Climate models do not justify even precautionary action

Unfor­tu­nately, the com­pan­ion arti­cle pub­lished by Scep­tic Mag­a­zine, in sup­port of the AGW propo­si­tions offers lit­tle more than a rep­e­ti­tion of well-knownIPCC assertions.

those who advo­cate extreme poli­cies to reduce car­bon diox­ide emis­sions inevitably base their case on GCM pro­jec­tions, which some­how become real pre­dic­tions in pub­lic­ity releases. But even if these advo­cates admit­ted the uncer­tainty of their pre­dic­tions, they might still invoke the Pre­cau­tion­ary Prin­ci­ple and call for extreme reduc­tions ‘just to be safe’. This prin­ci­ple says, ‘Where there are threats of seri­ous or irre­versible dam­age, lack of full sci­en­tific cer­tainty shall not be used as a rea­son for post­pon­ing cost-effective mea­sures to pre­vent envi­ron­men­tal degra­da­tion.’ That is, even if we don’t fully know that CO2 is dan­ger­ously warm­ing Earth cli­mate, we should cur­tail its emis­sion any­way, just in case.

How­ever, if the present uncer­tainty limit in Gen­eral Cir­cu­la­tion Mod­els is at least 100 degrees per cen­tury, we are left in total igno­rance about the tem­per­a­ture effect of increas­ing CO2. It’s not that we, ‘lack­full sci­en­tific cer­tainty,’ it’s that we lack any sci­en­tific cer­tainty. We lit­er­ally don’t know whether dou­bling atmos­pheric CO2 will have any dis­cernible effect on cli­mate at all.“Skep­tic Mag­a­zine, Patrick Frank

Frank’s arti­cle is well-written, well-documented and brim-full of insight into the lim­its of GCM mod­el­ing. He writes with­out jar­gon and with no more alge­bra than is absolutely nec­es­sary: remark­ably lit­tle as it turns out­for rea­sons that are illu­mi­nat­ing in themselves.

Unfor­tu­nately, the com­pan­ion arti­cle pub­lished by Scep­tic Mag­a­zine, in sup­port of the AGW propo­si­tions offers lit­tle more than a rep­e­ti­tion of well-known IPCC assertions.

those who advo­cate extreme poli­cies to reduce car­bon diox­ide emis­sions inevitably base their case on GCM pro­jec­tions, which some­how become real pre­dic­tions in pub­lic­ity releases. But even if these advo­cates admit­ted the uncer­tainty of their pre­dic­tions, they might still invoke the Pre­cau­tion­ary Prin­ci­ple and call for extreme reduc­tions ‘just to be safe’. This prin­ci­ple says, ‘Where there are threats of seri­ous or irre­versible dam­age, lack of full sci­en­tific cer­tainty shall not be used as a rea­son for post­pon­ing cost-effective mea­sures to pre­vent envi­ron­men­tal degra­da­tion.’ That is, even if we don’t fully know that CO2 is dan­ger­ously warm­ing Earth cli­mate, we should cur­tail its emis­sion any­way, just in case.

How­ever, if the present uncer­tainty limit in Gen­eral Cir­cu­la­tion Mod­els is at least 100 degrees per cen­tury, we are left in total igno­rance about the tem­per­a­ture effect of increas­ing CO2. It’s not that we, ‘lack­full sci­en­tific cer­tainty,’ it’s that we lack any sci­en­tific cer­tainty. We lit­er­ally don’t know whether dou­bling atmos­pheric CO2 will have any dis­cernible effect on cli­mate at all.“Skep­tic Mag­a­zine, Patrick Frank

Frank’s arti­cle is well-written, well-documented and brim-full of insight into the lim­its of GCM mod­el­ing. He writes with­out jar­gon and with no more alge­bra than is absolutely nec­es­sary: remark­ably lit­tle as it turns out­for rea­sons that are illu­mi­nat­ing in themselves.

Unfor­tu­nately, the com­pan­ion arti­cle pub­lished by Scep­tic Mag­a­zine, in sup­port of the AGW propo­si­tions offers lit­tle more than a rep­e­ti­tion of well-known IPCC assertions.

those who advo­cate extreme poli­cies to reduce car­bon diox­ide emis­sions inevitably base their case on GCM pro­jec­tions, which some­how become real pre­dic­tions in pub­lic­ity releases. But even if these advo­cates admit­ted the uncer­tainty of their pre­dic­tions, they might still invoke the Pre­cau­tion­ary Prin­ci­ple and call for extreme reduc­tions ‘just to be safe’. This prin­ci­ple says, ‘Where there are threats of seri­ous or irre­versible dam­age, lack of full sci­en­tific cer­tainty shall not be used as a rea­son for post­pon­ing cost-effective mea­sures to pre­vent envi­ron­men­tal degra­da­tion.’ That is, even if we don’t fully know that CO2 is dan­ger­ously warm­ing Earth cli­mate, we should cur­tail its emis­sion any­way, just in case.

