"We an't predict the weather 5 days out, so how can we predict the climate a few years out?"

            When someone says this to me I normally respond: "So are you saying we can't predict spring, summer, fall, and winter either?".  Climate and weather are very different entities.  Climate is defined as weather averaged over time.  This averaging removes a lot of the seemingly random and generally unpredictable nature of weather.  The goal of this page is simple.  This page intends to show that there are components of the climate that can very easily be predicted months, years and even decades in advance.  Once you realize this then you will start to see how certain long term trends can actually be easier to predict than some short term trends.  If you want a more detailed analysis I suggest you read up on the climate models or read Hansen's papers.  Real Climate has an excellent series of articles on modelling as well.  But most people need to learn how to crawl before they can walk.  This page does not intend to bring you up to expert speed but to merely give you a rough idea as to why long term predictions are possible.  A few things that can be predicted more than a few days out are:

1. The sun:
1. We can say with a high degree of confidence that the sun will remain stable over the next several thousand, if not million, years.  Therefore it's effects on the climate will likely be negligible.  
2. The near term cycle is pretty steady as well.  The variance in watts per square meter is only a fraction of a percent.

3. Sourece: Global Warming Art

2. The Asian monsoon: December to march
1. This is a reasonably stable and predictable shift in weather patterns.  Most meteorologists consider this pattern very predictable.
3. El Nino: ~4 year cycle.
1. Yes the intensity varies.  However, it does occur in a somewhat regular cycle so we can rather easily predict the number of occurrences in the next score or so.  And once we get past the "spring barrier" the *short-term* predictive skills of ENSO models go way up.  You could even say we have a much better ability to predict El Nino 8 months out than we do with predicting rain a few weeks out. There is still much to learn about El Nino, we only recently launched the ARGO probes and Jason satellites,  but this example goes a long way in proving that long term events are often easier to predict than a lot of short term events.
4. Planetary orbits
1. The Milankovitch variations are rather well understood as well as their effects on the planet.

2. Source: Global Warming Art

   
   
5. The composition of the atmosphere. 
1. The composition of the atmosphere is constantly changing.  CO2 and CH3 levels have varied considerably over the last million years.  This pattern has been very regular.  The fossil fuel forcing of CO2 which created the Keeling curve is also very well understood.  The fact that this is due virtually entirely to human activities is so well established that one rarely sees it questioned.  More info on that can be found here.
2. Keeling Curve

   Source: NOAA

3. CH3, CO2, and Ice core data.  Current Ice core data goes back a million years.  However, Realclimate has a nice article discussing the last 650,000 years analyzed through the EPICA core.  

   Ice Core Data

6. The Hurricane cycle  
1. There is a very well known multi-decadal cycle with the hurricanes.  If you are wondering why 2006 was absent of hurricanes in the Atlantic (the pacific had plenty) please read about Calima.
7. The last of which are the climate models.  
1. The models put forth by James Hansen in 1988 have had a very good track record with recreating the past and predicting 20 years into the future.

A few analogies I've heard others use:

1. Predicting how hot it will get in your car today is not something a lot of us can do.  However, everyone knows that a car with a black interior and rolled up windows will be much hotter than a white car with it's windows open.  The greenhouse gases are like the windows.  And the black interior is similar to the melting ice sheets leaving behind darker rock. 
2. You can't predict the stock market on any given hour, but you can predict how the market or a certain portion of the market will do in a year given a certain set of conditions.  The hourly flux is like the weather, the year long projections (under certain conditions) are like the climate.

If you would like to contact us, suggest a topic to be covered, contribute a relevant commentary,
or be part of this effort on a more permanent basis, please email:




Unique IP visits: