Wednesday, July 11, 2007

Solar Cycle Hit Piece

Reuters reports on a new study claiming that recent solar activity is not responsible for the rapid global warming in the last twenty years. However, warmening biases abound as we see AGW skeptics described as a "dwindling group [that] pins the blame on natural variations in the climate system, or a gradual rise in the sun's energy output." Of course, there is no evidence presented to bolster this dwindling numbers claim. In fact, there is a growing record of skeptics breaking their silence.

Warmening credentials are also established in the previous sentence: "Most scientists say emissions of greenhouse gases, mainly from burning fossil fuels in power plants, factories and cars, are the prime cause of the current warming trend."

Curiously, the report highlights the statistical analysis of the data, mentioning that they were "smoothed to take account of the 11-year sunspot cycle," but doesn't explain why the scientists would do that, or even which dataset they are analyzing. This appears to be a sly resurrection of the discredited "hockey stick" which had claimed a rapid rise in warming using shaky temperature proxies and shakier computer models. Again, I have to stress, model outputs are predictions based on hypotheses, not data.

Also, the report misrepresents solar variability theories on warming, which claim that peak solar output in the 1970's is related to a twenty year, or so, warming process of the oceans, which then affects mean global surface air temperature. Sad attempt, really.

Update: Thanks to commenter JC, who snagged a PDF of the actual paper in the Reuters story, we learned that solar cycle over the last twenty years was compared to the GISS (Goddard Institute Surface Stations) temperature reconstruction. The researchers employed a "running mean" statistical analysis on both the most recent solar irradiance data and the GISS to tease out any correlation. After removing the 11 year curve, interestingly, the researchers found an overall, yet slight, downward trend in solar output during this time frame. Why that is strange, is because only three years ago, NASA teased out an increasing trend with their analysis. Well, the minimum for sunspot activity just ended earlier this year, so the data in Lockwood's paper are the latest. Other researchers are definitely going to look at the statistical analysis, because Lockwood and Frohlich applied a variable cycle length formula to the running mean analysis. The sun's 11 year cycle isn't exactly 11 years, but variable as well, so other solar experts will want to make sure the downward trend isn't an artifact. This really is a big blow against the solar forcing crowd for climate change because a downward trend with rising temperatures doesn't make much sense. Regarding the lag in their theories, I've seen it as little as two years when tied to sunspots, 4 to 5 years when tied to the hydrological cycle in North America, and as many as 10 (or 20) when there was close association with sunspot activity and the temperature record before 1940. A twenty year downward trend "outlives" all those lag timeframes, at least with the simplistic formulations they present right now. That American Thinker piece might be on to something, if Merrifield can get his math right and add up all the solar cycles together to show why a pretty flat solar irradiance trend now still means a global mean temperature increase.

Problems with this perceived downward trend are the observed increases in solar irradiance detected on Mars and Neptune. Astronomers are going to be busy trying to discount lower albedos (reflectivity) on those respective planets' surfaces as the source of the higher reflected output. Heh. As long as these scientists keep tying their research to climate change, the funding will keep coming.

8 comments:

  1. Dude - I guess the British Royal Society didn't see this article in American Thinker.

    Dwindling group, and all, you know?

    ReplyDelete
  2. Dude:

    Yup, I saw that article linked in Jawa Report this morning.

    ReplyDelete
  3. The datasets used are listed in the report and include sunspot numbers, solar flux measured by NASA, neutron count rates (eg - cosmic ray flux) measured by the Climax neutron monitor and TSI satellite measurements compiled by PMOD.

    The smoothing of the 11 year sunspot cycle is a common statistical technique - you'll see it used in just about any study that examines long term trends in solar activity.

    Re the use of models, the whole point of the study was not to use model outputs but examine and compare empirical data. I have a PDF copy of the study and would be happy to email it to you if you're interested.