How­ever, if the present uncer­tainty limit in Gen­eral Cir­cu­la­tion Mod­els is at least 100 degrees per cen­tury, we are left in total igno­rance about the tem­per­a­ture effect of increas­ing CO2. It’s not that we, ‘lack­full sci­en­tific cer­tainty,’ it’s that we lack any sci­en­tific cer­tainty. We lit­er­ally don’t know whether dou­bling atmos­pheric CO2 will have any dis­cernible effect on cli­mate at all.“Skep­tic Magazine,Patrick Frank

Frank’s arti­cle is well-written, well-documented and brim-full of insight into the lim­its of GCM mod­el­ing. He writes with­out jar­gon and with no more alge­bra than is absolutely nec­es­sary: remark­ably lit­tle as it turns out­for rea­sons that are illu­mi­nat­ing in themselves.

Unfor­tu­nately, the com­pan­ion arti­cle pub­lished by Scep­tic Mag­a­zine, in sup­port of the AGW propo­si­tions offers lit­tle more than a rep­e­ti­tion of well-known IPCC assertions.

…those who advo­cate extreme poli­cies to reduce car­bon diox­ide emis­sions inevitably base their case on GCM pro­jec­tions, which some­how become real pre­dic­tions in pub­lic­ity releases. But even if these advo­cates admit­ted the uncer­tainty of their pre­dic­tions, they might still invoke the Pre­cau­tion­ary Prin­ci­ple and call for extreme reduc­tions ‘just to be safe’. This prin­ci­ple says, ‘Where there are threats of seri­ous or irre­versible dam­age, lack of full sci­en­tific cer­tainty shall not be used as a rea­son for post­pon­ing cost-effective mea­sures to pre­vent envi­ron­men­tal degra­da­tion.’ That is, even if we don’t fully know that CO2 is dan­ger­ously warm­ing Earth cli­mate, we should cur­tail its emis­sion any­way, just in case.

How­ever, if the present uncer­tainty limit in Gen­eral Cir­cu­la­tion Mod­els is at least ±100 degrees per cen­tury, we are left in total igno­rance about the tem­per­a­ture effect of increas­ing CO2. It’s not that we, ‘lack…full sci­en­tific cer­tainty,’ it’s that we lack any sci­en­tific cer­tainty. We lit­er­ally don’t know whether dou­bling atmos­pheric CO2 will have any dis­cernible effect on cli­mate at all.”  Skep­tic Magazine—Patrick Frank

Frank’s arti­cle is well-written, well-documented and brim-full of insight into the lim­its of GCM mod­el­ing. He writes with­out jar­gon and with no more alge­bra than is absolutely nec­es­sary: remark­ably lit­tle as it turns out—for rea­sons that are illu­mi­nat­ing in themselves.

Unfor­tu­nately, the com­pan­ion arti­cle pub­lished by Scep­tic Mag­a­zine, in sup­port of the AGW propo­si­tions offers lit­tle more than a rep­e­ti­tion of well-known IPCC assertions.

…those who advo­cate extreme poli­cies to reduce car­bon diox­ide emis­sions inevitably base their case on GCM pro­jec­tions, which some­how become real pre­dic­tions in pub­lic­ity releases. But even if these advo­cates admit­ted the uncer­tainty of their pre dic­tions, they might still invoke the Pre­cau­tion­ary Prin­ci­ple and call for extreme reduc­tions ‘just to be safe’. This prin­ci­ple says, ‘Where there are threats of seri­ous or irre­versible dam­age, lack of full sci­en­tific cer­tainty shall not be used as a rea­son for post­pon­ing cost-effective mea­sures to pre­vent envi­ron­men­tal degra­da­tion.’ That is, even if we don’t fully know that CO2 is dan­ger­ously warm­ing Earth cli­mate, we should cur­tail its emis­sion any­way, just in case.

How­ever, if the present uncer­tainty limit in Gen­eral Cir­cu­la­tion Mod­els is at least ±100 degrees per cen­tury, we are left in total igno­rance about the tem­per­a­ture effect of increas­ing CO2. It’s not that we, ‘lack…full sci­en­tific cer­tainty,’ it’s that we lack any sci­en­tific cer­tainty. We lit­er­ally don’t know whether dou­bling atmos­pheric CO2 will have any dis­cernible effect on cli­mate at all.”  Skep­tic Magazine—Patrick Frank

Frank’s arti­cle is well-written, well-documented and brim-full of insight into the lim­its of GCM mod­el­ing. He writes with­out jar­gon and with no more alge­bra than is absolutely nec­es­sary: remark­ably lit­tle as it turns out—for rea­sons that are illu­mi­nat­ing in themselves.

