Alternate Title: On Pinker et al 2005 and the Positive Trend in the NCEP-DOE Reanalysis-2 Surface Downward Solar Radiation Flux
Over at Tallbloke’sTalkshop, Tallbloke has a post about the Pinker et al 2005 paper “Do Satellites Detect Trends in Surface Solar Radiation?” I’ve included (cross posted) that post in this one. Second, a positive trend also appears in the NCEP-DOE Reanalysis-2 Surface Downward Solar Radiation Flux data that’s available through the NOAA NOMADS website, though the positive trend is not as significant. Not too surprising, the multi-model ensemble mean of the climate models prepared for the IPCC’s upcoming 5th Assessment Report (AR5) do not show an increase in downward solar radiation at the surface. Another model failure to add to the ever-growing list?
First, Tallbloke’s post: Keep in mind, the satellite-based surface solar radiation flux “data” presented in Pinker et al is not based on direct measurements. It is calculated primarily from ISCCP cloud amount data. [I’ve added figure numbers to the following illustrations for my post.]:
Pinker et al: Do Satellites Detect Trends in Surface Solar Radiation?
Errrmm, why I have not heard of this 2006 [sic] paper before? 1.6W/m^2 per decade is a not insignificant trend. Did the IPCC ignore this?
Do Satellites Detect Trends in Surface Solar Radiation?
R. T. Pinker,1 B. Zhang,2 E. G. Dutton3
Abstract
Long-term variations in solar radiation at Earth’s surface (S) can affect our climate, the hydrological cycle, plant photosynthesis, and solar power. Sustained decreases in S have been widely reported from about the year 1960 to 1990. Here we present an estimate of global temporal variations in S by using the longest available satellite record. We observed an overall increase in S from 1983 to 2001 at a rate of 0.16 watts per square meter (0.10%) per year; this change is a combination of a decrease until about 1990, followed by a sustained increase. The global-scale findings are consistent with recent independent satellite observations but differ in sign and magnitude from previously reported ground observations. Unlike ground stations, satellites can uniformly sample the entire globe.
Figure 1 (Figure 1 from Pinker et al)
[END OF CROSSPOST]
As blogger Roger Andrews noted in his April 11, 2013 at 3:29 am comment at Tallbloke’sTalkshop:
According to Pinker et al’s Figure 5, reproduced below for reference, solar radiation over the oceans increased by about 5 w/m2 between 1983 and 2001 while solar radiation over land areas decreased slightly. How does this happen?
Figure 2 (Figure 5 from Pinker et al)
Also, does anyone have an explanation for the abrupt +/- 10 w/m2 downward excursion in the ocean record in 1994?
INTRODUCTION TO THE NCEP-DOE REANALYSIS-2 SURFACE DOWNWARD SOLAR RADIATION
The NCEP-DOE Reanalysis-2 is discussed in the 2002 Kanamitsu et al paper NCEP-DOE AMIP-II Reanalysis (R-2). The reanalysis begins in 1979 and at present runs through March 2013. Surface Downward Solar Radiation (radiation from the sun at the surface of the Earth) is one of the NCEP-DOE Reanalysis-2 fields available through the NOAA NOMADS website here, where it’s identified as “DSWRFsfc * surface Downward solar radiation flux [W/m^2]” in the drop-down “Fields” menu. Kanamitsu et al identifies Chou (1992) “A solar radiation model for use in climate studies” and Chou and Lee (1996) “Parameterizations for the absorption of solar radiation by water vapor and ozone” as the bases for the surface downward solar radiation flux outputs.
The NOAA NOMADS website allows users to select global coordinates for the desired outputs, so we’ll look at a few subsets in addition to the global values. The outputs are presented by the NOAA NOMADS website as absolute values, so I’ve converted them to anomalies using the base years of 1981-2010.
Keep in mind, this reanalysis is the output of a computer model—one that includes measurements (data) as inputs. It’s is not “raw” observations-based data.
GLOBAL
Figure 3 shows the global surface downward solar radiation flux anomalies from January 1979 to present, based on the NCEP-DOE Reanalysis-2. I’ve also included a linear trend line, which shows that the global values increased at a rate of about 0.22 Watts/m^2/decade since 1979, which is significantly less than the trend for the shorter period presented by Pinker et al (2005). The output includes lots of variability, what some would consider noise, so I’ve smoothed the output values with a 13-month running average filter in Figure 4.
