date: Tue, 06 Jul 2004 09:28:24 -0600
from: Tom Wigley <wigleyatXYZxyz.ucar.edu>
subject: Re: Question re AR4 and long scenarios
SARAH -- SEE BELOW ....
>Thanks for input and your recent paper.
>I don't think I know a lot about emissions scenarios but if your
>IPCC overshoot scenario example is their level perhaps I am
YOU PROBABLY KNOW MORE THAN MOST WG1 PEOPLE.
>I think my role will be to report to and fro between WG1 ch 10?
>(projections) and think about what we will present.
YES -- IMPORTANT.
>Also I may have some leverage in who is invited to the meetings,
>but that won't be covered in the teleconference on Thursday I
>Re carbon cycle feedbacks
>1) is it a typo in your abstract - 'carbon feedbacks on the carbon
OOPS -- THANX FOR NOTICING THIS
>2) So in the paper you do use carbon cycle feedbacks - I think
>this is going to be a big issue in the TAR - how to justify not doing
YES, NO EXCUSE FOR NOT INCLUDING THESE. HOWEVER, SINCE ALL
COUPLED AOGCM/CARBON CYCLE MODELS WILL HAVE THESE
DIFFERENT, IT IS BETTER FOR COMPARABILITY (AND TO AVOID
OUTLIERS) TO DO THE CO2 OFF LINE.
>The scoping paper seems to cover an aweful lot -
>I don't think IPCC TAR delt with mitigation scenarios, WG1 didn't
>anyway except for the used before WRE and S profiles (too
>political) is this going to change now?
NOT READ THIS YET.
>Any thoughts appreciated.
SOME OTHER STUFF ....
FIRST, WHEN YOU DID THE AOGCM/MAGICC CALIBRATION, DID YOU
DE-DRIFT THE 1% RUN DATA?
SECOND, I HAVE BEEN WRITING UP SOME STUFF ON COMMITMENTS --
MAINLY TO PRE-EMPT IPCC, SINCE THEY WILL NOT BE ACCOUNTING FOR
UNCERTAINTIES. SEA LEVEL IS EVEN MORE OF A PROBLEM THAN I
THOUGHT. THERE ARE TWO SERIOUS ISSUES.
THE NON-MELT TERM IS LINEAR IN TIME. THIS MEANS IT GETS QUITE BIG
AFTER, SAY, 400 YEARS. MAYBE LINEAR IN TIME IS OK TO 2100, BUT AFTER
THAT IT COULD BE (IS) RATHER SILLY.
THE SECOND ISSUE IS ONE I RAISED BEFORE, DESCRIBED IN THE
MAGICC/SCENGEN MANUAL. THE RATHER DUMB JONATHAN METHOD FOR
GSICS IMPLIES AN UPPER BOUND FOR MELT, WHICH MUST BE IDENTIFIED
WITH THE TOTAL AVAILABLE AMOUNT OF GSIC ICE. THE LIMIT FOR THE
CENTRAL CASE IS ABOUT 19cm FROM 1880, WHICH, AS I SAID BEFOR, IS
ALMOST CERTAINLY TOO LOW. PERHAPS 50cm IS TOO HIGH (BUT NOT
ACCORDING TO MEIER), AND PERHAPS EVEN THE 30cm WE USED IN THE SAR
IS TOO HIGH (I THINK YOU SAID IT WAS). BUT ONLY 19cm (OR 17cm FROM
THIS BECOMES EVEN MORE RIDICULOUS IF ONE USES THE LOW AND HIGH
LIMITS RECOMMENDED IN THE TAR. FOR CHANGES FROM 1880 THE LOW IS
ABOUT 5cm (WHICH HAS GOT TO BE TOTALLY IMPOSSIBLE), WHILE THE
HIGH IS ABOUT 32cm (WHICH, WITHIN THE MANY ACCOUNTING
UNCERTAINTIES, HAS ALSO GOT TO BE TOO LOW.
THESE ARE COMPENSATING ERRORS -- TOO HIGH NON-MELT AND TOO LOW
GSIC. BUT THIS IS ALL PRETTY DISTURBING, AND ALL THE MORE REASON FOR
US (MAINLY YOU) TO GET SOME MORE CREDIBLE NUMBERS OUT.