Guest essay by Dr. Antero Ollila
Water vapor feedback has remained a topic of debate since 1990. The laymen do not know that water has an essential role in calculating the warming effects of GH gases. In all Anthropogenic Global Warming (AGW) models the Relative Humidity (RH) stays constant. It sounds like a very neutral and harmless assumption. When GH gases increase the atmospheric temperature, the constant RH means that the absolute amount of atmospheric water also increases. Because water is about a 15 times stronger GH gas than carbon dioxide, this small increase of water content increases the temperature as much as GH gases. According to IPCC, the radiative forcing of GH gases is doubled by water (AR3 and AR4). IPCC calls this feature a positive feedback of water.
According to my spectral analysis calculations, the positive water feedback effect would be this magnitude. But in AR5 (20139 IPCC reports in Science Basis , p. 666: “The contribution of water vapor to the natural greenhouse effect relative to that of carbon dioxide (CO2) depends on the accounting method, but can be considered to be approximately two to three times greater.” There are no references to any scientific publications. This raises a question, if IPCC really knows how to calculate the effects of positive water feedback. The knowledge of water feedback in not getting more accurate but more inaccurate according to IPCC. It is a little bit of strange development after all the money used for climate change research.
The opponents of AGW theory have pointed out the RH measurements do not show constant RH trends as we can see in figure 1, (NOAA, http://www.esrl.noaa.gov/gmd/aggi/)
Figure 1. Relative Humidity (RH) trends from 1948 to 2012
I have called this figure a sight test. If you see that the trends are essentially at the horizontal position and not bending downward, you should test your eyes with the doctor. This is an example but not the only one that IPCC denies the direct measurements, when they do not fit into their theories.
The warming calculations are based on the climate models both in simple as well as in computer based General Circulation Models (GCM). The warming effects of GH gases are very small and there are many other factors. The ongoing pause in the global temperature is a good example of these other factors, which IPCC call natural effects. It is therefore quite difficult to show in the real climate, if there is positive water feedback or not.
Luckily there is one experiment which was organized by the Mother Nature itself. This was the eruption of the Mount Pinatubo. The main eruption began on the island of Luton in the Philippines on the 3rd of June, 1991 and concluded on the next day. Four large explosions generated eruption columns reaching the heights of up to 24 km in the stratosphere. The estimate of the stratospheric mass increase was 14 – 20 Mt of SO2, which created 21-40 Mt of H2SO4–H2O aerosols. The eruption also injected vast quantities of minerals and metals into the troposphere and stratosphere in the form of ash particles. The aerosols created a global layer of sulfuric acid haze over the globe and the global temperatures dropped about 0.5 °C in the years 1991 – 1993. Because of the aerosols and ash particles, the incoming solar radiation decreased 6 W/m2. At the same time there was a maximum increase of downward longwave radiation flux of 4.5 W/m2 caused by the very same aerosols and ash particles. Totally the radiative forcing at the surface was in maximum 1.5 W/m2. This radiation anomaly can be compared to the decrease caused by the doubling of the CO2 concentration from 280 ppm to 560 ppm. According to IPCC data this change is 3.7 W/m2.
I have carried out a dynamic analysis of the temperature effects caused by the eruption. I wanted to test two options for the climate sensitivity parameter (CSP). The radiative forcing (RF) at the top of the atmosphere has a linear relationship to the global mean surface temperature change dT:
dT = CSP*RF (1)
IPCC uses still equation (1) in its latest report AR5 but IPCC no longer keeps the value of CSP as almost constant. There is no information in AR5 as to, what the real value of CSP is or in which way it varies. The CSP value of 0.5 K/(W/m2) has been used in the former reports of IPCC and it includes the positive water feedback. Actually the CSP value of 0.5 still has the decisive role, because IPCC reports in AR5 that the transient climate sensitivity value is likely to lie in the range 1 to 2.5 °C giving the average value 1.75 °C. This value is almost the same as calculated by equation (1): dT = 0.5 K/(Wm-2) * 3.7 Wm-2 = 1.85 K. The value of 0.27 K/(W/m2) has been used showing no water feedback.
I carried out two simulations by a simple dynamic model to test these two CSP values. The results are depicted in figure 2. It is very clear that the CSP value of 0.5 gives results which deviate from the real response of the global temperature decrease.
Figure 2. Simulation of the Mount Pinatubo eruption using different climate sensitivity parameters.
There are two former studies about the dynamic temperature response in the Mt. Pinatubo eruption. Hansen et al. applied GCMs by name SI94 and GRL92 in their simulations published in 1992. Soden et al. also applied a GCM in the research study published in 2002. They also included the absolute atmospheric water content as a variable. The major results were that the GCM simulations could calculate the dT values close to the measured value, if the positive water feedback was included. The water content was calculated using the NASA Water Vapor Project (NVAP) values.
So there are research results which show very different results. What could be the reasons? The reasons are rather simple to point out. All other researchers including me have used maximal solar irradiation decrease value of -6 W/m2 but in these two studies the researchers have used the value -4 W/m2. In the same way other researchers have used the maximal deviation value of -0.5 C during the eruption but in these two studies the value of ~-0.7 C has been used. Soden et al. have included the RH change during the eruption and they have been able to show that there is positive water feedback needed to explain the temperature decrease.
Figure 3. Relative Humidity trends according to NCEP/NCAR Reanalysis and NVAP-M datasets.
