Reliability of regional climate model trends
Formal Metadata
| Title |
Reliability of regional climate model trends
|
| Title of Series | |
| Author |
|
| License |
CC Attribution 3.0 Unported:
You are free to use, adapt and copy, distribute and transmit the work or content in adapted or unchanged form for any legal purpose as long as the work is attributed to the author in the manner specified by the author or licensor. |
| Identifiers |
|
| Publisher |
|
| Release Date |
2013
|
| Language |
English
|
Content Metadata
| Subject Area | |
| Abstract |
A necessary condition for a good probabilistic forecast is that the forecast system is shown to be reliable: forecast probabilities should equal observed probabilities verified over a large number of cases. As climate change trends are now emerging from the natural variability, we can apply this concept to climate predictions and compute the reliability of simulated local and regional temperature and precipitation trends (1950–2011) in a recent multi-model ensemble of climate model simulations prepared for the Intergovernmental Panel on Climate Change (IPCC) fifth assessment report (AR5). With only a single verification time, the verification is over the spatial dimension. The local temperature trends appear to be reliable. However, when the global mean climate response is factored out, the ensemble is overconfident: the observed trend is outside the range of modelled trends in many more regions than would be expected by the model estimate of natural variability and model spread. Precipitation trends are overconfident for all trend definitions. This implies that for near-term local climate forecasts the CMIP5 ensemble cannot simply be used as a reliable probabilistic forecast.
|
Related Material
Video is accompanying material for the following resource
00:00
Video
Electric power distribution
Climate
Model building
00:06
Climate change
Paper
Musical development
Weather forecasting
Climate model
Rain
00:24
Metal
Climate change
Weather forecasting
Winter
Day
Scale (map)
Temperature
Climate model
Angeregter Zustand
Rain
01:15
Spread spectrum
Spread spectrum
SEED
Global warming
Desertion
Climate change
Oceanic climate
Pattern (sewing)
Weather forecasting
Musical ensemble
Temperature
Asbestos
Climate model
Remotely operated underwater vehicle
Morning
Model building
Ground (electricity)
Model building
03:14
Cell (biology)
Ranking
Pattern (sewing)
Rubber stamp
Ranking
Temperature
Musical ensemble
Scale (map)
Climate
Funksender
Map
00:06
this paper is about to reliability of climate models the idea is very simple the climate change is not so strong locally did we can starts trying to verify the models and develop people too little time France's
00:19
if you see it the weather forecast for tomorrow it says 60 percent chance of rain and where people
00:25
have checked that if you take a lot of days with a forecast of 60 percent chance of rain then all 60 per cent of those days appear to be retained 40 per cent at state try we tried to do the same with climate model trends yes the
00:38
temperature of the metal corrected for changes in places where the canvas was observed in the way it was observed in everything and you see upward trend to the 19 forties little bit downwards in the seventies and then steeply up to about now and this looks very similar to the global mean temperature the global mean temperature has less variability but it does hell social import Trenton 19 forties down invented the seventies and Fortran now so an easy way to describe the temperature in Netherlands is to say that this winter proportional to the global mean temperature is about 2 times faster if you look at the scale here the regression of this this is about to so we can make a histogram
01:18
of the trends in the climate models over here these new once on a climate models the purple 1 so the observations the seed observations show 2 times faster trend then the global mean 1 and the models are between halls and welcome to but to big spread in the bank's practice for 2 reasons 1st of all the natural variability of course is whether that's why the curve the Netherlands is so much more noisy than the global mean temperature and the 2nd 1 is model spread novel climate models and said some of them predicted more warming in the Netherlands and salt and predict less warming of the Netherlands so they also give a range of uncertainty and a question we want to answer is just like in weather forecasts it says here that it's 80 per cent of the models range is is actually correspond to 80 % you cannot do that for a single point we have to do with the whole world with started 1950
02:13
because from 1950 onwards we the observations more and you see a pattern of climate change here there's more warming in the Arctic northern Canada as said area there's more morning the deserts this arid deserts and worse than that of the Midwestern US Australia and that's less warming in the oceans especially the Pacific Ocean north Atlantic Ocean some notion here you see the model warming and you see basically the same pattern and the models and altering so badly see involves told warming in the Arctic you see this is getting warmer usually oceans getting less world but the details I'm not saying that if you look here at the percentage of the local warming in the ensemble you see that the models underestimated warming in ACM here in the West Pacific and eastern Indian Ocean and overestimated the warming trend in the Pacific Ocean and off the coast of the US the question is this this due to chance Arestis due to problems with the models were with the way the models are
03:14
so we got all this information together we make something called a rank histograms and that's the flat-lined 5 per cent of the map here should be outside the law about in sample 5 % should be in the next 5 % etcetera etcetera and we see that instead we have 1 a 10 per cent that is new and actually close to 20 per cent of that is outside his cell but the other side when we only expect 5 per cent for our conclusion is that the logical patterns are very similar to that a small skills to regional scales where many people want to have climate for cost FIL ensembles overconfident there more regions where the observed trends in the mobile transmitter is greed they would expect from pure chance
