- JoNova - https://joannenova.com.au -

Scafetta 2013: Simple solar astronomical model beats IPCC climate models

Nicola Scafetta has a new paper (in long line of papers) on a semi-empirical model which has a better fit than Global Circulation Models (CGM) favored by the IPCC. We ought be careful not to read too much into it, but nor to ignore the message in it about the grand failure of the GCM’s. Scafetta used Fourier analysis to find six cycles, then uses those six cycles to produce a climate model he runs for as long as 2000 years which seems to match the best multiproxies. In terms of discovering the absolute truth about the climate, this is not an end-point way to use Fourier analysis, as it is just “curve fitting”  With six flexible cycle frequencies (plus amplitude and phase) there are 18* 6 tuneable parameters, more than enough to model any wiggly line on a graph, and there are scores of astronomical cycles to pick from. *.[Nicola Scafetta replies to this below, pointing out he uses the “6 major detected astronomical oscillations”, and their phases are fixed. I am happy to be corrected. His model is more useful than I thought. Apologies for the misunderstanding.   – Jo]

But Scafetta’s work suggests it’s madness not to pay attention to astronomical cycles, and points to major flaws in the IPCC simulations. Compare the two types of models: Scafetta’s simple model uses [natural astronomical] cycles and assumes there is a connection [there might be, it is speculative] but curve fits to produce predictions**. The unverified IPCC models assume CO2 has a powerful influence (backed up by laboratory experiment, but not backed up with empirical data from the climate) — then the IPCC assume powerful positive feedbacks that more than double the effect of CO2 (without empirical evidence to support those assumptions) and in a sense, curve-fits volcanic, solar, and aerosols to flex the line to match the data. We know the IPCC models don’t work, they don’t hindcast the last 2000 years, and didn’t predict the last 20. It obvious from Scaffetta’s work that we ought be investigating these natural cycles, and that the IPCC models are hopelessly incomplete.

1. The IPCC depends on the claim that their models include all the important forcings. Their attribution claim has always been “we can’t model the recent temperature rise without using CO2 forcings”. This is argument from ignorance, and Scafetta shows just how ignorant it is.

2. IPCC models don’t produce natural cycles. IPCC models are missing important natural forcings (if only we knew what they were). Scafetta takes the thermometer records, and the paleoclimate records, points at natural cycles, some of  which are well known and long established, some of which are purely speculative, and shows how the IPCC models do not produce the same natural cycles. If those cycles (or ones like them) have a physical cause it means the IPCC models don’t include those forcings. A monster flaw.

3. Look at the “pause”, the long plateau in temperatures. The IPCC favoured models failed to predict it (von Storch). The natural cycles might explain the flatness in global surface temperatures since 2000. A simple solar-astronomical-model based on these natural patterns outperforms the inadequate, over-rated, billion-dollar-IPCC models.  The caveat being that in a chaotic system the true natural cycles may be difficult to discover.

4. Natural cycles may be driven by the orbits of planets and their effects on the sun. This is speculative, but very much worth discussing. According to Scafetta, there may be natural cycles of 9.1, 10–11, 19–22 and 59–62 years. (Several of these cycle lengths also appear in Ian Wilson’s work on a mechanism where lunar tides in our atmosphere may help trigger ENSO conditions). It is believable that the resonant effect of the orbits of planets acts on the solar dynamo, in ways we do not yet know, affecting it’s luminosity and magnetic field, and that these small solar changes are then amplified on Earth’s climate. (See, e.g.  Svensmark and cosmic rays, or Lam et al 2013, who found the solar wind may influence Rossby waves and atmospheric pressure.)

5. Monopolistic science funding has taken years to not find the answer. Many research programs and grants have focused on making a CO2 driven model work. How much money have governments spent figuring out role of natural cycles in a climate that has always changed? If governments could tax planets, there might be 23 solar-system coupled climate models, and they might just work a whole lot better than the CO2 ones.