In Figure 3 the NVAP dataset values as well the NCEP/NCAR (National Center for Environmental Prediction / National Center for Atmospheric Research) values are depicted. The NVAP water content trends show great seasonal changes of about 3 TPW mm. Soden et al. have reported that there has been ~0.75 TPW mm peak reduction during the Pinatubo eruption. The graphs show that the peak reduction estimate can be regarded a correct estimate. But this choice of using the peak values only can be questioned, because the trend line of NVAP-M values show an increased rate of absolute water content and it is an opposite trend! The figure 3 shows that the RH measurements are not accurate enough to be used as evidence about RH feedback. The average value of these two datasets show practically zero trend during the eruption.
The results of Hansen et al. and Soden et al. can be explained by proper data selection, which is called cherry picking. I have used the most commonly applied data values and the results show no water feedback. This result means that the Climate Sensitivity (CS) is 0.27 K(W7m2) * 3.7 W/m2 = 1 K. Many researchers using different methods have found that the CS would be in the range of 1 to 1.2 C, if the RF value of 3.7 W/m2 of CO2 (increase from 280 ppm to 560 ppm) has been applied. There is only one question remaining, is the RF value of 3.7 W/m2 correct. According to my analysis this value is calculated in the atmosphere, where there is constant RH and therefore positive water feedback. My calculations using three different methods show that the RF value of doubling the CO2 concentration is 2.12 W/m2. Therefore climate sensitivity is only 0.27 * 2.12 = ~0.6 C assuming constant absolute water content in the atmosphere.
The paper:
Physical Science International Journal, ISSN: 2348-0130,Vol.: 9, Issue.: 4
Climate Sensitivity Parameter in the Test of the Mount Pinatubo Eruption
Antero Ollila1*
1Department of Civil and Environmental Engineering (Emer.), School of Engineering, Aalto University, Otakaari 1, Box 11000, 00076 AALTO, Espoo, Finland.
Abstracts
The author has developed a dynamic model (DM) to simulate the surface temperature change (ΔT) caused by the eruption of Mount Pinatubo. The main objectives have been 1) to test the climate sensitivity parameter (λ) values of 0.27 K/(Wm-2) and 0.5 K/(Wm-2), 2) to test the time constants of a simple first-order dynamic model, and 3) to estimate and to test the downward longwave radiation anomaly (ΔLWDN). The simulations show that the calculated ΔT of DM follows very accurately the real temperature change rate. This confirms that theoretically calculated time constants of earlier studies for the ocean (2.74 months) and for the land (1.04 months) are accurate and applicable in the dynamic analyses. The DM-predicted ΔT values are close to the measured value, if the λ-value of 0.27 K/(Wm-2) has been applied but the λ-value of 0.5 K/(Wm-2) gives ΔT values, which are about 100% too large. The main uncertainty in the Mount Pinatubo analyses is the ΔLWDN flux, because there are no direct measurements available during the eruption. The author has used the measured ERBS fluxes and has also estimated ΔLWDN flux using the apparent transmission measurements. This estimate gives the best and most consistent results in the simulation. A simple analysis shows that two earlier simulations utilising General Circulation Models (GCM) by two research groups are depending on the flux value choices as well as the measured ΔT choices. If the commonly used minimum value of -6 Wm-2 would have been used for the shortwave anomaly in the GCM simulations, instead of -4 Wm-2, the ΔT values would differ from the measured ΔT values almost 100%. The main reason for this error seems be the λ-value of 0.5 K/(Wm-2).
Full paper here: http://sciencedomain.org/abstract/13553
Quantitative Radiative Forcing with Relative Humidity
Ferenc Miskolczi does a quantitative Line By Line (LBL) radiative forcing calculation using the available balloon humidity measurements with numerous latitude/longitude blocks and elevation sections. e.g., see:
The greenhouse effect and the infrared radiative structure of the Earth’s Atmosphere
No, he doesn’t. He says the TIGR data are too unreliable. He says:
“In this article we use the GAT atmosphere as the representative temperature and humidity structures of the global average climate. For studying possible long term changes in the global average optical thickness (due to changes in GHG content of the atmosphere) the TIGR2 archive is not suitable. “
He’s just using a standard atmosphere.
Thanks for clarifying Nick. I was thinking of previous graphs/analyses i had seen where he uses the more detailed data, (NOT constant over time) :
Compare: Greenhouse Effect and the IR Radiative Structure of the Earth’s Atmosphere
(One substantial issue are the adjustments from TIGR 2 to 2000.)
Summary fig. H2O vs CO2: “Former NASA scientist defends theory refuting global warming doctrin
Here’s another graph of NOAA humidity vs time for 3 elevations
Michel Moon realized that I am the disciple of Miscolczi, because we have found the same kind of results about the GH gas effects. It is true that the results about the GH effects calculated by me and by Miskolczi are close to each other. They are independent calculations but they have been calculated by different spectral analysis tools. I have used Spectral Calculator and the water absorption of this tool is little bit weaker (no continuum absorption/emission included). It means that my results are a little bit more “conservative” than those of Miskolczi: for example 11 % versus 9 % concerning the portion of CO2 in the GH phenomenon).
If you are able to get Nick Stokes to back down it will be a huge first. You have him on the run, keep it up. The NASA NVAP-M study results from satellite seem to have been suppressed.
“The truth shall set us free!!!”
Why would YOU use ANY tool that does not compute the real effects. Water droplets in clouds of just a few mils (0.1 mm) are near perfect black body (isotropic) radiators at the cloud Temperature. Spectral Calculator calculates one dimensional beam absorption coefficient.
It does not correctly predict radiant energy transmission in a three dimensional atmosphere.
G
Well, maybe I got too exited, when I learned by trial and error how to add an image on the comments. I like to show an image, which shows in the form of table, that my spectral analyses give the same results as by Kiehl & Trenberth, when using the same wrong atmosphere. There are also the results carried out in the correct average global atmosphere. The difference is in the amount of water. Water is on the driving seat, when we think about the GH effects.
To David L. Hagen. It is true that Miskolczi has used in his latest studies more detailed atmospheric models than me. I have used a single column calculations. Anyway, the results are pretty close to each other. The question of water content is not quite sure yet. If we use the balloon measurements since 1948, the absolute water content has decreased and compensated the warming effects of CO2. I would be cautious and wait another 10 years. The accurate RH measurements started around 1980 with the new Humicap technology from Vaisala, and now the accuracy and reliability is much better.
Water content depends on latitude and season. It might complement your present work to study dependence of sensitivity on latitude. Would such a study be able to definitively separate the effects of CO2 from those of water vapor?
Separation of warming effects of water and other GH gases is not a problem. I have carried out tens of such spectral analysis calculations. These calculation do not solve the problem here: in which way the water in the atmosphere behaves as the GH gases increase the temperature. Stephen Wilde offers one explanation and I think that Miskolczi has introduced a pretty complicated theory with numerical calculations. This theory shows that (with simple terms) the GH effect of the average global clear sky is constant. It means that water eliminates the warming effects of GH gases. I have no scientific skills to judge Miskolzci’s theory. I have noticed that it has been ignored but I have seen only one trial to explain, why it cannot be true.
So far I keep it clear the the assumption of constant RH is not correct.
The RH graphs at different altitudes. I am sorry that the url address is not to the direct address of the data page of NOAA.The data base, what I have used in creating the graphs in my story are here: . Hopefully it went right.
So, something went wrong. I write it here. http://www.esrl.noaa.gov/psd/cgi-bin/data/timeseries/timeseries1.pl
Miskolczi’s observations appear to be correct but as far as I know he offers no explanation.
An explanation arising from the fact that higher humidity allows condensation out at a lower height and warmer temperature (so as to radiate more to space) is simple, neat and complies with basic physics.
A powerful and thought-provoking thread. Thanks to all.
Regarding:
“There is only one question remaining, is the RF value of 3.7 W/m2 correct. According to my analysis this value is calculated in the atmosphere, where there is constant RH and therefore positive water feedback.”
I read Dr. Roy Spencer’s blog, and there he does not dispute the 3.7 W/m^2 per 2xCO2 figure or claim it includes any feedbacks. He uses it as a pre-feedback figure. Given his views and expertise, I expect him to say it includes a positive feedback if it does.
I have calculated the RF value of CO2 in the same way as Myhre et al. and I got the formula RF = 3.12*ln(C/280), where C is the concentration of CO2 in ppm. (Myhre: RF = 5.35*ln(C/280). There are three papers referred by IPCC and the other two papers are those of Hansen et al. and Shi. These three papers give different formulas but the results are very close to each other. Strange enough, one of these papers namely Shi, says this way (exact quote): “Water feedback is treated by assuming a fixed relative humidity (FRH) in the model…”. I draw a conclusion that also Myhre’s and Hansen’s calculations have been carried out in the fixed RH conditions, which means doubling the RF value. That is the only explanation I can find.
I have noticed that Myhre’s equation has been commonly used in various analyses as expressing the right RF of CO2 but I have not found anybody carrying his/hers own calculations. Those who say that Myhre’s formula is correct, I would ask: have you calculated by yourself and what are the results of your calculations?
There is a special paper of mine about this issue: http://www.seipub.org/des/paperInfo.aspx?ID=17162
Regarding:
‘Strange enough, one of these papers namely Shi, says this way (exact quote): “Water feedback is treated by assuming a fixed relative humidity (FRH) in the model…”. I draw a conclusion that also Myhre’s and Hansen’s calculations have been carried out in the fixed RH conditions, which means doubling the RF value. That is the only explanation I can find.’
However, water vapor feedback is a feedback. Assumption of constant relative humidity is a simplistic way of evaluating this feedback. The 3.7 W/m^2 per 2xCO2 is a pre-feedback figure that even Dr. Roy Spencer goes along with in his blog.
I noticed your paper stating water vapor presence in terms of Average Global Atmosphere – but representative water vapor presence is less because water vapor’s effect (like all greenhouse gases) is sublinear, and most of the world’s atmosphere has below-average atmospheric concentration of water vapor (which has a superlinear function of temperature (at any specific relative humidity).
If relative humidity is decreasing as a result of warming, then the atmosphere should be getting less cloudy. That would make the cloud albedo feedback positive.
That is s good point. I think that there are two competing theories for AGW theory and they are the Sun theory (Svensmark) and the Astronomical Harmonic Climate Model (Ermakov et al.). In both theories the clouds have an important role in amplifying the effects of the original changes. It is a general understanding the the clouds provide the last resort against the overheating of the Earth. The higher temperatures should mean more clouds. The cooling effect of cloudiness is about -0.1 C per 1 % increase in cloudiness.
The last resort is the Planck feedback
I have added an image above illustrating the absorption bands of GH gases in the average global atmosphere. There is also another image showing the effects of increased concentrations of GH gases. According to my calculations, the effects of water are very close to linear showing now decreasing in the elevated concentrations. I add here an image showing the absorption bands of water and CO2 in the tropical climate. Because there is so much water, the portion of CO2 about the total absorption (=GH effect) is only 5.9%, when in the average atmosphere it is about 11 % and in the polar cap area (above 60 latitude) it is about 24 %.
Relative humidity can be decreasing while absolute humidity remains constant. I believe that is more nearly the case. If so, then, the tonnage of water in the atmosphere remains constant. Any excess precipitates out. There is no a priori reason to believe cloud albedo is anything but constant overall.
