From: "Mitchell, John FB" To: 'Mike Hulme' Subject: RE: GEC paper Date: Fri, 30 Apr 1999 17:23:15 +0100 see inserts jfbmitchell@meto.gov.uk Hadley Centre for Climate Prediction and Research The Met. Office, Bracknell RG12 2SZ UK Tel +44 1344 856613/6656 Fax+44 1344 856912 > -----Original Message----- > From: Mike Hulme [SMTP:m.hulme@uea.ac.uk] > Sent: Friday, April 30, 1999 12:31 PM > To: Mitchell, John FB > Subject: RE: GEC paper > > John, > > Could you have a quick look at this paragraph (see below) from the GEC > fast-track paper. I do not understand: > > a) why CO2-doubling forcing for CM2 is cited (see your original email at > the end of this message) as 3.26Wm-2 when I thought it was 3.471Wm-2 (I'm > sure I've seen 3.471Wm-2 cited elsewhere for HadCM2). [Mitchell, John FB] 3.471 in longwave, 3.26 when shortwave also taken into account. Unfortunately modellers do not always make clear how they have estimated their CO2 forcing. > and > > b) why the forcing curves in the plot William Ingram sent show higher > forcing in CM2 than CM3 (by about 0.5Wm-2) when the CO2-doubling forcing > is > *lower* in CM2 cf. CM3. [Mitchell, John FB] HadCM2 is 1%/year increase in CO2 which is only approximately equivalent to IS92a. Hadcm 3 is "95a" - in fact "95a" I think differs only from in the conversion of the 92a emissions to concentrations, so strictly speaking is not an emissions scenario. As far as I know, Tom never did explain why his concentrations in 1995 were different form the ones Jonathan and I derived using his 1992 model- I think CH4 liffetimes and the CO2 sink were the main factors. > [is this solely due again to the difference between IS92a and IS95a > concentrations?] > > and > > c) why the global-mean warmings in CM2 and CM3 are quite similar when CM3 > has a higher sensitivity than CM2 (3.3 to 2.5K over the next century) and > CM3 also has a higher CO2-doubling forcing (3.74Wm-2 to 3.26Wm-2, or > 3.47Wm-2 - see a)). Surely this should lead to faster warming in CM3 cf. > CM2? [Mitchell, John FB] See above - HadCM2 uses 1%/year increase in CO2, which gives a greater forcing than HadCM3, even after the effect of explicit trace gases is added in. (about 0.5Wm-2 by 2100). The greater climate sensitivity does not make as big a difference as one would expect. The difference in CO2 forcing per doubling is not the issue- the net forcing is, and that has ben calculated taking the difference in CO2 response into account M aybe I have misinterpreted something here. > Thanks, > > Mike > > ______________ > > Paragraph from GEC paper ...... > > "In HadCM3, greenhouse gas concentrations were increased from their 1860 > values up to present (1990) as observed and then following the IPCC > emissions scenario IS92a (Leggett et al., 1992) from 1990 to 2100. Only > one simulation was carried out. The increase in radiative forcing during > the twenty-first century is very similar to HadCM2, being only 0.5 Wm-2 > (about 10%) smaller by 2100 than in the HadCM2 experiment (Figure 2). > Note > that the ratio of the increases in CO2 concentration (HadCM2/HadCM3) is > much greater than the ratio of the changes in radiative heating. There is > a greater increase in heating in HadCM2, so a greater increase in CO2 is > required to produce the same fractional increase in heating. Also, > because > the heating due to doubling CO2 in HadCM2 is less than in HadCM3 (3.26 > Wm-2 > compared to 3.74 Wm-2), a larger increase in CO2 is required to give the > same change in heating. Note also that the increase in forcing varies as > the logarithm of the change in CO2 concentration." > > > At 14:54 09/04/99 +0100, you wrote: > >Hi Mike. > > > >2xCO2 > >HadCM2 3.26 Wm-2 including stratospheric adjustment and allowance for > >solar absorption. > >hadCM3 3.74 Wm-2 as above. > > > > > >Gordon C., C. Cooper, C. Senior, H. Banks, J. M. Gregory, T.C. Johns, > J.F.B. > >Mitchell and R. Wood, 1999. Simulation of SST, sea ice extents and ocean > >heat transports in a coupled model without flux adjustments. Climate > >Dynamics (provisionally accepted) > > > >Note year is 1997 > >Gregory, J. M. and J.F.B Mitchell, 1997. The climate response to CO2 of > the > >Hadley Centre coupled OAGCM with and without flux adjustment, J Geophys > >Lett., 24, 1943 -1946. > > > >I will try and look at then text now > >John > >jfbmitchell@meto.gov.uk > >Hadley Centre for Climate Prediction and Research > >The Met. Office, Bracknell > >RG12 2SZ UK > >Tel +44 1344 856613/6656 > >Fax+44 1344 856912 > > > >> -----Original Message----- > >> From: Mike Hulme [SMTP:m.hulme@uea.ac.uk] > >> Sent: 09 April 1999 14:11 > >> To: Mitchell, John FB > >> Subject: RE: GEC paper > >> > >> John, > >> > >> Here is a Word 6 version of the GEC paper. You need to give me two > >> references (Gregory and Mitchell 1998 and Gordon et al 1999?) and check > >> through the bits I have added. See especially what I have worded about > >> CO2 > >> concentrations in Section 7 - quite what we cite for HadCM3 I'm not > sure. > >> It depends what the impacts people say about the sensitivity of their > >> results to CO2 concentrations. I also have a question in the text in > >> Section 5 for you. > >> > >> Figure 10 is not made yet - I thought I would produce this inter-model > >> comparison plot for the Amazon given the interesting results we were > >> getting there. > >> > >> I will wait for your