Bishop Hill Diary dates, Winchester Science Festival edition
Jul 21, 2015 Diary dates Some of the events at the Winchester Science Festival this coming weekend may be of interest to readers.
This Sunday, we have a talk by Tamsin Edwards:
How can we predict the future of our planet?
Every week the papers report new predictions about future climate change – whether about rising sea levels, more severe heat waves, or better wine making in the UK. Scientists make these predictions using highly complex computer models and measurements of the past. But it’s rare to hear much about how they do this, or the sometimes surprising stories behind the science and technology. What do climate models have in common with car radios? Did you know we use the same algorithms as email spam filters to reject “junk predictions”? Tamsin will lift the curtain on this cutting-edge research area.
On Saturday there is a panel discussion, which I understand will feature Tamsin and Richard Tol.
You are definitely wrong! Certainty & uncertainty in the scientific debate.
Dallas Campbell and friends
From climate change to GM crops, alternative energy to drug use, the world of science is full of controversial topics that divide opinion. TV presenter and science populariser Dallas Campbell will bring together a panel of experts to discuss why we believe what we believe, and why certain areas attract so much controversy.
Lastly, on Sunday afternoon, statistician Norman Fenton, who was featured at BH when he appeared on Horizon recently, is doing a talk too:
Fallacies of probability and risk
From climate change and surgical decisions to motor insurance or forensic evidence, Bayesian reasoning and analysis offers us a remarkable insight into why the world behaves the way it does and how we can prepare ourselves for future events.
Reader Comments (26)
If that's the best they can do no wonder climate "science" is such a crock.
Mike Jackson, filters on junk predictions.
If their spam filters have been used to reject junk predictions, how much money are they now after to get better filters for the rejection of junk predictions?
Would Mann's Holy Hockey Stick have made it through their junk prediction software?
Tamsin Edwards - the little Brian Cox wannabee who would prefer makeup to models. The use of the word 'we' should tell you who pays her rent.
Saturday line-up: Dallas Campbell (mod), Tamsin Edwards, Chris French, Suzi Gage, yours truly
Never heard of two of the panel
Christopher Charles French is a British psychologist specialising in the psychology of paranormal beliefs and experiences, cognition and emotion.
Suzi Gage is a post-doctoral researcher at the University of Bristol, investigating associations between substance use and mental health.
Is that them?
@SandyS
Correct.
I would rather hear Tamsin talk about ***Validation*** of climate models, rather than about how some of the algorithms are the same as those used elsewhere. How do we know that climate models have any ability to predict the future climate, rather than just being a random collection of respectable algorithms?
SandyS, Richard Tol
If a panel needs expertise on paranormal activity and substance abuse, climate science has more problems amongst its rank and file, than it is letting on.
Is this to examine why forecasts are always 'higher' than expected, and to converse with long 'dead and buried' tree ring data?
Slightly tangential, but there was an illuminating clash on Woman’s Hour this morning between the gynaecologist Professor John Studd and a psychiatrist, who were discussing the treatment of 50-something women with depression.
Studd has been arguing for years (decades, apparently) that the cause is nearly always hormonal and treats it accordingly with HRT patches. Most GP’s, however, prescribe anti-depressants and when those don’t work, refer such patients to psychiatrists, who use CBT, which mostly doesn’t work either.
The problem, as with many branches of medicine (and climate science), is that the specialists are entrenched and unwilling to consider other explanations. Professor Studd has written extensively about his work, but can never get published in psychiatric journals and is always turned down when he offers to lecture to anyone outside his own field. I expect he'd get on with Nigel Lawson...
Where’s Douglas Adams’s point-of-view gun when you need it?
"associations between substance use and mental health"
I wonder if that explains Dr Lew's output..?
@SandyS/Richard Tol: I suspect it is the old "climate deniers are all mentally ill" stick being swung at us, as they have to be not to believe the sky is falling in!
Could each panel member be asked to nominate 3 IPCC approved, peer reviewed papers, that have caused the most problems for climate scientists to reconcile with factual evidence?
Mann's ghost writings for others should be included, given the panel.
'What do climate models have in common with car radios?'
They are both tuned?
" Did you know we use the same algorithms as email
spam filters to reject “junk predictions” "
At some point you still have to train the algorithm manually to tell it was is junk and what isn't. It doesn't seem to make much sense applied to a GCM.
ie
https://en.m.wikipedia.org/wiki/Naive_Bayes_spam_filtering
Rob
Basically all cooling scenarios that arise from heavy-handed aerosol forcing, are discarded. You simply can't have too much warming though. As ever with Bayesian rules in the absence of actual expertise (ie a history of getting things right from time to time) it is all highly subject to confirmation bias.
Chris French is sound. He is been a debunker in the skeptics society more than 20 odd years.
- Though I don't know his current views on climate.
- Unfortunately there was a decline in the skeptics movement originating out of parts of the American skeptics movement who have lazy skeptic attitude relying on "scientists say" rather than "science says" and go around sneering and dismissing who don't agree with such science authority.
"Suzi Gage is a post-doctoral researcher at the University of Bristol, investigating associations between substance use and mental health." --SandyS
I hope she is familiar, then, with adrenaline addiction, the constant search for excitement where there is none. Adrenaline addicts are prone to make statements like, "We have only five years left to stop the Earth from boiling like a pot of gruel! Let the panic begin!"
"What do climate models have in common with car radios?"
What you dial into them totally determines what you get out of them.
Would that Josh were there when "Tamsin will lift the curtain on this cutting-edge research area."
He could capture the scene as The Wizard of Oz utters the ultimate line, "Pay no attention to that Man(n) behind the curtain!"
http://www.criticalcommons.org/Members/pcote/clips/the_wizard_of_oz-title1_2.mp4/view
I used have an insatiable interest in all things science. I think I've been cured. The paragraph about climate models makes me feel a bit queasy.
Anyone thinking of going to this? Not sure the talks interest me that much but would make a good venue for a meet...
Winchester was also the seat of King Canute... just an observation :-)
I'll be in moi combine harvester oo arr, five miles east of the venue....All welcome for a harvest tea.
@Alan the Brit
They can try and swing that at me.
I'll be in moi combine harvester... unbunging it? ~ inch forecast Friday ;-)
Yeah, just seen the forecast. Join me for a harvest moan, instead. And as you know, we farmers very rarely moan....
You can tell what tone to expect by the choice of other events they FRAME it with
EXHIBITORS
include the company chaired by Lord Deben
"Veolia will be showing how they turn waste into useful resources. Come and have a go with their hands-on model of an energy recovery facility and see what they do with what YOU throw away at home."
..and how they make use your Tax Money that you throw away on eco-subsidies ..meyybe
I think the choice of events at the science festival and the regular Scientifique events there is influenced by the fact that some of the organisers are connected to Basingstoke Transition Town etc.
I mentioned an August event in the discussion thread
A somewhat, and often used effective spam algorithm is:
1. check email sender ID against blacklist.
2. If on blacklist send email to spam folder.
3. If not on blacklist detect spam in email by keyword inspection. If spam send to spam folder if not spam send to inbox.
4. recipient reads email. If recipient reports as spam put sender on blacklist.
I guess this algorithm would work for a number of research paper authors (i leave you to ID your favorite culprits) but a lot of junk would get through - at least the first time! But at least it doesn't require a lot of automagical "scifience".