Why not ?? Earths ‘steady state’ Temperature depends on the total earth albedo. There is NO cloud albedo. Albedo refers to the total earth rejection of incoming solar spectrum radiant energy. It is not some localized reflection coefficient.
Clouds do more than simply reflect some solar spectrum radiant energy, as part of earth’s albedo. Clouds also radiate efficiently at LWIR thermal radiation wavelengths.
Albedo by definition is solar spectrum ‘reflected’ radiant energy only.
I put ‘reflected’ in apostrophes just so you know that it is not really reflected by clouds; it is strongly refracted by ordinary optical processes in near spherical water droplets, so it is scattered by multiple refractions, which results in much back scattering. The original solar photons exit the earth as part of the albedo. If the leaving photons did not arrive from the sun and exit, instead of remaining on earth, they are NOT part of albedo.
G
Relative humidity determines cloud presence more than absolute humidity does. Note cloudiness in tropical and polar regions, where the temperature difference at any given relative humidity means the absolute humidity is different by an order of magnitude or more.
Lower RH, means fewer clouds and more SW absorption by the ocean, which means more warming.
http://www.climate4you.com/images/HadCRUT3%20and%20TropicalCloudCoverISCCP.gif
Interesting. Vaguely complementary.
Why are you plotting global HadCRUT, land+sea “average” temps against tropical cloud. Wouldn’t it be a better examination of the effect to look at tropical SST ?
The problem with Tropical SST is that it is very constant, held constant by evaporation. Temperature is the wrong metric to use for the energy transfer, for the Tropical SST at least.
Measuring temperature and radiation losses is more appropriate for the higher latitudes. Also Clouds increase the albedo in the Tropics but decrease the albedo in the higher latitudes, simply because of the incident angle.
Increasing cloud coverage in the Tropics decreases the amount of energy entering the system and decreases the cloud coverage in the upper latitudes increasing the energy loss (less albedo).
Whereas Decreasing cloud coverage in the Tropics increases the energy absorption and increases the amount of clouds in the upper latitudes slowing the energy loss.
The location and amount of clouds are the Climatic controllers.
That’s all sounding to get rather contorted and hand-wavey. That tropical SST is fairly constant was my point. You suggestion that tropical cloud kinda casts a shadow on higher latitudes but does not affect tropical SST .. Hmmm. I’d want some solid data to convince me of that.
Greg “That’s all sounding to get rather contorted and hand-wavey. That tropical SST is fairly constant was my point. You suggestion that tropical cloud kinda casts a shadow on higher latitudes but does not affect tropical SST .. Hmmm. I’d want some solid data to convince me of that.
No, what I am saying is that tropical clouds are perpendicular to the sun for maximum reflection. High latitude clouds see a ‘lower’ sun and as a result trap a portion of surface reflected radiation. And the extra water vapor from the tropics increases the amount of clouds in the upper latitudes.
Dr Ollila, thank you for your essay, and link to the published paper.
I have carried out a dynamic analysis of the temperature effects caused by the eruption. I wanted to test two options for the climate sensitivity parameter (CSP). The radiative forcing (RF) at the top of the atmosphere has a linear relationship to the global mean surface temperature change dT:
Have you tested your model against “out-of-sample” data?
Sorry but because I am not a native English speaker, I could not understand your question. Could you try other wording?
Have you used the model to predict the outcomes of events that had not happened when you wrote and parameterized the model?
Hello matthewrmarler, I think that now I understood your question. In this Pinatubo study the model is very simple indeed concerning the temperature effects caused by the radiation flux changes. It is mainly a dynamic model, because the input varies along the time.
I have two papers, which you might be interested in, because there are future projections. The first paper I have a model describing the fluxes between the atmosphere, the biosphere and the ocean and there are also are two projections of the CO2 concentration trends for the future:
http://sciencedomain.org/issue/1288
Another paper introduces two competing theories for AGW and the future forecasts of these models:
http://www.scienpress.com/journal_focus.asp?main_id=59&Sub_id=IV&Issue=1564
Dr Ollila. I have had a look at your paper “Cosmic theories of Greenhouse gases…” You say this includes future forecasts for different models – am I correct in thinking that Figure 8, where it says “The black curve is the combined effect of SDI, the Sun and GH gases.”, that this is the future forecast for the model you describe here?
To seaice1. Yes in Fig. 8 there is a future forecast up tio 2050. Why it is ending there? The available data of two major cosmic forces was available only up to 2050.
Dr. Ollila,
Thank you for more information about “Global climate changes as forecast by Goddard Institute for Space Studies three-dimensional model” by Hansen et el, Journal of Geophysical Research, 20 August 1988. It is perhaps the most frequently cited example of a successful test of climate models.
Its skill is evaluated more or less in “Skill and uncertainty in climate models” by Julia C. Hargreaves, WIREs: Climate Change, July/Aug 2010 (ungated copy).
What might we find from a rerun of Hansen’s model using your suggestions? We’ll never know! Hargreaves reported that “efforts to reproduce the original model runs have not yet been successful”. The dog ate the model, so it can’t be rerun with updated observations or different assumptions.
Very sloppy archiving for the project to save the world.
“The dog ate the model”
There is no reason to believing archiving is the problem. Complex codes need maintenance as hardware, software libraries etc change. The original program evolved into GISS Model E, so it was maintained in that sense, but reviving a 30yo version would be a major task. It probably used something like SCSS for code management, and IMSL libraries etc. Getting old versions of those working would be a challenge. And the original people probably won’t be available.
Nick Stokes
Certainly. But how do you solve this part of the “running old code, getting BAD solutions” problem?