Figure 3
##########
Figure 4
MODEL-REANALYSIS COMPARISON
Surface downward solar radiation is one of the variables from the CMIP5-archived climate models that are available through the KNMI Climate Explorer. Refer to the Radiation Variables, where it is identified as “rsds”. As you’ll recall, the CMIP5 archive is being used for the IPCC’s upcoming 5th Assessment Report (AR5). According to the multi-model ensemble mean of the CMIP5-stored models, using the RCP6.0 scenarios, the downward solar radiation at the surface should have been flat (decreasing slightly) during this period, if manmade greenhouse gases were responsible for global warming. But based on the Pinker et al (2005) paper and the NCEP-DOE Reanalysis-2, surface downward shortwave radiation increased significantly. The linear trends of the CMIP5 multi-model mean and the NCEP-DOE reanalysis are compared in Figure 5. Both model outputs have been smoothed in Figure 6.
Figure 5
##########
Figure 6
REANALYSIS COMPARISON – TROPICS VS EXTRATROPICS
In this preliminary look at the surface downward solar radiation outputs from the NCEP-DOE Reanalysis-2, I’ve also broken down the globe into 3 subsets: the Tropics (24S-24N) and the Extratropics for the Northern and Southern Hemispheres (24N to 90N and 90S to 24S, respectively). The Northern Extratropics show little trend in surface downward solar radiation, Figure 7, while Tropics at 0.28 Watts/m^2/Decade and Southern Extratropics at 0.32 Watts/m^2/Decade both show significant increases in downward solar radiation at the surface. I’ve smoothed the outputs for the three regions in Figure 8.
Figure 7
##########
Figure 8
The tropical and extratropical regions are presented individually in Figures 9 through 11 for anyone interested.
Figure 9
##########
Figure 10
##########
Figure 11
ADDITIONAL READING
On the thread at Tallbloke’sTalkshop, blogger Ned Nikolov in his April 10, 2013 at 11:37 pm comment also recommended Wild (2009) Global dimming and brightening: A review, and Pallé et al (2009) “Interannual variations in Earth’s reflectance 1999–2007”.
CLOSING
If there has been a rise in surface downward solar radiation since the late-1970s, early-1980s somewhere between the magnitudes presented by the NCEP-DOE Reanalysis-2 and Pinker et al (2005), then the hypothetical impacts of manmade greenhouse gases have obviously been overstated—especially when one considers that the IPCC’s climate models used for the attribution of global warming failed to simulate this rise in downward solar radiation at the surface over that period.
Another climate model failure comes as no surprise. In our earlier model-data comparisons, the CMIP5-archived models failed to properly simulate:
1. Satellite-era Sea Surface Temperatures
2. Global Land+Sea Surface Temperatures Since 1900
3. Satellite-Era Precipitation
To paraphrase Edwin Starr’s song “War”:
Climate Models. What’re they good for? Absolutely nothing.
‘Climate Models. What’re they good for?’
Telling one how little one understands about climate?
No matter that the forecasts are wrong. The UN needs them for planning purposes.
Frankly, I doubt if there is any significance to the trend lines on any of these charts. I don’t see any error bars or uncertainty statements, but as noisy as these charts look, I will bet dollars to donuts that the slopes are not statistically significantly different from zero.
Are the ‘First’ and ‘Second’ points in the opening two paragraphs transposed just to filter for people still drinking their coffee this morning? 😉
Excellent work Bob, thanks for taking this up. The calculation I did a few years ago of the extra forcing on the ocean required to achieve the 1993-2003 increase in sea level due to the steric component shown by the Colorado.edu satellite altimetry was around 4W/m^2. On WUWT, other commenters were convinced I’d made an order of magnitude error and that it should be 0.4W/m^2
I note that the Palle et al, Wild and Pinker papers go a long way to vindicating my position. Clearly the Colorado dataset is biased high for the seal level rise rate, but equally clearly, the biggest cause of the increase in rise rate was diminished cloud cover, not increased airborne co2.
Another convincing piece of research suggesting greenhouse gas warming effects could have been overestimated. But will our government’s scientific advisors draw it to their master’s attention? I wonder.
Geoff Withnell:
Huh? The error bars you refer to lie on an intra-yearly scale, not a multi-decadal scale since drift has never been reported as an issue. The climate system is noisy…are you expecting otherwise? If the data set were not noisy, I would be concerned.