IPCC Climate models don’t match the turning points

I have long said that it was obvious the CO2 theory does not fit the data because the models were not able to reproduce any turning points in our climate. The models don’t explain why the world was warm in the medieval times, cool 300 years ago in the little ice ages, nor do the models explain the shorter 30 year cool periods in the last 150 years either.

Fig 17 (below) shows how climate models (GCM simulations) fail during the last 13 years, overdo the volcanic cooling spikes, fail to reproduce the well known cooling period from  1880-1910.

Fig 17 A reproduction of Fig. 1 in Gillett et al. (2012)with additional comments that highlight the major mismatches between the GST record (black) and a set of simulationsmade with
CanESM2. The figure highlights problems common to all CMIP5 GCMs. From Scafetta (2013a).

Scafetta reviews papers showing records of some natural cycles go back thousands of years:

“Quasi-decadal, bidecadal and 60-year oscillations and other longer oscillations have been detected in numerous records covering centuries and millennia. For example, Jevrejeva et al. (2008) and Chambers et al. (2012) showed a quasi 60-year cycle in the sea level rise rate since 1700; Klyashtorin et al. (2009) showed that numerous climate indexes present a long-term 50–70 year oscillations during the last 1500 years; Knudsen et al. (2011) showed a persistent quasi 60-year cycle in the Atlantic Multidecadal Oscillation throughout the last 8000 years; a quasi 20-year and 60-year oscillations also appear for centuries and millennia in some Greenland temperature records (Davis and Bohling, 2001; Chylek et al., 2012).

Fig. 18 (below) reproduces Fig. 10 in Scafetta (in press) that shows two relatively global climatic indexes since 1700: the global sea level record (Jevrejeva et al., 2008) and the North Atlantic Oscillation (NAO) reconstruction (Luterbacher et al., 1999, 2002). The right panels show the multi-scale acceleration analysis (MSAA) of these two records and highlight the presence of a common major quasi 60-year oscillation since 1700. This oscillation is revealed by the alternating green and red colors indicating that the local acceleration of the records varies from negative to positive values, that is, there is an oscillation.”


Fig 18 [A] Global sea level record (Jevrejeva et al., 2008) (left) and its MSAA colored diagram (right). [B] North Atlantic Oscillation (NAO) (Luterbacher et al., 1999; Luterbacher et al., 2002) (left) and itsMSAA colored  diagram(right). In [B] the colors are inverted. Note the common quasi 60 year oscillation since 1700 indicated by the alternating green and red regions within the 30–100 year scales. From Scafetta, in press.

Climate models cannot reproduce the medieval warm period

The blue line represents the newer multiproxy studies of the last thousand years showing the Medieval Warm Period and the Little Ice Age. For a few years the hockey-stick shaped type of reconstruction was popular and Crowley’s model fitted it. (See the bottom of Graph A). But the Crowley model does not fit the Ljundqvist (2010) nor Leohle (2008) reconstruction (top, Graph A). Scafetta suggests that if Crowley’s model had more sun, less volcano’s, less aerosols, and less CO2, it would fit the Moberg reconstruction with Hadley temperatures from 1850 onwards (see Graph B).

“The climate models that predicted a very small natural variability and that were used to fit the hockey stick temperature records cannot fit the  recent proxy GST reconstructions casting doubts on their accuracy. Still recent millennium simulation studies using modern solar models (that is, Wang et al., 2005) are able to predict only hockey-stick temperature graph showing an average cooling from the 900–1300 MWP to the 1300–1800 LIA up to ~0.3 °C, and just half of the empirically measured 11-year solar signature on the climate (see Feulner and Rahmstorf (2010) and IPCC (2007) Fig. 6.14: http://www.ipcc.ch/publications_and_data/ar4/wg1/en/figure-6-14.html).”