In 1986, they thought TSI = 1371 watts/m^2,
In 1988 ACRIM1 had 1367
Then in 1996, the calculation had to be based on TSI = 1365 watts/m^2,
Then in 2010, the calculations had to be based on TSI = 1362 watts/m^2
Now, it the same “robust” calculations keep coming up with the same forcing from a theoretical, fundamental physics basis but while chasing a input heat energy that has dropped by more than 3 times the “net forcing” that has been calculated for a doubling in CO2 by man, how accurate the robust fundamental calculations?
Should they now not be predicting massive cooling of -6 watts/m^2 forcing, since the input energy has dropped by that much during their season of running GCM?
“Should they now not be predicting massive cooling of -6 watts/m^2 forcing”
No. For a start, the forcings are per unit area of Earth surface. So all those numbers need to be divided by four. But solar is part of the forcing scenario. Hansen said in the 1988 paper that there was no evidence to that time to base a forecast variation, so they took it i to be constant. The variations you cite are not actual changes in TSI, but mostly changes in measurement value with the instruments of the time. TSI itself didn’t change from 1371 to 1362 W/m2. The historic variation they now use for CMIP 5 is:
http://data.giss.nasa.gov/modelforce/solar.irradiance/tsi_CMIP5.png
That is a variation of range about 1.5 W/m2, which after dividing by 4 gives a forcing oscillation of about 0.4 W/m2.
Nick,
Thanks for the additional color on this. I’ve twice had IT run old code (once for an audit, once for an arbitration), the latter about 30 years old. Neither was a big project, but the context was in a large corp — not a science institution.
I’ll ask my one of my cyber co-authors for an opinion on this.
Nick,
I got an answer.
The team I worked with were skilled, had little turnover, and a consistent systems environment.
Still, if we’re talking about saving the world this looks like an important project with trivial costs. What’s the cost of allowing the current gridlock to continue for another 28 years?
A common attitude seems to be trillions for prevention and mitigation but not a cent for testing. Penny wise and all that.
As I described earlier, there are three studies on the RF effect of CO2. Hansen et al. applied a model but Myhre et al. calculated the RF effects for three different atmospheric profiles (tropical, NH, SH) for the RF effects for the change from 280 ppm upwards. Shi applied broad band calculations. My method is actually the same as used by Myhre et al.
I’ve just plotted the NCEP monthly area weighted data with a 12mo low-pass filter.
It looks a bit different to figure 1.
http://climategrog.files.wordpress.com/2016/03/ncep_rh1.png
Higher more rarified altitudes show marked drop in RH until 1977 ( the end of cooling ) , then rise to about 1998. Mid troposphere seems to generally counter this.
The lowest level in this graph : 600, 700, 850mb show a slight increase in the last 15-20 years. Beginning of cooling ? Maybe this data has not been ‘corrected’ yet.
Actually this looks very different from figure 1. Is that the area weighting or partly the fact that I low-pass filtered the monthly data instead of using their ‘seasonal averages’.
Perhaps Dr. Ollila could clarify what he plotted.
850 and 925 mb levels of NCEP data
Sorry to coming so late to comment this issue. Partially the delay is due to my different time zone.
I did not check the properly the latest version of Figure 1, because I did not used these values anywhere in my calculations otherwise than the other water content graph. Figure 1 was used only to show that even by eye it is possible to see that the assumption of the constant RH is not correct. I created this figure first time for my paper published in 2013. Now after your comments I noticed that the data has been updated. The original data set can be found in this address:
http://www.esrl.noaa.gov/psd/cgi-bin/data/timeseries/timeseries1.pl
The data set is NCEP Reanalysis published by NOAA. I have used the global values with seasonal average and area weight grids. The version in my story was based on the earlier version. The newest version is from 13th of May 2014. The climatology in the newest version is 1981-2010, when it was in the older version 1971-2000. The main difference seems to be that the new values are about 2 % on the higher level. Anyway there are the same downward trends as in the older version. Here is a figure with older and newer graphs:
The earlier data in that sequence, particularly, is to be taken with scepticism. Not the fault of NCEP – there just isn’t much reliable data. Here is the note that goes with that data:
“The humidity analysis is believed to be the weakest of the primary atmospheric analyses; i.e., Z, T, U, V, and Q. The other primary variables (Z, T, U and V) and their gradients have to be internally consistent. One can not change T without changing U, V, and Z. As a result, the internal consistency provides a check and constraints on these fields. The humidity analysis, however, is unconstrained (except for q <= q-sat) and is effectively produced by a univariate analysis with no dynamical constraint on the gradients. Then there is the sampling problem. The humidity has many small-scale features and a single measurement may not be representative of a grid box average.
By the way, did I mention instrument errors?"
Thanks for some very informed input on this discussion. A rarity here. ( As on most blogs )
There has been a lot on RH recently and when I read that at NCEP, I realised the whole thing was pretty hopeless. I posted those graphs because they seem at odds with the starting point of this post : figure 1 “sight test”.
Dr. Ollila initially gave a bad link for the origin of what he plotted and he still has not said what he actually plotted and it was not marked on his graph. This is rather unsatisfactory.
It seems that he could have plotted a non area weighted average and thus heavity skewed the data towards changes in the polar regions.
Can we can hope that the data is not complete garbage or the result of model assumptions in the NCEP reanalysis ( which is not to be ruled out )?
I
1. Re quantitative importance of water vapor for the Greenhouse Effect: From infrared (IR) spectra of outgoing radiation as seen by satellites looking down on a cloudless Earth, water vapor is roughly twice as important as CO2. Sample spectra are found in Grant W. Petty’s excellent book “A First Course in Atmospheric Radiation, Second Edition” and Fig. 3 at http://climateaudit.org/?p=2572 .