Unfor­tu­nately, the com­pan­ion arti­cle pub­lished by Scep­tic Mag­a­zine, in sup­port of the AGW propo­si­tions offers lit­tle more than a rep­e­ti­tion of well-known IPCC assertions.

…those who advo­cate extreme poli­cies to reduce car­bon diox­ide emis­sions inevitably base their case on GCM pro­jec­tions, which some­how become real pre­dic­tions in pub­lic­ity releases. But even if these advo­cates admit­ted the uncer­tainty of their pre­dic­tions, they might still invoke the Pre­cau­tion­ary Prin­ci­ple and call for extreme reduc­tions ‘just to be safe’. This prin­ci­ple says, ‘Where there are threats of seri­ous or irre­versible dam­age, lack of full sci­en­tific cer­tainty shall not be used as a rea­son for post­pon­ing cost-effective mea­sures to pre­vent envi­ron­men­tal degra­da­tion.’ That is, even if we don’t fully know that CO2 is dan­ger­ously warm­ing Earth cli­mate, we should cur­tail its emis­sion any­way, just in case.

How­ever, if the present uncer­tainty limit in Gen­eral Cir­cu­la­tion Mod­els is at least ±100 degrees per cen­tury, we are left in total igno­rance about the tem­per­a­ture effect of increas­ing CO2. It’s not that we, ‘lack…full sci­en­tific cer­tainty,’ it’s that we lack any sci­en­tific cer­tainty. We lit­er­ally don’t know whether dou­bling atmos­pheric CO2 will have any dis­cernible effect on cli­mate at all.”  Skep­tic Magazine—Patrick Frank

Frank’s arti­cle is well-written, well-documented and brim-full of insight into the lim­its of GCM mod­el­ing. He writes with­out jar­gon and with no more alge­bra than is absolutely nec­es­sary: remark­ably lit­tle as it turns out—for rea­sons that are illu­mi­nat­ing in themselves.

Unfor­tu­nately, the com­pan­ion arti­cle pub­lished by Scep­tic Mag­a­zine, in sup­port of the AGW propo­si­tions offers lit­tle more than a rep­e­ti­tion of well-known IPCC assertions.

…those who advo­cate extreme poli­cies to reduce car­bon diox­ide emis­sions inevitably base their case on GCM pro­jec­tions, which some­how become real pre­dic­tions in pub­lic­ity releases. But even if these advo­cates admit­ted the uncer­tainty of their pre­dic­tions, they might still invoke the Pre­cau­tion­ary Prin­ci­ple and call for extreme reduc­tions ‘just to be safe’. This prin­ci­ple says, ‘Where there are threats of seri­ous or irre­versible dam­age, lack of full sci­en­tific cer­tainty shall not be used as a rea­son for post­pon­ing cost-effective mea­sures to pre­vent envi­ron­men­tal degra­da­tion.’ That is, even if we don’t fully know that CO2 is dan­ger­ously warm­ing Earth cli­mate, we should cur­tail its emis­sion any­way, just in case.

How­ever, if the present uncer­tainty limit in Gen­eral Cir­cu­la­tion Mod­els is at least ±100 degrees per cen­tury, we are left in total igno­rance about the tem­per­a­ture effect of increas­ing CO2. It’s not that we, ‘lack…full sci­en­tific cer­tainty,’ it’s that we lack any sci­en­tific cer­tainty. We lit­er­ally don’t know whether dou­bling atmos­pheric CO2 will have any dis­cernible effect on cli­mate at all.”  Skep­tic Magazine—Patrick Frank

Frank’s arti­cle is well-written, well-documented and brim-full of insight into the lim­its of GCM mod­el­ing. He writes with­out jar­gon and with no more alge­bra than is absolutely nec­es­sary: remark­ably lit­tle as it turns out—for rea­sons that are illu­mi­nat­ing in themselves.

Unfor­tu­nately, the com­pan­ion arti­cle pub­lished by Scep­tic Mag­a­zine, in sup­port of the AGW propo­si­tions offers lit­tle more than a rep­e­ti­tion of well-known IPCC assertions.