comments before sending it to Martin and the other > >> impacts people, but I must do this by the 19th April at latest. > >> > >> I think I understand where the various CO2 numbers come from now. > >> > >> Regards, > >> > >> Mike > >> > >> > >> At 11:59 09/04/99 +0100, you wrote: > >> >Dear Mike, > >> > I think we have traced where the different CO2 values have come from > >> > HadCM2 HadCM3 > >> > assumed 'correct' assumed 'correct' > >> > 2020s 441 470 457 434 > >> > 2050s 565 590 574 528 > >> > 2080s 731 770 712 638 > >> > > >> > The left hand HadCM2 value we think comes from SA90 - Peter Cox will > >> >check. The second HadcM2 value is notional- I don't think the > >> inconsistency > >> >between the the columns matters that much, since there is no "correct" > >> >HadCM2 value. > >> > The HadcM3 values do matter. The right hand side value is > >> >what was used in the model, and what Willaim took from the TOM Wigley > as > >> >being the SAR IS95a values. I do not know where these are publicaly > >> >available, and I have asked Dave Griggs that if we use new scenarios > (eg > >> >SRES) in the TAR, they are publicly available and well documented. The > >> left > >> >hand column appears to be from the 1992 IPCC supplement.(The annex by > >> >Mitchell and Gregory). This used the then current UEA enrgy > >> balance/carbon > >> >cycle model to convert CO2 emissions to concentrations. I presume the > >> >discrepancy comes from changes to the carbon cycle model and anything > >> elses > >> >affecting the conversion from emissions to concentrations. > Unfortunately, > >> as > >> >far as I can tell, the SAR never refers to these or explains why the > >> >concentrations are different. > >> > This could easily happen again. The situation with the new > >> >SRES scenarios to me seems rather chaotic, anad again they are > emissions > >> >scenarios, not concentration scenarios. The initial GCM runs will use > CO2 > >> >concentrations from one particular model. The TAR may report (probably > >> will > >> >report) different values since they will use a different model. The > best > >> >thing is to talk to the people who set up the GCM run to find out > exactly > >> >what was used in the model > >> > > >> > With best wishes > >> > John > >> > > >> > > >> >jfbmitchell@meto.gov.uk > >> >Hadley Centre for Climate Prediction and Research > >> >The Met. Office, Bracknell > >> >RG12 2SZ UK > >> >Tel +44 1344 856613/6656 > >> >Fax+44 1344 856912 > >> > > >> >> -----Original Message----- > >> >> From: Mike Hulme [SMTP:m.hulme@uea.ac.uk] > >> >> Sent: 08 April 1999 17:35 > >> >> To: N.W.Arnell; Sari Kovats; Matt Livermore; parryml@aol.com; > >> Andrew > >> >> White; jfbmitchell@meto.gov.uk; gjjenkins@meto.gov.uk; > >> >> r.nicholls@mdx.ac.uk > >> >> Subject: HadCM3 CO2 concentrations > >> >> Importance: High > >> >> > >> >> Dear Fast-trackers, > >> >> > >> >> In putting the scenario paper together for the GEC issue, John > Mitchell > >> >> and > >> >> I have come up with slightly different CO2 concentrations for HadCM2 > >> and > >> >> HadCM3 to what we had earlier assumed. These CO2 concentrations > will > >> >> really have to appear in the scenario paper to be consistent with > the > >> GCM > >> >> experiments. Given the differences from the values (I think) you > have > >> all > >> >> used in the impacts work, what significance does this have for your > >> work? > >> >> > >> >> > >> >> HadCM2 HadCM3 > >> >> assumed 'correct' assumed 'correct' > >> >> 2020s 441 470 457 434 > >> >> 2050s 565 590 574 528 > >> >> 2080s 731 770 712 638 > >> >> > >> >> > >> >> The difference is that the assumed HadCM2 concentrations are > 20-30ppmv > >> too > >> >> low while the assumed HadCM3 concentrations are 20-70ppmv too high. > >> >> > >> >> The assumed HadCM2 concentrations came from Cox and Friend (they had > >> >> already run Hybrid with these concentrations before the FT work got > >> under > >> >> way, so we adopted their values). I cannot yet trace where the > assumed > >> >> HadCM3 concentrations came from, but the 'correct' values are what > both > >> >> John Mitchell and the IPCC (1996 report) have calculated for the > IS92a > >> >> scenario. > >> >> > >> >> Your suggestions on how best to handle this inconsistency would be > >> >> appreciated. How big a difference do these differences make to your > >> >> impacts? > >> >> > >> >> Thanks, > >> >> > >> >> Mike > >> >> > >> >> > >> >> > >> > ************************************************************************** > >> >> *** > >> >> Dr Mike Hulme > >> >> Reader in Climatology tel: +44 1603 593162 > >> >> Climatic Research Unit fax: +44 1603 507784 > >> >> School of Environmental Science email: m.hulme@uea.ac.uk > >> >> University of East Anglia web site: > >> >> http://www.cru.uea.ac.uk/~mikeh/ > >> >> Norwich NR4 7TJ > >> >> > >> > ************************************************************************** > >> >> *** > >> >> Annual mean temperature in Central England during 1999 > >> >> is about +1.5 deg C above the 1961-90 average > >> >> *************************************************** > >> >> The global-mean surface air temperature anomaly for 1998 > >> >> was +0.58 deg C above the 1961-90 average, the warmest year yet > >> recorded > >> >> > >> > ************************************************************************** > >> >> *** > >> > << File: gec.fasttrack.doc >> > >