2. However, water vapor does not provide a 200% positive feedback because doubling CO2 does not double water vapor. Increasing temperature from 15.0 to 15.6 Celsius increases saturated water vapour from 12.788 to 13.290 mm Hg [Handbook of Chemistry and Physics], an increase by a factor of 1.039, or by 3.9%. Multiplying by a weighting factor of 2 would increase this only to 8%, not 200% (the main absorption lines of both CO2 and water vapor are highly saturated). This factor of 8% increase would also apply at 50% relative humidity instead of 100%. For any starting temperature from 40.0 to -10.0 Celsius, an increase by 0.6 Celsius increases water vapor by only 3.2% to 5%. This relatively constant % increase results from the fact that saturated water vapor is roughly an exponential function of absolute temperature, and exp[K(T+0.6)] = exp(KT).exp(0.6K) = constant.exp(KT) for any T, since K and therefore 0.6K and exp(0.6K) would both be constants.
3. I have used 0.6 K instead of 1 K for the climate sensitivity on doubling CO2 (not including feedbacks) because the main gases of the atmosphere (N2, O2, Ar) are nonpolar molecules that cannot and do not emit any significant infrared (IR) radiation. Therefore the outgoing TOA (Top Of the Atmosphere) IR is not radiated from a black body surface at 10 km (or 5.5 km) altitude, but is better understood as that part of the black body IR emitted from the 15.0 C (288.2 K) solid and liquid surface of the Earth which is NOT absorbed by greenhouse gases such as CO2, water vapor, ozone, etc. The Stefan-Boltzmann law says that at 288.2 K and emissivity 0.98, the surface emits 383.34 W/m^2. Adding a radiative forcing of 3.7 W/m^2 which must be emitted to compensate at steady state for increased absorption on doubling CO2, the warmer Earth must emit 387.04 W/m^2. Using the Stefan-Boltzmann law backwards, this corresponds at emissivity 0.98 to a new temperature of 288.893 K. This gives 288.893 – 288.2 = 0.693 degrees for the temperature sensitivity not including feedbacks. However, Jack Barrett at his excellent website http://www.barrettbellamyclimate.com/ (see the section “The hard bit”) has rerun the MODTRAN simulated IR spectrum to 70 km altitude (not the 20 km assumed to be the TOA). The results show slightly increasing CO2 emission from the stratosphere on doubling CO2 (because of the temperature inversion caused by the absorption of incoming UV and visible Solar radiation by ozone, the photons escaping at central CO2 frequencies come from a higher altitude, i.e. at a higher temperature), equivalent to about 0.11 degrees. So I have rounded off the climate sensitivity not including feedbacks to 0.693 – 0.11 = 0.6 degrees.
4. In addition, increasing water vapor is likely to increase cloud cover, which provides a net negative feedback. Assuming this negative feedback only just cancels (not overwhelms) the 8% positive feedback due to increased water vapor absorption, the climate sensitivity INCLUDING WATER VAPOR AND CLOUD FEEDBACKS is likely to be only about 0.6 degrees, not the 3 degrees assumed in the literature and by the IPCC. This explains why there has been an 18 year hiatus/pause in global warming, even as CO2 continued to increase recently. Increasing CO2, manmade or natural, is not anywhere the danger as previously thought, and limited financial and manpower resources ought to be used for other problems.
5. The literature also got the mechanism of the greenhouse effect wrong. The troposphere does not consist of spherical shells of decreasing temperature at increasing altitude, each emitting a Planck black body spectrum until finally “the IR photons escape to outer space at 10 km”. At 10 km, the Stefan-Boltzmann law at 220 K predicts an emission of only 130 W/m^2, nowhere near the observed 240 W/m^2. Calculating an altitude of 5.5 km, where the temperature is 256 K which would correspond to emission of 240 W/m^2, assumes emission from a Planck black body shell, and the observed spectra simply do not follow Planck curves. Instead, the observed spectra are net ABSORPTION spectra (NOT EMISSION spectra), where some of the photons emitted by the 288.2 K black body surface are absorbed at CO2, water vapor and ozone frequencies. That energy is not simply re-emitted as photons to the next layer higher up, but is transferred during non-radiative collisions to the main molecules of the troposphere (N2, O2, Ar) which make up 99.9% of dry air and do not emit any significant IR. The energy ends up increasing the translational and rotational energies of these molecules which outnumber CO2 by 2500:1. I.e. the troposphere warms up, the greenhouse effect. Even Petty’s book title betrays lack of understanding, since it includes “Atmospheric Radiation” [i.e. Emission, not Absorption].
If the result and theory is correct, there are usually more than one way to get the same results. In my original paper I have used two tools (Spectral Calculator and MODTRAN) and two different methods plus the traditional energy balance equation with a pen and paper. All these methods come to the CS of 0.27….0.3 K.
Relevant links:
Richard Telford – The Peer Review of Ollila (2016
Scholarly Open Access – Finnish Man Uses Easy Open-Access Journals to Publish Junk Climate Science
ScienceDomain Reviewer 1st round comment
cienceDomain Reviewer Final Evaluation
Predatory journal. $500 bucks and you too can be a peer-reviewed published scientist.
It might be useless to start the show my replies to every comment of reviewers. I would rather ask, what do you think about those two papers, which have been referred to be solid evidence about positive water feedback during the Pinatubo eruption? There are two points, which make the analysis of positive water feedback to work: 1) selection of a lower solar irradiation decrease -4.5 W/m2 instead of -6.0 W/m2 used by any other researchers, 2) using the humidity decrease in a very “creative” way, which has turned the trend upside down. As usually this review process included relevant questions but also some, which were not so relevant. For example the main reviewer wanted to disapprove the paper, because I had used also UAH MSU temperature as a reference beside the average temperate of 4 temperature data sets. According to the reviewer, UAH MSU is not a surface temperature. If we are accurate as the Sun, it is a correct comment. But the authors of the paper of Soden et al. had also used UAH MSU temperature and only it.This happened during those days, when there was no pause in the temperature and trends of UAH MSU were close to other data sets. This is of course my personal opinion but I would be ready to give my paper for a re-review together with the paper of Soden et al. It would be interesting to see the results.