Fig. 23. [A] Comparison between the original energy balance model prediction by Crowley (2000) versus the hockey stick temperature graph by Mann et al. (1999) implying a MWP as warm as the 1900–1920 period, and two non-hockey stick recent paleoclimate GST reconstructions (Loehle and Mc Culloch, 2008; Ljungqvist, 2010) showing a far larger preindustrial variability and a MWP as warm as the 1940–2000 period. [B] (Bottom) the volcano, solar and GHG + Aerosol temperature signature components produced by Crowley (2000) model are scaled to fit (Moberg et al., 2005) 1850 by HadCRUT4, which also shows a MWP as warm as the 1940–1970 period. See Scafetta (2013a, 2013b) for more details.

“Fig. 24A shows the proposed solar model versus the extra-tropical Northern Hemisphere temperature reconstruction by Ljungqvist (2010) (black)”

The blue line at the bottom of A is the 115 year oscillation. Fig 24B is the solar model (red) compared to HadCrut4.

Fig. 24. Scafetta (2012c) three-frequency solar model (red). [A] Against the Northern Hemisphere temperature reconstruction by Ljungqvist (2010) (black). The bottom depicts a filtering of the temperature reconstruction (black) that highlights the 115-year oscillation, h115(t), (blue). [B] The same solar model (red) is plotted against the HadCRUT4 GST (black)merged in 1850–1900 with the proxy temperature model by Moberg et al. (2005) (blue). The green curves highlight the quasi millennial  oscillation, h983(t), with its skewness that approximately reproduces the millennial temperature  oscillation. Note the hindcast of the Maunder and Dalton solar minima and relative cool periods, and the  projected quasi 61-year oscillation from 1850 to 2150. Adapted from Scafetta, 2013a.

Scafetta’s solar-astronomical model suggests the temperature will be pretty stable between now and the 2030’s (which, given the dominant 60 year Pacific Decadal Oscilation has been suggested by many including Akasofu). In fig 27 the left graph are the IPCC model projections (depending on how much CO2 we emit). On the right graph are the solar-astronomical model projections — which include CO2 emission “scenarios”.

Fig. 27. [A] All CMIP5 model projections versus the HadCRUT4 GST record. [B] The solar–astronomical  semi-empirical model, Eq. (13) with β = 0.5, against the HadCRUT4 GST record: a common 1980–2000 baseline and annually resolved records are used in the large figures while the monthly HadCRUT4 GST record is used in the inserts. The figure highlights the better performance of the solar–astronomical semi-empirical model versus the CMIP5 models.


Fig 28 is a close up of the last 30 years and the next few.

Fig. 28. Eq. (13) with β = 0.5 (blue) and the original CMIP5 ensemble mean model (red) against six global temperature estimates (HadCRUT3, HadCRUT4, UAHMSU, RSSMSU, GISS and NCDC), which were baselined with HadCRUT4 from Jan/1980 to Dec/1999. Temperature data from: http://www.ncdc.noaa.gov, http://www.metoffice.gov.uk, http://data.giss.nasa.gov, http://www.remss.com/, http://vortex.nsstc.uah.edu/.

Scafetta draws a mud-map of possible natural factors and their interactions. This is speculation, “possibles”. We need more research.

Fig. 29. Network of the possible physical interaction between planetary harmonics, solar
variability and climate and environment changes on planet Earth.
Adapted with permission after Mörner, 2012, see also Scafetta, 2013a.

“Fig. 26 compares the four CMIP5 ensemble average projections (panel A) and the solar–astronomical semi-empirical model using
β = 0.5 in Eq. (13) (panel B) against the HadCRUT4 GST record: a common 1900–2000 baseline is used. The figure highlights the superior performance of the solar–astronomical semi-empirical model versus the CMIP5 ensemble mean models.”

Fig. 26. [A] The fourCMIP5 ensemble average projections versus the HadCRUT4 GST record. [B] The solar–astronomical semi-empiricalmodel, Eq. (13) with β = 0.5, against the HadCRUT4 GST record: a common 1900–2000 baseline is used. The figure highlights the better performance of the solar–astronomical semi-empirical model versus the CMIP5 models, which is particularly evident since 2000 as shown in the inserts.

