so, your denialist friend says, "No computer model can predict climate"

[greenman3610]

show him this graph from the IPCC.

It shows computer model results.
The blue band in each graph is modelled temperature
predicitions using just "natural" forcing - sunlight, ocean currents, particulates,
clouds.
The pink band is the prediction when man made forcings, co2, methane,etc,
are added to the mix.
The black line is ACTUAL measurements of temperature.
IN each case. globally, and for each continent, land and sea, the
results are impressive.

[Kagemusha]

1. So um... when were computers invented again? What year?

These graphs go back to 1900 after all.

I'm not doubting the results (from that alone). I'm just pointing out what catches my attention offhand.

[greenman3610]

3. the models look backward as well as forward

[salin]

4. ?

a computer "model" graphs data entered and the date of when computers were invented is rather irrelevant. The question isn't when computers were invented - but rather was this data collected back in 1900 - listing data sources would add weight to the charts.

[Kagemusha]

6. ...So why does the pink band extend back before computers?

I understand why the data goes back to 1900. It would be rather difficult not to. Why the graph treats that data as "within predictions", I'm not so sure of. (Mind you, I'm plenty computer literate but I'm not a statistician.) Your explanation makes it sound largely like it's to make a more compelling visual image.

[salin]

7. I am not clear how when computers were invented

figures into the question? If one has data (for anything) one enters it and can do modeling - so in my mind the date of the invention of computers is sort of irrelevant. Thus, I think I am just missing your bigger point (blame it on no coffee, yet this morning for salin ;-)

[BlooInBloo]

13. You can project backwards in time as easily as you can forwards...

... And you can test the validity of your projection methodology when you project back, since you have the "correct answers" (instead of having to wait and see).

[sarge43]

9. National weather services were nearly universal by 1900.

Ours was established in 1870. The International Meteorological Organization, now the World Meteorological Organization, began the international exchange of weather data in 1873. The data on world wide weather for the past century and a half is massive and valid. If you want the daily temperature in Kiev during September 1905, it's merely a matter of digging.

[salin]

10. thanks. Good info.

[Kagemusha]

11. Which is totally besides my point but whatever.

Apparently people feel modeling temperatures of years past using data of years past proves something deep and moving about predictive capability for the future. That's apparently what I'm not getting. To me, it'd be more impressive to predict the future and have the future data prove the models right. Proving that the models successfully "predict" unchangeable historical data somehow is just not impressing me like it should.

[muriel_volestrangler]

16. The models are started at the beginning of the period, and then left to run

So, given the starting conditions of around 1900, they successfully predict the present day temperatures - over a period of a century or so - if both natural and anthropogenic variables are taken into account (feeding into them the actual data on carbon dioxide emissions etc. over the 20th century, and getting them to work out resulting temperatures, rainfall and so on). So we know they could start at 1900, and predict the climate in 2000.

Therefore, we have good confidence they can start at 2000, and we can then feed into them scenarios of future carbon dioxide emissions, and see what they predict will happen to the temperatures in the next 100 years.

[yellowjacket]

2. So let me get this straight...

When talking to someone who doubts the accuracy of computer models in predicting climate change, the best way to convince them is to show them another computer model? Brilliant.

[greenman3610]

5. what you show them is, that the computer models have been

stunningly accurate.
most people do not realize that we have been successfully modelling
climate for the last 30 years, and those models have only
gotten better with time.

[yellowjacket]

8. So 30 years of accuracy means that I should

extrapolate that out to mean that they would have been accurate in 1900 and will continue to be accurate in 2200? Thirty years is but a blip in history and I remain skeptical that human understanding of all the forces involved in climate change is sufficient to model it on a computer. Just a thought.

[Toots]

15. Gee maybe you could do a peer reviewed article to be published in leading science journals.

There have been almost a thousand such articles written and reviewed by the top scientists throughout the world and not a single one of them denies that 1. the earth is warming...2.Human activity is a factor in it. These are the same scientists that put man on the moon and created the atomic bomb. Their entire life is dedicated to research and not making any unfounded claims. They have been studying these conditions for decades using every resource available to mankind and have issued dire warning to all humanity. Are you so arrogant as to believe human activity has no consequence. Maybe it is more a matter of ulterior motives than intelligent discourse. :shrug:

[On the Road]

14. Yes, Displaying Data Demonstrating

that taking into account greenhouse gases results in more accurate results than not taking them into account.

How else would you do something like this? Wait and see for twenty years?

[BlooInBloo]

12. Even the worst batter'll get a hit, if he swings enough times.