It should be noted that the revieweer who did not want to show his name seems to be John Abraham
He is frequent contributor at the “unreliable” alarmist web site SkS
https://www.skepticalscience.com/posts.php?u=2502
and frequent contributor to the fact-distorting, alarmist newspaper web site of The Guardian, where every climate article now seems to carry the subtitle : ” Climate Consensus – the 97%”
http://www.theguardian.com/profile/john-abraham
He is qualified to raise pertinent criticisms but seems to be an activist as much as scientist.
The UAH issue seems to be a key plank in the SkS and warmist agenda to discredit datasets that don’t show the “right” message. I guess RSS is now OK, now that they don’t correct for diurnal drift any more.
I have sent the following reply to the web page of Telford:
Yes, in the open peer review system the possible weaknesses are readily available. I admit that three of the reviewers gave only formal comments. The first reviewer was very detailed in his/her comments trying to turn every stone to find out reasons to reject the paper. It was very obvious what is the attitude of the reviewer based on this comment: “On line 175 the author(s) claim that prior researchers included a positive water vapor feedback. There really is no doubt about this.” In the section “Results and discussion” I have a summary showing that a former paper of Soden et al. (it is called a solid evidence about the positive water feedback) is based on the proper selection of flux changes not used by the other researchers. Soden et al. have used in a very “creative” way the humidity trends in order to turn upside down the trend. I am ready to give the paper of mine to any review process together with the paper of Soden et al. It would be interesting to see the results.
Yes, I admit that I visit sauna three times a week. But you are wrong about the money. I did not pay $500 but only $50 from my own pockets. What we should think about those papers published with the fee of $2000 – $3000?
My original profession is not a civil engineer. I used to be a process and automation engineer. You may say that I have no idea about the climatology. At least I know what is the water content of the average global atmosphere. Kiehl & Trenberth used the US Standard Atmosphere 76 in calculating the portion of CO2 in the GH phenomenon and getting the result of 26 %. Have you any idea what is wrong with this atmosphere?
Dr. Antero Ollila
http://multi-science.atypon.com/doi/pdf/10.1260/0958-305X.25.8.1439
Each method shows that, on average, water vapour contributes approximately 96% of current greenhouse gas warming. Thus, the factors controlling the amount of water vapour in the air also control the earth’s temperature.
TOTAL BACK RADIATION OF ALL GHG Figure 7 is FAQ 1.1 Figure 1 from page 96 of AR4. It shows the radiation balance for the earth and that the back radiation of all of the greenhouse gases is 324 W m-2. This is the value used to calculate the RF [radiative forcing] of CO2 at 378 ppmv as (8.67/324)/100 = 2.7% back radiation of the total of all of the greenhouse gases.
From Table 1, CO2 accounts for 2.7% of the global warming while all of the other gases account for approximately 0.7% for a total of approximately 3.4%. It becomes evident that, on average, water vapour accounts for approximately 96% of the current global [greenhouse effect] warming. This is an important finding because it leads to the conclusion that the factors controlling the average level of water vapour in the atmosphere also control atmospheric temperature.
[O]n average, each molecule of CO2 is surrounded by approximately 23 molecules of water vapour at ground level. … If the warming effect of water molecules and CO2 molecules were the same, then the contribution of CO2 would be (1/22.7) = 4.4% of that of water vapour. But from the previous section, water molecules are 1.6 times more effective at warming than CO2 molecules. Using this value and the ratio of 22.7:1, the contribution of CO2 to warming of the atmosphere is approximately (1/22.7)/1.6 = 2.8% of that of water vapour. As water vapour is approximately 96% of the total RF of all of the GHG, the contribution of CO2 is approximately 4% less than this, i.e., 2.69%. If the average RH were 60%, the contribution of CO2 would be ((1/27.4)/1.32) x 0.96 = 2.65%. For practical purposes, these values are the same as the 2.7% obtained by the quadratic model.
The increased absorption band areas are directly related to the warming effects of each GH gas. As far as I know the spectral analysis is the only way to find out the small changes of the absorption bands. The absorption / emission processes at the molecular level are so complicated that they cannot be calculated accurately enough with no other methods. I think that many people do not know that GCM’s do not apply spectral analyses in their calculations but they utilize results of real spectral analyses. By the way the absorption spectra in HITRAN database has been checked in the real climate conditions and the error is less than 1 %.
Yes, that’s the real physics part. The rest fluff and frig-factors typically misrepresented as “known, basic physics” .
Lower RH at higher altitudes could be a sign that CO2 causes cooling, not warming.
I wrote a tongue-in-cheek post about CO2 causing global cooling at http://wattsupwiththat.com/2016/01/08/how-thunderstorms-beat-the-heat/ and everybody scoffed. It predicted cooling caused by a greater cloud cover. Cloud cover is increased because more CO2 in the lower levels reduces the IR flux to the higher levels. That means less absorption by water at higher levels, cooler water and more condensation (clouds).
What I didn’t mention was that it predicts lower cloud levels. If the enthalpy of the atmosphere is greater very near the surface and lesser above then it will lower cloud levels and increase condensation. Lower cloud levels mean less absolute humidity at higher levels. There are a number of possible explanations why this would mean less relative humidity too.