    ReplyDelete
  4. JC:

    I would love it if you emailed me a copy. You can find my email address in my profile at the top of blog. I'll also update the post with the info from the paper.

    My main gripe about science reporting is how unclear they make everything, and the biases that creep in when using vague generalizations. Using an 11 year smoothing window for the statistical analysis (like applying a least square fit to boost signal to noise ratio) is obvious to anyone that knows about the 11 year sunspot cycle, but that wasn't mentioned in the story.

    I'll read the paper first, but can you tell me how they concluded that solar flux variability did not affect air temperature? Of course we're going to have hard numbers in regards to the sun. It's the "temperature" of the earth that's squishy and is reliant on GCMs, which was the point of my post. If I find out they just compared the solar record to a temperature record produced by the 13 GCMs approved by the IPCC, I'm going to be disappointed.

    ReplyDelete
  5. Joe, I've just emailed the PDF. I concur re media reporting of science - that's why I always try to read the original studies and if required, contact the authors direct. In this case, the paper hadn't even come out yet when it was first reported in Nature so I emailed Mike Lockwood direct and he emailed me the PDF a few days before it was published in the Royal Proceedings journal (nifty scoop :-).

    What the study does is compare the trends in solar activity with temperature (GISS surface measurements) and find the trends diverge around 1980. Eg - before 1980, temperature and solar activity show good correlation but from 1980, they go off in different directions, which leads him to conclude that solar influence on global warming is minimal.

    Personally, I think the study is useful in that it incorporates the latest satellite measurements of TSI and sunspot numbers to include the latest solar minimum which just happened over the last month or two. But he's actually retreading ground that has been well covered in previous studies (eg - Foukal 2006, Scafetta 2006, Lean 1999, Waple 1999, Frolich 1998, etc)

    ReplyDelete
  6. Joe, thanks for the credit (although I'm just the messenger, was lucky to get a copy of Mike Lockwood).

    The upward trend you're talking about is probably Willson's reconstruction of satellite data which shows a very slight upward trend. When Scafetta analysed Willson's reconstruction, he concluded "since 1975 global warming has occurred much faster than could be reasonably expected from the sun alone" (Scafetta 2006). Judith Lean at PMOD did an alternative reconstruction on the same satellite data and found no upward trend - secular solar trends are practically level. Then Mike Lockwood's latest analysis shows a very slight downwards trend. The fact that they're arguing about whether there's a trend or not shows just how slight the trend is.

    Re Mars and Neptune, their warming is clearly not related to solar variations. Mars' albedo has been decreasing over the last 30 years due to observed dust storms. Neptune's brightening is a seasonal response (it's orbit is 164 years).

    ReplyDelete
  7. I think that what everyone is missing is that the temperature proxy data shows that end of the 20th century has been cooling. The instrumented readings say we are warming, but then these are the same readings that say there is no urban heat island effect.

    This gives rise to the question, are the temperature proxies wrong?

    Another issue is that the study it trying to do a solar activity to temperature and ignores the possible linkage of solar, cosmic rays, clouds, to temperature.

    As to the neutron count rates, Nir J. Shaviv shows in various papers that the low energy cosmic rays are not the ones in question and that the Climax does not monitor the proper energy levels. This is a bit of mis direction.

    ReplyDelete
  8. Anon:

    You bring up some good points. Currently, there is an audit going on of the siting of the surface stations, and they're finding that quite a few of these "rural" stations which had the rural cooling correction applied to them (so they were corrected upward), are actually placed near asphault, air conditioning units, and brick walls.

    As to this certain paper, it did seem to me that they stuck the cosmic ray hypothesis to be tested as almost an afterthought, associating neutron count with solar wind. The higher the neutron count, the stronger the solar wind, which meant few cosmic rays hitting the earth. This was the weakest part of the paper, because, as you say, it did not account for higher energy particles, and indeed, did not directly measure them.

    ReplyDelete

Please don't comment on posts more than 4 years old. They will be deleted.