8 Comments

  • Mitchell Porter wrote:

    Frank’s arti­cle has been briefly dis­cussed on at least two blogs by cli­ma­tol­o­gists, and his argu­ment dis­missed as invalid. They say that in order to derive his cumu­la­tive error (”±100 degrees per cen­tury”), he makes math­e­mat­i­cally erro­neous assump­tions about the GCMs. It is going to require a lit­tle more study on my part to truly grasp the prob­lem, but it seems to involve, in part, treat­ing a num­ber derived from aver­ag­ing over many sim­u­la­tion runs as if it were the prop­erty of an indi­vid­ual sim­u­la­tion run.

    http://www.realclimate.org/index.php/archives/2008/05/what-the-ipcc-models-really-say/#comment-86894

    http://initforthegold.blogspot.com/2008/03/why-is-climate-so-stable.html?showComment=1210739520000#c8782544473864969372

    Since the arti­cle is gain­ing a lit­tle atten­tion, I imag­ine some­one will write a more thor­ough rebut­tal at some point.

  • Thanks for the ref­er­ences Mitchell. I read the debate between Frank and Tobis care­fully. I can’t say that I’m able to adju­di­cate on the tech­ni­cal issues of cli­mate mod­el­ing. But I’ve had a lot to do with eco­nomic mod­el­ing over the years (Tobis has a low opin­ion about eco­nomic ratio­nal­ity, by the way). If that expe­ri­ence is rel­e­vant then, read­ing between the lines, I’d say that Tobis is evad­ing the issues.

    It seems to me that he dis­misses Frank’s argu­ment about the rel­e­vance of uncer­tain­ties in ini­tial con­di­tions (iter­ated through a model) with­out actu­ally answer­ing it. His response to Frank is really ad hominem: ‘You don’t under­stand cli­mate … what you say holds for weather i.e. instan­ta­neous pre­dic­tion, but not for cli­mate. Your claims would be con­sid­ered naive by my colleagues/fraternity etc etc’.

    I’ve always found that sort of argu­ment uncon­vinc­ing. If an eco­nomic mod­eler was unable to respond to crit­i­cisms of model mech­a­nisms except by say­ing, ‘it’s not done that way’, or ‘that’s not the issue we’re inter­ested in’—as they some­times did— I started to pay much less atten­tion. I found that after that sort of response the debate becomes ‘hieratic’.

    On the whole I think Pat Frank makes a very plau­si­ble argu­ment. But I rec­og­nize that it’s impossible—for now, at least—to ‘carry the day’ in this debate. There are too few phys­i­cal facts/mechanisms and too much ques­tion­able use of statistics.

  • Mitchell Porter wrote:

    Frank’s crit­ics are unan­i­mous that his main “result” — the spec­tac­u­lar ampli­fi­ca­tion of error — is purely an arte­fact of his approx­i­ma­tion and has noth­ing to do with the behav­ior of the GCMs he is discussing.

    Here’s what I gather is happening.

    A GCM has a large num­ber of adjustable para­me­ters. To get these tem­per­a­ture trends, the para­me­ters are set, a large num­ber of indi­vid­ual runs are per­formed, and then they are sta­tis­ti­cally aggre­gated. So you’ll get a mean and a vari­ance. The vari­ance is appar­ently con­stant over time.

    Frank takes this model out­put and makes it the input to some­thing much sim­pler, a lin­ear fit to the evo­lu­tion of the mean tem­per­a­ture as pre­dicted by the GCMs. Unsur­pris­ingly, errors in the ini­tial con­di­tions of his lin­ear equa­tion then grow lin­early with time. But this demon­strates absolutely noth­ing about the GCMs. What he (or some­one) needs to do is to look at what uncer­tainty in those model para­me­ters does to the vari­ance of the pre­dicted temperature.

  • Mitchell Porter wrote:

    P.S. I see what you mean about Tobis vs eco­nom­ics, though (I’m just view­ing his other posts).

  • Pat Frank wrote:

    Mitchell wrote, in #1, “…[the error] seems to involve,… treat­ing a num­ber derived from aver­ag­ing over many sim­u­la­tion runs as if it were the prop­erty of an indi­vid­ual sim­u­la­tion run.

    In the first place, the SRES sim­u­la­tions on which the error oper­ated are com­posedof aver­aged mul­ti­ple GCM runs. So, the quoted objec­tion is out of place.

    Sec­ondly, indi­vid­ual GCM runs exhibit far more error than do aver­ages of many runs and so, if any­thing, the uncer­tainty in the long-range pre­dic­tions of indi­vid­ual runs would be even larger than the esti­mate given in the Skep­tic article.

    That is, the given uncer­tainty rep­re­sents a lower-limit estimate.

