You (and Curry) have a point, but I'm not sure it's the correct point. Yes, I expect that good software V&V practices will become a part of climate modelling as used for policy making, just as we see in the nuclear industry.
Will it make climate models more correct? Probably somewhat, at great expense. Still, I remember Steve Easterbrook finding that the informal QA methods used in climatology already produce remarkably high quality code.
And of course we all know that the real issues with climate modelling quality are in representing the physics, rather than formal code correctness. This is optimizing the trees for the forest, and probably a very subopimal use of resources if overall quality is the objective.
Yet, I have no doubt we are going there. It's politically inevitable. But let's be explicit: it's about creating the appearance of rigor. As such, it is just another (small) aspect of scientists having to get much better at public relations. What you have written about a lot, Michael.
Join us tomorrow for Climate and Sustainability: Moving by Degrees from 8:00 a.m. to 5:30 p.m. PT at ( http://www.elabs7.com/ct.html?rtr=on&s=fj6,lqj6,dv,j0of,3cjy,js8a,lyyo ) movingbydegrees.org and on twitter @ ( http://www.elabs7.com/ct.html?rtr=on&s=fj6,lqj6,dv,kskt,7ldk,js8a,lyyo ) #MPclimate
Participate in this one-time-only event by posting questions and ideas to the day's moderators and participants. Stream live video of keynote speech by Andrew Revkin, NYT Dot Earth blogger.Participate in discussions with panelists Joe Romm, Dr. Michael E. Mann, Michael Levi and others.Watch exclusive online-only interviews with Dr. Stephen Schneider, Stanford University; Eric Pooley, BusinessWeek; and Elizabeth Kolbert, The New Yorker.
( http://www.elabs7.com/ct.html?rtr=on&s=fj6,lqj6,dv,u5g,c3r6,js8a,lyyo ) View the complete agenda. Spread the word to invite others and join attendees from across the nation for this online event.
With regards to the comment from Willis that she was responding to... How much science would be "discredited and disowned" if all data were required to still be around?
Your top quote is from publicity from Naoimi Oreskes' important book, and you point out the first sensible thing I have heard from Judith Curry.
ReplyDeleteToday I feel hopeful!
You (and Curry) have a point, but I'm not sure it's the correct point. Yes, I expect that good software V&V practices will become a part of climate modelling as used for policy making, just as we see in the nuclear industry.
ReplyDeleteWill it make climate models more correct? Probably somewhat, at great expense. Still, I remember Steve Easterbrook finding that the informal QA methods used in climatology already produce remarkably high quality code.
And of course we all know that the real issues with climate modelling quality are in representing the physics, rather than formal code correctness. This is optimizing the trees for the forest, and probably a very subopimal use of resources if overall quality is the objective.
Yet, I have no doubt we are going there. It's politically inevitable. But let's be explicit: it's about creating the appearance of rigor. As such, it is just another (small) aspect of scientists having to get much better at public relations. What you have written about a lot, Michael.
Hope this is an appropriate place, if not feel free to bump it or dump it.
ReplyDeleteWednesday all day on National Public Radio's website:
( http://www.elabs7.com/ct.html?rtr=on&s=fj6,lqj6,dv,j0of,3cjy,js8a,lyyo )
Join us tomorrow for Climate and Sustainability: Moving by Degrees from 8:00 a.m. to
5:30 p.m. PT at (
http://www.elabs7.com/ct.html?rtr=on&s=fj6,lqj6,dv,j0of,3cjy,js8a,lyyo )
movingbydegrees.org and on twitter @ (
http://www.elabs7.com/ct.html?rtr=on&s=fj6,lqj6,dv,kskt,7ldk,js8a,lyyo ) #MPclimate
Participate in this one-time-only event by posting questions and ideas to the day's
moderators and participants. Stream live video of keynote speech by Andrew Revkin,
NYT Dot Earth blogger.Participate in discussions with panelists Joe Romm, Dr.
Michael E. Mann, Michael Levi and others.Watch exclusive online-only interviews with
Dr. Stephen Schneider, Stanford University; Eric Pooley, BusinessWeek; and Elizabeth
Kolbert, The New Yorker.
( http://www.elabs7.com/ct.html?rtr=on&s=fj6,lqj6,dv,u5g,c3r6,js8a,lyyo ) View the
complete agenda. Spread the word to invite others and join attendees from across
the nation for this online event.
With regards to the comment from Willis that she was responding to... How much science would be "discredited and disowned" if all data were required to still be around?
ReplyDelete