I’m stickin’ by my guns. CO2 causes cooling and it’s because water condenses at lower temperatures than CO2. More CO2 absorbing the IR below, less IR reaching the water vapor above. Using the climatist assumptions about IR radiation, that means more evaporation of water below and more condensation of water above. That in turn means higher albedo and more insolation reflected back to space. That means global cooling.
“I’m stickin’ by my guns. CO2 causes cooling and it’s because water condenses at lower temperatures than CO2. ”
That sounds like negative feedback , which will reduce or negate GHG effect of CO”, it will not produce actual cooling.
http://www.sdiarticle2.org/prh/PSIJ_33/2016/Rev_PSIJ_23242_Joh_ANON_v1.pdf
Many serious criticisms of this paper by the main reviewer seem to have gone unaddressed.
University of St Thomas? John Abraham?
see above, warmist activist:
https://www.skepticalscience.com/posts.php?u=2502
http://www.theguardian.com/profile/john-abraham
It is not necessary to have the LW component explicitly . Just use ERBE total flux in both directions. The difference is the energy budget anomaly which comprises both the aerosol effect and the climate feedback.
That was the method suggested Dr Roy Spencer which I used in my study ( link to Spencer’s blog included.).
I have replied to each comment of every reviewer but it looks like that the magazine does not publish the author’s comments.
That is certainly not the impression that we get from review documents posted . Maybe you should ask for a refund 😉
True, it is misleading if you made full replies.
If there is a lag of six months in the aerosol spreading , I don’t see how you can take that is a globally valid figure. Most of the energy comes into the system in the tropics. There the spread happened within a month or two.
There is detailed gridded AOD data for four different atmospheric heights available here:
http://data.giss.nasa.gov/modelforce/strataer
That had peaked already and dropped to about 50% within 6mo.
Using just the tropics, I found ( as did Spencer et al ) a lag of about 12-13mo in the radiative effects ( reflecting the surrface climate reaction ) . That corresponds to a time constant of about 8 mo.
If you are effectively subtracting 6 months from that as the time it takes to reach high latitudes that will not represent the majority of the disruption to the energy budget and will seriously skew any estimation of time constant.
It is unclear from figure 1 in the paper when the minimum on UAH happens since there is still a strong 6 monthly residual in your anomaly data, though I would estimate 1992.8 by eye, you need to determine this better.
That would also agree my and Spencers lag correlation for the tropics of 12 to 13 mo.
AOD and ERBE data outside the tropics are patchy. Aerosol spreading involves notable delays and global data sets averaging land and SST are or questionable validity for energy budget calculations and will in any case bias towards the land of NH .
https://judithcurry.com/2016/02/10/are-land-sea-temperature-averages-meaningful/
I would suggest focussing on tropics where data is more continuous, physically consistent and where the ‘action’ is in terms of energy budget.
https://climategrog.wordpress.com/2015/01/17/on-determination-of-tropical-feedbacks/
I have defined the dead time from the global UAH MSU monthly temperature, and it is quite easy to find out, when the global temperature starts to decline; that delay is 1.6 months. If you look at the measured temperature and the simulated temperatures, they follow pretty well each other.
[Oops. Misplaced this comment above]
It is not necessary to have the LW component explicitly . Just use ERBE total flux in both directions. The difference is the energy budget anomaly which comprises both the aerosol effect and the climate feedback.
That was the method suggested Dr Roy Spencer which I used in my study ( link to Spencer’s blog included.).
This is a quote from my study: “There has been a special GEWEX project to assess the surface radiation budget data sets based on the available data at the top of the atmosphere (TOA). By studying the GEWEX results, the author’s conclusion is that the LWDN fluxes could not be estimated reliably in this project based on the other existing flux data.” The ERBE flux is the easiest substitute for the real LWDN flux.
Thanks for the reply. I saw that.
You may like to look at the ERBE in more detail. There was some very silly assumptions made about there being constant meteorology throughout the day in the tropics ! Obviously by someone who’d never been there.
This led to a massive alias of the diurnal cycle that comes out at 36d in the tropics ( 72d outside that ). This affects the upward SW : hence the derivative LW flux.
https://climategrog.files.wordpress.com/2015/02
Pretty ain’t it ?
It was pointed out by Trenberth that taking the usual monthly averages would result is a circa 6mo alias in the data. ( 198d I think is more accurate ). This is
why ERBE only provide 36d and 72d datasets now. That avoids the alias but there may be some residual oddities. At some stage they provided monthly so be careful to avoid that.
It will probably be necessary to fit a spline of do a Cat-Mull interpolation if you need monthly data.
I got the pretty picture by extracting the daily data. It took a fair bit of effort adapting some FORTRAN code but it was informative.
BTW the red line you can just see popping above the others is the Pinatubo event.
Regarding climate sensitivity to the Pinatubo eruption, and Earth’s climate sensitivity: The Pinatubo eruption’s cooling and the warming after its effects faded occurred over only a few years, which is period much too short to get the atmosphere in equilibrium with the part of the ocean that is above the thermocline. To do that requires decades. The .27 degree per W/m^2 that is indicated above sounds to me like somewhat over half the equilibrium climate sensitivity, which I think is around .35-.4 degree per W/m^2.
(I expect a lower figure than IPCC does because positive cloud albedo feedback requires water vapor feedback less than that of constant relative humidity. The zero-feedback except Planck figure is about .33 degree per W/cm^2.)
I have used in this paper the time constant of 2.74 months for the surface ocean (=mixing layer of 75 m in depth). The temperature below this depth has not time to react for this kind of a rapid change. The model calculated temperature response of the dynamic model follows very well the real temperature change proving that the time constants of land and sea are correct. The differences in using CSP 0.27 versus 0.5 are very big.