Baseless assumptions like yours above do not further scientific understanding.
It has no sense to compare Pinker trend until 2005 with the trend of the longer record until 2012, when there is a clear peak around 2005. The peak in solar matches well with North Atlantic SST/OHC.
“clearly, the biggest cause of the increase in rise rate was diminished cloud cover, not increased airborne co2”
Yeah but, no but, yeah but, what if it were the CO2 wot dun it, ‘cos ti sucked up the extra ‘eat and wotsits and like it made the clouds go down and all?
“Also, does anyone have an explanation for the abrupt +/- 10 w/m2 downward excursion in the ocean record in 1994?”
No explanation by a layman’s guess. It could be the combination of Mt Pinatubo’s (persistent) aerosols increasing low cloud cover: http://www.sciencestew.com/articles/climate/CloudCover.jpg
I am willing to bet that equatorial DSR has predictive capabilities of global DSR. If that is the case, the most important measure we need to focus on is equatorial DSR. I would concentrate on seasonal measures to determine which set of 3 to 4 month running averages or just a seasonal set of months (IE AO-style), turns out to be the most powerful (IE MJJ? JJA? MJJA? etc). This would be an important addition to El Nino/La Nina measures since many times these measures are in neutral territory. However, figuring out with the equatorial clouds are doing during “neutral” conditions should be very useful.
philr1992 says:
April 12, 2013 at 6:46 am
Geoff Withnell:
Huh? The error bars you refer to lie on an intra-yearly scale, not a multi-decadal scale since drift has never been reported as an issue. The climate system is noisy…are you expecting otherwise? If the data set were not noisy, I would be concerned.
Baseless assumptions like yours above do not further scientific understanding.
I’m not skilled enough in the art of climate science not to make some of Mr. Withnell’s baseless assumptions. The plots look awfully noisy. When I see data like that and am doing chemstry or process studies, I tend not to draw any conclusions until I have reduced the noise. When I have my environmental hat on and start to make data quality assumptions, I usually drink until I get over it. It sometimes seems that climate science is best taken with a good single malt.
Could opposing conclusions be drawn from the same data?
I like that- “Clearly the Colorado dataset is biased high for the seal level rise rate”- lol!
Great post Bob, looks like yet another huge hole in the IPCC ‘zombie conjecture’ of CO2 driving our climate – just a shame their conjecture is unfalsifiable. Nothing much left of it now, apart from a few threads of bandage blowing in the wind, and lots of corrupt politicians with their fingers in their ears shouting “La..la..la….” at the top of their voices.”
It’s time they called time on this scam, otherwise the public are likely to take matter into their own hands and get rid of the lot of them!
Geoff Withnell April 12, 2013 at 6:08 am
Frankly, I doubt if there is any significance to the trend lines on any of these charts. I don’t see any error bars or uncertainty statements, but as noisy as these charts look, I will bet dollars to donuts that the slopes are not statistically significantly different from zero.
OT but strangely, just as I was reading the words “dollars to donuts” a speaker on the radio said “dollars to donuts”. Douglas Adams would have enjoyed the moment.
Bob: Yes the plots are noisy, as they should be. The climate system is very noisy, removing noise is essentially just smudging/blurring reality.
Notice how variations in tropical/northern subtropical cloud cover precede variations the ONI by 6-18 months. This clearly demonstrates the fact that angular momentum (tied to Rossby wave amplitude and the annular modes) drives the pressure/circulation fields that modulate ENSO.
quote
Also, does anyone have an explanation for the abrupt +/- 10 w/m2 downward excursion in the ocean record in 1994?
unquote.
Yep, that’s surely a mystery. A real problem. Looks like the ocean areas are short of aerosols and something supplied them. Pinatubo, maybe. But why is the whole plot falling? That’s a mystery right enough. It’s almost as if there’s a shortage of aerosols over the oceans. Now whatever could cause that? You’d need to think of a mechanism that stops the waves breakling and reduces phytoplankton DMS production. Or something.
Nope, can’t think of anything.
JF
[sigh] Google NASA shiptracks.
Read the source paper, Pinker 2005. The “data” are so heavily processed that they have estimated errors in the 10 to 30 watt per square meter range. The errors are larger than the entire Y-scale ranges on these plots. Show some error bars, or at least give some indication of the real world variances. You need much larger Y-scales and try to apply some real world common sense to this.