    Regard­ing your follow-up com­ment, the line in Fig­ure 2 is not a lin­ear fit to the dis­played GCM out­puts.  It is an entirely inde­pen­dent cal­cu­la­tion, as shown in the deriva­tion given in the Sup­port­ing Information. 

    Gavin on Real­Cli­mate repeat­edly and erro­niously dis­missed that line as a fit, which it is not, and which dis­missal rests upon a misrepresentation.

    The line in Fig­ure 2 shows that all one need do to repro­duce global aver­age tem­per­a­ture pro­jec­tions of com­plex GCMs, with respect to the effect of increas­ing GHG’s, is to lin­early prop­a­gate the GHG forcing.

    This in itself is an extra­or­di­nary result, given all the heavy-breathing about how cli­mate is rife with com­plex feed­backs and non-linear chaotic excur­sions. One sees none of that in the GCM-produced tem­per­a­ture trends.  It’s as though, in GCMs, all the cli­mate feed­backs aver­age out to zero—at least across 80 to 120 years. That hardly seems realistic.

    As the Fig­ure 2 line closely repro­duces GCM global aver­age tem­per­a­ture out­puts, it is entirely jus­ti­fi­able to use the same gen­er­at­ing equa­tion to esti­mate the uncer­tainty in those outputs.

    It seems as though the uncer­tainty in inter­me­di­ate cli­mate states are typ­i­cally not prop­a­gated step-wise for­ward in time-wise GCM pro­jec­tions. One won­ders on what grounds this exclu­sion can pos­si­bly be justified.

    Mitchell, you also sug­gested that, “What he (or some­one) needs to do is to look at what uncer­tainty in those model para­me­ters does to the vari­ance of the pre­dicted tem­per­a­ture.

    In the arti­cle, I men­tioned search­ing the cli­ma­tol­ogy lit­er­a­ture with­out luck for such a study. Don’t hold your breath wait­ing for a cli­mate mod­eller to do it. The para­me­ter uncer­tain­ties (among oth­ers) would have to be prop­a­gated through the physics. The project would be dif­fi­cult, and extremely revealing.

  • Thanks for the clar­i­fi­ca­tions, Patrick.

    I thought the demon­stra­tion in your Fig­ure 2b (and the ‘com­plex’ for­mula that accom­pa­nied it) was hilarious.

    Peter

  • Pat Frank wrote:

    Thanks Peter, and thanks for your inter­est. It’s been quite a sleigh-ride. So far, I’m still aboard.

    Quite clear is it that the IPCC have been play­ing fast-and-loose with proper uncer­tainty esti­ma­tion, and that they are a closed and rather para­noid shop in con­tra­dic­tion to their own rules and standards.

    Over at ClimateAudit.org, Steve McIn­tyre has been fol­low­ing David Holland’s FOIA attempts to have the IPCC release its reviewer notes, in accor­dance with its sup­posed com­mit­ment to open­ness and trans­parency. Mr. Hol­land is being stonewalled by bureau­cratic stratagem.

    So, if you have some influ­ence, or come to a posi­tion of gov­er­nance again, it would be of great ben­e­fit to use those pow­ers to require a full and inde­pen­dent audit the IPCC.

    Espe­cially how the IPCC have accom­mo­dated reviewer objec­tions and fol­lowed their own rules of data qual­ity.  Steve McIn­tyre has doc­u­mented some seri­ous lapses at his site.

    Aus­tralia could take the lead in call­ing for an audit, or even doing it in-house, and the IPCC could not legit­i­mately refuse. What, after all, have they to hide if all is above-board? An hon­est bro­ker wel­comes an audit.

    Far too much is rid­ing on cli­mate change to allow the IPCC to keep its hand hidden.

  • Pat Frank wrote:

    Peter, just to let you know I’ve up-dated the eval­u­a­tion of cli­mate model cloud error to the CMIP5 GCMs, at WUWT here.

    The CMIP5 GCMs will be used to val­i­date the IPCC’s human attri­bu­tion state­ments in the up-coming AR5.

    The more advanced CMIP5 GCMs proved no bet­ter at repro­duc­ing cloudi­ness than were the CMIP3 GCMs that were eval­u­ated in my 2008 Skep­tic paper

    As a result, the growth of error is rapid and obvi­ates the reli­a­bil­ity of any pro­jec­tion of future cli­mate or of future sur­face air temperature.

    Cli­mate mod­els remain unable to resolve the effects (if any) of anthro­pogenic GHGs and remain entirely unable to attribute his­tor­i­cally recent cli­mate warm­ing to any cause at all.

Leave a Reply

Your email is never shared.Required fields are marked *