 Scafetta’s model suggests climate sensitivity is more likely 0.3 °C to 1.8 °C  [*between 1 and 2.3 C].

[The smaller climate sensitivity figures apply to the paragraph below]

“The proposed semi-empirical model may produce  more reliable projections for the 21st century, which are far less alarmist than the current CMIP5 projections. Under the same anthropogenic emission scenarios, the model projects a possible 2000–2100 warming ranging from 0.3 °C to 1.8 °C. This range is significantly below the original CMIP5 GCM ensemble mean projections spanning from about 1 °C to 4 °C.”

I think Scafetta’s model is useful for showing the public how easy it is to get very different results in climate modeling, and to point out the major flaws in the GCMs. Curve fitting cycles of unknown mechanism is dubious, but not more so than using assumptions of feedbacks for which there is no empirical evidence. It’s all muddy modeling.

Above all, it’s sheer craziness to ignore resonance in astronomical cycles. Tallbloke’s Talkshop has a post on this paper too. Tallbloke’s comments are particularly interesting.

Nicola Scafetta’s site is: http://people.duke.edu/~ns2002/#astronomical_model.

SPPI has a another earlier related reprint published from July 2013.



Scafetta, Nicola (2013) Discussion on climate oscillations: CMIP5 general circulation models versus a semi-empirical harmonic model based on astronomical cycles Earth-Science Reviews  Volume 126, November 2013, Pages 321–357

*UPDATE: Nicola Scafetta replies


thank you for the nice presentation of my work.

There are only a couple of misunderstanding I would like to clarify.

1) You writes: “Scafetta used Fourier analysis to find six cycles, then uses those six cycles to produce a climate model he runs for as long as 2000 years which seems to match the best multiproxies. In terms of discovering the absolute truth about the climate, this is not an end-point way to use Fourier analysis, as it is just “curve fitting”. With six flexible cycle frequencies (plus amplitude and phase) there are 18 tuneable parameters, more than enough to model any wiggly line on a graph, and there are scores of astronomical cycles to pick from.”

This is not fully correct. My methodology is very similar to that currently used to predict ocean tided
that make use of up 40 harmonics. See here: http://en.wikipedia.org/wiki/Theory_of_tides#Tidal_constituents

The methodology is based on the identification of the most relevant astronomical oscillations and then only the amplitudes of the oscillations are fit upon the temperature data. In my case, I use the 6 major detected astronomical oscillations. These are not randomly chosen from an infinite set of possible astronomical oscillations, but are the very major gravitational and electromagnetic oscillations of the heliosphere. Also the phases are fixed by the astronomical oscillations.

As done with the ocean tides only the amplitudes of the oscillations are fit on the data. The phases of the decadal and multidecadal cycles (4 cycles on 6 used) are only optimized on the data because the fitted phase almost identically corresponds to the theoretical ones, as Figure 3 shows. Moreover the phase of the secular and millennial cycles (figure 24) are fixed only by astronomical considerations.

Thus, the only real free parameters are the 6 temperature oscillation amplitudes.

Thus, the good correlation that you find in my graphs (e.g Figure 26) is not due just to curve fitting, but mostly to the fact that the climate presents oscillations synchronized to the major oscillations of the heliosphere.

The model, moreover, is tested on its hindcast ability. The good correlation would exist also by fitting the amplitude just in the period 1850-1950. Also in this case the model reconstructs-forecasts the observed 1950-2013 variability.

2) you write: ” Scafetta’s model suggests climate sensitivity is more likely 0.3 °C to 1.8 °C”.

This is incorrect. That refers to the projected warming range from 2000 to 2100. The estimated climate sensitivity range is between 1 and 2.3 C. (This is an upper limit as explained in the paper).

Beside these two details, your article is a very good summary of my paper.

I am keeping a forecast experiment at my web-site where my model forecast since 2000 is compared against the IPCC GCMs

Nicola Scafetta


**I’ve edited this sentence to fit the updated response.

8.8 out of 10 based on 60 ratings