Bishop Hill Climate models and rainfall
Oct 10, 2014 Climate: Models Climate: WG2 I have a new briefing paper out at GWPF looking at the question of rainfall, flood and global warming. Here's the press release:
London, 10 October: A new briefing paper from the Global Warming Policy Foundation reviews the scientific literature on rainfall and floods and finds little evidence that there have been significant changes in recent years and little support for claims that they will become worse in future.
Despite claims to the contrary, there has been no significant change in rainfall trends in recent years both at global and UK levels. It remains very difficult to make strong claims about any changes there have been because of high natural variability in rainfall patterns, particularly in the UK.
Rainfall is a particularly difficult area for climate models, which have limited ability to recreate what is seen in the real world. Since these climate models are the main basis of claims that extreme rainfall and flooding events are being adversely affected by man-made global warming and that rainfall will become worse in the future, policymakers should treat such modelling with extreme caution.
Author Andrew Montford said, “We are constantly bombarded with insinuations that storms and floods are caused by or ‘linked to’ climate change.”“In reality these claims are usually based on climate models, which have a demonstrable inability to tell us anything reliable about rainfall. The scientific evidence shows that a simple extrapolation of rainfall averages over time can give better rainfall predictions than climate models,” he added.
Here's the paper.
Reader Comments (33)
"high natural variability in rainfall patterns, particularly in the UK"
Or what the Met Office, in its saner moments, calls "unsettled weather".
Roger Pielke, Jr. has been making the case for several years that the scientific literature does not support claims of increasing extreme weather. His book will be published November 1, The Rightful Place of Science, Disasters and Climate Change.
http://www.amazon.com/Rightful-Place-Science-Disasters-Climate/dp/0692297510/ref=sr_1_1?s=books&ie=UTF8&qid=1412947243&sr=1-1
It's chaotic. Not only do models fail to get it right, they cannot possibly ever get it right except by coincidence.
See rgbatduke's recent post at WUWT for an explanation of why.
Which should prompt the question from policy-makers; "Just what do the models have a demonstrable ability to tell us anything about? That the world doesn't blow up in any model runs? The models are certainly not inconsistent with the world not blowing up." I'm struggling to think of anything useful predicted by climate, not weather, models.
The last time I looked at the Wet Office's propaganda sorry I meant scientific based website, which was few years ago now, the level of technical output was astonishingly impressive, along the lines of "some areas will have more rain, some areas will have less rain, some areas will have warmer temperatures, some areas will have cooler temperatures, some areas will have more drought, some areas will have less drought"! I dare say little has changed.
The observational evidence for an increase in extreme flood events at a given location is relatively easy to determine. One of the standard reports of streamflow data is annual peak flow. The set of annual peak flow values can be analyzed for trends over the period of record for that flow gauge. This is becoming standard practice for engineers analysing flow records to determine if the usual assumption of stationarity is valid at that location.
In work that I am aware of in Western Canada, the trends in annual flow peaks are slightly up or slightly down depending on the location. In no cases have the trends been determined to be statistically significant. I am sure similar work has been done elsewhere in the world. It has probably not been reported because significant trends have not been found.
The reason that there appear to be trends is because of a hydrological phenomenon known as "clustering". Though we like to think that flood events occur randomly in time, they often cluster with, for example, a decade or more of high annual flood peaks flowed by a relatively quiet period. This is most likely due to the PDO, AMO and other oscillations that we have not yet detected. The unknown unknowns.
Where the models become reality. Not uncommon in petroleum geology where geologists believe the models and over look the reality of what exists in drilled wells.
Judy Curry has a big article on the Wall Street Journal editorial page this morning: "The Global Warming Statistical Meltdown."
It is very good on the models.
Prof. Curry's article is here
And on her own site here...
What would be the reason for any changes in these events? At least we could stop looking at global average figures and highlight the actual local changes in the UK. So where has warmed the most in the country? as I guess this should have the greatest increase in 'extreme weather'.
I have a model that it 100% reliable in telling me about rainfall - Big Ears the pheasant. I only have to look out of the conservatory and I can tell from how pissed-off and soggy he looks, exactly how heavy it's been raining. And, unlike the so-called experts at the Met Office, he literally works for peanuts, although he does crap on the doormat if we forget to feed him.
So once again a fact centered review of the climate shows skeptics are correct: There are no significant changes in rainfall patterns.
I think it was not based on models but by scurrilous journalists, politicians and a very few outlier scientists just asserting it. I've not heard of any papers that made any connection, just conference announcements. That no GCM is capable of such forecasts is in fact well known. The only model I heard of was by '13 degrees' Myles Allens climateprediction.net whose name must be ironic because up until July they had a job offer for:- Research Job: Changing Risks of Droughts, Water Scarcity and Heatwaves Using Ensemble Climate Modelling. That was based on funding obtained based on the previous years aftcasts of more drought in exactly the same regions that they later predicted 25% increased flooding for. The green menace had previously been trying to build a case that fracking was going to cause water shortages in 'water-stressed' areas. Presumably that fake argument is now dead-in-the-water so to speak! Contradiction, thy name is climate science!
The reason that there appear to be trends is because of a hydrological phenomenon known as "clustering". Though we like to think that flood events occur randomly in time, they often cluster with, for example, a decade or more of high annual flood peaks flowed by a relatively quiet period. This is most likely due to the PDO, AMO and other oscillations that we have not yet detected. The unknown unknowns.