For now, Withnell is just stating the obvious. There is not enough information to say anything, except that the trends are zero until shown otherwise.
bw:
re your post at April 12, 2013 at 9:41 am
http://wattsupwiththat.com/2013/04/12/the-sun-was-in-my-eyes-was-it-more-likely-over-the-past-3-plus-decades/#comment-1273391
I would not go there if I were you. The same is true of almost all climatological time series.
Almost nothing used as data in climatology has been measured to be significantly different from zero at 99% confidence, and – upon investigation – most claims of 95% confidence turn out to be unjustifiable.
Richard
I’m unable to open the paper, what is the spectral range of the instruments involved?
Solar dimming and brightening isn’t a new or obscure issue, there are many papers on the subject. For instance, Martin Wild, JGR 2009, “Global dimming and brightening: A review.” The article just before the Pinker article in Science is titled “From Dimming to Brightening:
Decadal Changes in Solar Radiation at Earth’s Surface.”
Human emissions of SO2 is the precursor for tropospheric ammonium sulfate due to human activies (volcanos are also a major, though intermittent, source of sulfate). Sulfate relects sunlight (as a whitish haze) on reasonably clear days, and causes clouds to become whiter than they would otherwise be on cloudy days, also increasing amoungs of sunlight reflected. Sulfate is not the only human emission which affects how much sunlight gets through to the earth’s surface, but it is the most important one. Coal and oil both have sulfur to varying degrees, depending on source and type of coal and oil. Power plants can remove virtually all sulfur emissions with the use of “scrubbers,” but scrubbers didn’t begin to be used in the US until the 1970s. Refineries can remove the amounts of sulfur in residual oil, gasoline, diesel, etc. So the amount of SO2 emissions in a given year depends upon how much of each fuel is burned, and the degree of SO2 controls at different parts of the system.
SO2 emissions from human activites rose and then stabilized for much of the 1970s through early 1980s. When the Soviet Union (and eastern Europe) governments fell, very dirty, heavily subsidized industries could no longer be heavily subsidized. So from 1989 through about 1991, there was a large drop in tropospheric sulfate from industrial activities. Throughout the 1990s, sulfate in the US and Western Europe fell because of acid rail laws. The 1990s is the decade when dimming stopped and brighening occured.
Sulfate levels began to climb again around 2000, as China began its long ramp up of coal use (with no SO2 controls) to power their economy. See figures in “Anthropogenic and natural contributions to regional trends in aerosol optical depth, 1980–2006”, by Streets et al, JGR 2009.
Another paper is “Aerosol and cloud effects on solar brightening and the recent rapid warming” by Ruckstuhl et al. GRL, 2008. Here is the Abstract:
“The rapid temperature increase of 1 degree C over mainland Europe since 1980 is considerably larger than the temperature rise expected from anthropogenic greenhouse gas increases. Here we present aerosol optical depth measurements from six specific locations and surface irradiance measurements from a large number of radiation sites in Northern Germany and Switzerland. The
measurements show a decline in aerosol concentration of up to 60%, which have led to a statistically significant increase of solar irradiance under cloud-free skies since the 1980s. The measurements confirm solar brightening and show that the direct aerosol effect had an approximately five times larger impact on climate forcing than the indirect aerosol and other cloud effects. The overall aerosol and cloud induced surface climate forcing is ~ +1 W meter squared per decade and has most probably strongly contributed to the recent rapid warming in Europe.”
There is also a 2010 collection of 23 articles which appeared in AGU journals on the issue of solar dimming and brightening:
http://www.agu.org/contents/sc/ViewCollection.do?collectionCode=DIMBRIGHT1
I’m also unable to open the paper, giant crabs, fattened on CO2, are attacking me…. great discourse by the way, thanks for posting!
Jean Parisot and CRS, Dr.P.H., let’s try another link:
http://tallbloke.files.wordpress.com/2013/04/science-2005-pinker-850-4.pdf
Please advise.
John says:
April 12, 2013 at 9:54 am
So do you also then agree that the EPA Clean Air Act has led to global warming and, since asthma rates have also exploded since the act, asthma attack too?
Pamela Gray says: “I am willing to bet that equatorial DSR has predictive capabilities of global DSR.”
Tropical and global DSR anomalies have a correlation coefficient of 0.89.