November 05, 2009
Nitrogen Cycle Added To Climate Model

What is missing from climate models?

OAK RIDGE, Tenn., Oct. 9, 2009 -- For the first time, climate scientists from across the country have successfully incorporated the nitrogen cycle into global simulations for climate change, questioning previous assumptions regarding carbon feedback and potentially helping to refine model forecasts about global warming.

My own reaction: amazement. We are in the year 2009 and only now the nitrogen cycle gets added to climate models? What other important factors are not yet in climate models? Does anyone know? I'm looking for a knowledgeable reply, not a rant. What is the state of climate models? What are the prospects for more accurate models 5, 10, 20 years from now?

These scientists expect more rapid climate change as a result of adding the nitrogen cycle.

The results of the experiment at the Department of Energy's Oak Ridge National Laboratory and at the National Center for Atmospheric Research are published in the current issue of Biogeosciences. They illustrate the complexity of climate modeling by demonstrating how natural processes still have a strong effect on the carbon cycle and climate simulations. In this case, scientists found that the rate of climate change over the next century could be higher than previously anticipated when the requirement of plant nutrients are included in the climate model.

ORNL's Peter Thornton, lead author of the paper, describes the inclusion of these processes as a necessary step to improve the accuracy of climate change assessments.

You might think climate models can be really accurate even without the nitrogen cycle. But recall a recent post I did about biofuels boosting nitrous oxide emissions and thereby causing big warming effects. Where the nitrogen goes and in what form is very important for the climate.

MBL, WOODS HOLE, MAA report examining the impact of a global biofuels program on greenhouse gas emissions during the 21st century has found that carbon loss stemming from the displacement of food crops and pastures for biofuels crops may be twice as much as the CO2 emissions from land dedicated to biofuels production. The study, led by Marine Biological Laboratory (MBL) senior scientist Jerry Melillo, also predicts that increased fertilizer use for biofuels production will cause nitrous oxide emissions (N2O) to become more important than carbon losses, in terms of warming potential, by the end of the century.

Fertilizer usage is going to rise anyway due to increasing human population and industrialization. So how much will nitrous oxide emissions increase in the 21st century? Also, how much will fertilizer run-off and phytoplankton blooms increase? Will the Antarctic continent become livable as a result? Always look on the bright side of life.

Share |      Randall Parker, 2009 November 05 08:10 PM  Climate Trends


Comments
Bruce said at November 6, 2009 8:03 AM:

Randall, there are about 30 or so climate models the warmenizers refer to.

If you graph them, they form a nice fan shape.

Some come in higher than measured. Some below. Some near some of the time.

None of them are accurate. But the warmenizers love to point to graphs and point out reality falls somewhere among them all.

Its all a big cruel snake oil joke.

odograph said at November 6, 2009 8:05 AM:

I think it's more important how well models are clustering on good predictions than how many factors they have.

Compare to economics. There are fiscal stimulus models but they diverge immediately, and the actors (Krugman and Mankiw) break into their traditional ideological lines. That is an epic fail. Does climate do better? If they are clustering for scientific and non-ideological reason, I'd say so.

Alex said at November 6, 2009 1:56 PM:

I think there is a classic point to be made there. How can one trust the veracity of the models. The folks model investments markets had a huge incentive to get it right, and we had 'epic fail.' Where is the incentive for climate modelers? How are people to judge success or failure?

-Alex

Reid of America said at November 6, 2009 2:26 PM:

It is not possible to forecast climate with any skill. The IPCC admits this in their fine print and calls the results of the models "scenarios". They are not projections or predictions or forecasts. To do those error bars must be propagated in the computations. If you propagate error bars in the computations the models quickly become useless under any rational use of mathematics hence they are "scenarios"

Roger Pielke, Sr. has some serious articles about the deficiencies of climate models. Essentially they are only useful for process studies but completely useless in forecasting the real world. All 20 IPCC models have failed. The all failed to predict global cooling after the 1998 peak year. Proof positive that climate modeling has no predictive skill.

Some links to Pielke's articles:
http://pielkeclimatesci.wordpress.com/2008/11/28/real-climate-misunderstanding-of-climate-models/
http://pielkeclimatesci.wordpress.com/2008/11/14/are-multi-decadal-climate-forecasts-skillful-2/
http://pielkeclimatesci.wordpress.com/2008/11/13/what-are-climate-models-what-do-they-do-2/

Many more articles are listed here:
http://pielkeclimatesci.wordpress.com/category/climate-models/

th said at November 7, 2009 4:19 PM:

odograph writes:

"and the actors (Krugman and Mankiw) break into their traditional ideological lines. That is an epic fail. Does climate do better? If they are clustering for scientific and non-ideological reason, I'd say so"

Are you serious? You can't detect anything ideological about the science of global warming that despite being so miserably wrong that the urgency to pass legislation to stop it can't wait another day?

LAG said at November 8, 2009 5:51 PM:

What!? I thought this was settled science. Al Gore-ical, where are you?

John Moore said at November 9, 2009 8:17 PM:

It shouldn't be surprising that there are large missing factors in climate models. They don't have the ability to calibrate even the simpler models. Adding nitrogen simply creates yet another parameter to be fiddled and fudged. It will increase model complexity, but it is unlikely to increase model skill. Even if it does, there's no way to measure it.

The models do not exist to provide forecasts, because it is utterly and completely impossible to do forecasting beyond a few weeks using physical simulation models. They are used to attempt to measure sensitivity to certain conditions (say, CO2 concentration).

Lynn said at June 19, 2011 8:33 AM:

The nitrous oxide from wildfires is now a major feedback loop, not yet incorporated into the models. Oops.

http://www.loe.org/shows/segments.html?programID=11-P13-00024&segmentID=1

Post a comment
Comments:
Name (not anon or anonymous):
Email Address:
URL:
Remember info?

                       
Go Read More Posts On FuturePundit
Site Traffic Info
The contents of this site are copyright