Oct 10, 2014 at 3:00 PM | Unregistered Commenterpotentilla
Something I observed in the 80s and 90s when I was in the UK. They had 8 years of hosepipe bans followed by 6 yrs of too much rain. They then enlarged their reservoirs and the stress was relieved. No more hosepipe bans. The bans occured because the water reserve was right at the limit in times of "drought" (3 weeks without rain)
Salopian,
The fact that your pheasant has big ears is surely a worrying sign of Climate Change. Pheasants never used to have big ears. They have perhaps adapted, in order to pick up the cries of "wolf, wolf!"
Clustering is symptomatic of a purely random process. An unclustered 'smooth' distribution is symtomatic of a non-random process.
I recall a useful paper on rainfall trends in the UK by Stuart Lane - possibly ~2008. I took from it that floods etc are caused by collective memory loss
The degree of clustering in hydrologic data is greater than what would be expected from a purely random process. This was first discovered in the 1930s by Dr. H.E. Hurst who worked for the Ministry of Public Works in Cairo for 62 years. He examined about 1000 years of data from the River Nile (the longest flow record in the world) and developed what has become known as the Hurst Phenomenon. There is extensive recent literature on the Hurst Phenomenon including an excellent paper by Demetris Koutsoyiannis on climate variability, see link below. Another of his papers is mentioned in the Bishop's write-up.
http://www.tandfonline.com/doi/abs/10.1623/hysj.48.1.3.43481#.VDg3hSldVZU
I have looked at weather and Extreme events in the UKback to the 13 th century. What comes over loud and clear is that extremes in the past are much worse than they are today which, with few exceptions, is a benign period of climate.
The most noticeable climate states in this country are storms, droughts, hot weather, but most noticeable of all are the extreme flooding events and rain which often caused flooding and famine.
What is also noticeable is that the most extreme events, especially rainfall, occurred during colder periods of our climate not the warmer ones.
Obviously snow and ice also figured at times but in the scope of seven hundred years they are a relatively minor part of the whole range of weather experiences and restricted to certain periods.
Tonyb
Tonyb
Climate models tend to overestimate the cooling by water evaporation by up to 3%. What impact this approximation has on results (in a chaotic system) is unknown, but modelers are too busy to evaluate it.
http://judithcurry.com/2013/06/28/open-thread-weekend-23/#comment-338257
Tonyb, slightly O/T, but it is way past time that I thanked you for the work that you have put in.
If there is a living person who has made more unpaid effort to investigate the historical temperatures of England, then I would like to read about that person.
And if there was a "McIntyre Medal" for devotion to data, then you would receive my nomination.
best regards,
michael hart in deep Central-England.
potentilla
I notice the link you posted was to a paper entitled
Climate change, the Hurst phenomenon, and hydrological statistics
An interesting title when examining a 1000 years of data (10% of the current inter-glacial), perhaps a hint to how the grant application was worded? Reading the full text will have to wait for the moment.
Anyone ever correrlated weather patterns against the rotation of the planets in our solar system .
Today's Guardian reporting that a 100 000 UK homes could easily go Carbon neutral.
Easily achievable the people living inside them can't afford to heat them.
No amount of empirical evidence will persuade people like Ed Davey. Last night, on R4's Any Questions, he was bragging about the LibDem's success in increasing the amount of renewable electricity capacity in this country (sic). Thankfully, his party will disappear at the next election.
Ed Davey can increase the renewable electricity CAPACITY until he's blue in the face - but when there's no wind there's NO ELECTRICITY (current wind output 0.46GW/1.46% of modest demand)
When will he 'get it'..? Before next May..? No chance...
I've always favoured the Stone Hanging on Fence method of weather forecasting..
Stone casting a shadow - sunny.
Stone wet - raining.
Stone swinging about - windy.
Can't see stone - foggy
Stone covered in white stuff - snow.
Works for me...
SandyS noted:
"An interesting title when examining a 1000 years of data (10% of the current inter-glacial), perhaps a hint to how the grant application was worded? "
I agree many grant applications are worded with climate change in the title to increase the chances of funding. However the implication that Demetris Koutsoyiannis has played this game is completely unfair. The paper I linked to addresses natural climate variability, a topic of great interest to engineering hydrologists given that most hydrological records are relatively short. Natural climate variability is also of interest to BH readers given that alarmists attribute 100% of the warming since the Little Ice Age to anthropogenic causes. It is likely that much of the warming is due to natural variability.
@sherlock1 stone yellow = UEA student was nearby
- heard Labour MP on Any Questions "You UKIP people are all racists..now the water question , I say 'if it's brown flush it down' ...er..hang-on sorry"
An excellent and timely monograph. The models have been poor with temperature 'projections', and worse with precipitation. In the UK, local and national authorities with responsibilities for infra-structure and much else besides are blindly using UKCP09 guidelines for the weather they must plan for. Guidelines relying on computer models that demonstrably do not work. Great to see the spotlight there.
I did some quick statistics on our wet weather, you can see the results on my website, comments welcome.
Has The UK Had Exceptional Winter Rainfall Or Is It Just Weather As Normal?
Andrew, congratulations for being the prime target of the attention of the top passive-aggressive climate alarmist - Anders at ATTP! He and his merry band of miserabilists really hate your study.