secular increases in cosmogenic and geomagnetic proxies of solar activity do not necessarily imply equivalent secular trends of solar irradiance.” Following that, GISS goes on to explain the reasons for their continued use of the erroneous TSI data set, “The fact that proxies of solar activity do not necessarily imply long-term irradiance change does not mean that long-term solar irradiance change did not occur.” Refer to:

http://pubs.giss.nasa.gov/docs/2007/2007_Hansen_etal_3.pdf

(Note: The Hansen et al file is 24MB.)

IPCC KEY TO SOLAR FORCINGS AND REFERENCES

The following is the IPCC’s Key to the Solar forcings and references from page SM.9-12 of the Supplement to Chapter 9 of AR4:

####

SOL = solar irradiance

L95: Lean et al. (1995).

L95 (C00): temporally varying solar constant based on Lean et al. (1995) (Crowley, 2000).

L00: Lean (2000).

L02: Lean et al. (2002).

HS: Hoyt and Schatten (1993).

SK: Solanki and Krivova (2003).

####

Data for the two Lean and the Hoyt and Schatten reconstructions are easy to track down. The Lean et al (1995) data is available here:

ftp://ftp.ncdc.noaa.gov/pub/data/paleo/climate_forcing/solar_variability/lean_irradiance/lean1995data.txt

The Lean (2000) data:

ftp://ftp.ncdc.noaa.gov/pub/data/paleo/climate_forcing/solar_variability/lean2000_irradiance.txt

Note that there is a dataset included in Lean (2000) in which the minimums do not vary significantly. It is listed in the second column and identified as “11yrCYCLE”.

The Hoyt and Schatten (1993) data is part of the TSI reconstruction and composite comparison by Leif Svalgaard. It’s available in .xls format here, listed as Hoyt:

http://www.leif.org/research/TSI%20(Reconstructions).xls

The Crowley (2000) paper listed in the IPCC references is “Causes of climate change over the past 1000 years.”

http://www.sciencemag.org/cgi/content/abstract/289/5477/270

Crowley (2000) refers to a version of the Lean et al (1995) data: “An updated version of a reconstruction by Lean et al. (5) that spans the interval 1610-1998 was used to evaluate this mechanism.” The 1995 and 2000 versions of the Lean reconstruction are part of this post, and since I’ll be looking primarily at the effect of the data from 1900 to 1940, how Lean Crowley updated the last few years of data is not pertinent.

The Solanki and Krivova (2003) paper referenced by the IPCC is “Can solar variability explain global warming since 1970?”

http://www.mps.mpg.de/homes/natalie/PAPERS/warming.pdf

The data from Solanki and Krivova (2003) is difficult to find online (or I haven’t yet found it yet). The Solanki and Krivova (2003) data, however, is described by the National Center for Scientific Research (France) as, “The basic solar constant time series for the 20th Century simulations is constructed by Solanki and Krivova (2003). This data set is characterised by a 2-3 W/m2 increase in solar constant since the Maunder minimum. In the period 1850-2003 most of the total rise of about 1.5 W/m2 takes place in the period 1900-1950. Furthermore the solar cycle (and the variations therein over time) is included. The Solanki and Krivova (2003) time series is very similar to Lean (2000) but with some minor differences, mainly pre 20th century.”

http://www.cnrm.meteo.fr/ensembles/public/data/Descriptionsolar.doc

Again, I’ll be illustrating the effect of the erroneous data on the first part of the 20th Century, so any “pre 20th century” differences don’t come into play.

ADDITIONAL COMPARISONS OF TSI DATA USED BY THE IPCC

Figure 3 is a graph of the current understanding of the long-term variations in TSI, represented by the Svalgaard data (purple). Also included are the reconstructions of Hoyt and Schatten (green), Lean et al 1995 (blue), and Lean 2000 (red). The data begins in 1851.5 and runs the length of the individual datasets. The two Lean datasets and the Hoyt and Schatten data are available through the above links. The Svalgaard TSI data is also included in the linked spreadsheet from Leif.org. It’s referred to as the Leif data in Column C.

Note how sharply the Hoyt and Schatten (green) data rises from 1890 to 1950, but the current understanding of TSI variability is that there was no rise in the solar cycle minimums as illustrated by the Svalgaard (purple) curve. The two Lean datasets also have a significant rise from 1900 to 1960. I’ve “normalized” the Lean 1995 (blue) data in Figure 3 by subtracting 1.1 watts/meter^2 to show that it does follow the same general curve as the Lean 2000 data, with some minor differences, until SC20.

http://s5.tinypic.com/fp6qyp.jpg

Figure 3

AND WHAT EFFECT DOES THIS HAVE ON THE IPCC GCMs?

Assume for example that the GCMs are set to reflect the currently accepted climate sensitivity for variations in TSI, so that the minimum-to-maximum variation in the past three solar cycles results in a 0.1 deg C change in global temperature anomaly. If the solar cycle amplitude for those three cycles is approximately 1 watt/meter^2, then the scaling factor is 0.1. So in Figure 4, the TSI datasets have been scaled by that amount.

http://s5.tinypic.com/2mg6rll.jpg

Figure 4

The Hoyt and Schatten data would reflect a global temperature rise of approximately 0.3 deg C from 1890 to 1940, and that’s a significant portion of the actual rise in global temperature anomaly for the same period. The effect is the same for both Lean et al datasets, but to a lesser extent. But keep in mind, the rise in TSI minimums from the late 1800s to the mid-1900s does not exist. Refer again to the Svalgaard data.

And to refresh your memory on just how much global temperatures rose during the first part of the 20th Century, Figure 5 is a graph of HadCRUT3GL data from January 1850 to December 2007.

http://s5.tinypic.com/296kpdz.jpg

Figure 5

AND A LOOK AT TRENDS FROM 1900 TO 1940

Figure 6 compares trends in the scaled TSI data of the Svalgaard dataset from 1900 to 1940 to the trends of the three other TSI datasets. Again, the datasets have been scaled by a factor of 0.1 to reflect the impact of TSI on global temperatures. Due the variations in the solar cycle maximums, there is a slight trend in the Svalgaard data of ~0.009 deg C/decade from 1900 to 1940. The trend due to the incorrect variations in solar cycle minimums, on the other hand, for the Lean 2000 data is approximately ~0.026 deg C/decade, and for the Lean et al 1995 data, it’s ~0.027 deg C/decade. Then there’s the Hoyt and Schatten data with a trend from 1900 to 1940 of ~0.056 deg C/decade.

http://s5.tinypic.com/t7bknm.jpg

Figure 6

CLOSING COMMENT

As noted in past posts and in blog comments on this subject, if the natural climate forcings used to recreate the temperature rise in the first part of the 20th Century are erroneous, then the anthropogenic forcings used to recreate the global temperature variations in the latter part should not be assumed to be correct.