December 05, 2022

How can we predict the climate 50 years from now if we can't forecast the weather next week?

This is a question asked by a commenter, Skeptic517, on my blog yesterday.

It is a good question, one that I suspect most climate activists can not answer.  

Certainly, none of the responders knew the answer.

So what do you think?   Are long-term climate predictions hokum?

And now the answer.

Climate prediction is possible and reasonable because the nature of the prediction is very different than weather forecasting.

If you like technical terms, forecasting weather is an initial value problem, while climate prediction is predominantly a boundary value problem.  I will explain this in a second.

And the nature of the prediction is very different. 

For weather prediction, we predict the exact state of the atmosphere at a certain time.  The high in Spokane will be 68F next Thursday.   The low center will be 978 hPa and located in central Iowa.    Specific in time and space.

An 84-h weather forecast of sea level pressure (solid lines) and precipitation (shading)

For climate prediction, we don't forecast the surface weather map for 4 AM on January 4, 2090.  That would be nonsensical.

Instead, we forecast mean conditions, often over a broader area.   Will the average temperatures in spring over Washington State be warmer or cooler than current values?  Will the precipitation averaged over ten years of winter be greater at the end of the century than the recent ten-year average?

Projected change in annual mean surface air temperature from the late 20th century to the middle 21st century


That kind of thing.

The essential insight you need is to understand the different natures of weather versus climate prediction.

Weather forecasting:  an initial value problem

Weather prediction, an initial-value problem, starts with a comprehensive, 3-D description of the atmosphere called the initialization.   Then large supercomputers are used to solve the equations describing atmospheric physics to forecast the exact state of the atmosphere in the future at specific times. 

Forecast accuracy declines with time and by roughly two weeks nearly all predictability is lost.....something described theoretically by Professor Edward Lorenz of MIT.    

Here is a plot of the loss of forecast skill over time for the U.S. models(for around 18,000 ft over the northern hemisphere).  The blue color is for the leading U.S. global model, the GFS. Forecast kill drops rapidly between 5 and 10 days.


Climate forecasting:  a boundary value problem

Forecast skill for specific weather features is lost after roughly 2 weeks because the atmosphere essentially loses memory of the initial observed state of the atmosphere.

In climate forecasts for extended periods of time, the key constraint is not the initial conditions, but the amount of radiation coming into and out of the atmosphere.    If we know, how much radiation is coming into and out of the top of the atmosphere, the climate models can produce a realistic average climate for those conditions.

The amount of radiation emitted and absorbed by the atmosphere is greatly controlled by the composition of the atmosphere....which we have to assume (e.g., how much CO2, methane, and particles in the atmosphere).

Such projections are only as good as our estimate of the amount of greenhouse gases in the atmosphere in 50 or 100 years.   Big uncertainty!  But we do the best we can.




So climate prediction DOES make sense.

I am skipping some subtleties: for example, the initial state can have some influence on climate simulations.   But you can try to deal with that issue by running an ensemble of many climate predictions each starting slightly differently.

Anyway,  now you know the answer!




December 03, 2022

A Forecast Failure and the Next Snow Event

 The weather prediction business is a humbling one.     

Yes, forecasting systems have gotten far better, but occasionally they disappoint, particularly for difficult situations like this morning's snow event.

The model runs on Friday indicated that Puget Sound country was on the edge for sn ow, both in terms of temperature and precipitation.

Below is the precipitation forecast (amount of liquid water predicted for the 24h ending 4 AM today--Saturday.   No precipitation over Seattle's eastern suburbs and around 0.15 inches over the western side of northwest Seattle.   

With the typical ratio of 10 to 1 for snow to rain amounts, NW Seattle should have had only had about 1.5 inches at most.  Substantial portions of north Seattle had 4 inches.  For wet snow, the ratio is even lower--so there was clearly a substantial underestimate of precipitation.

Here are the observed amounts for the same period.  Ouch.

Much more precipitation than predicted over Seattle and the eastern suburbs

The models had predicted a sharp eastern edge of the rainfall (over Seattle), but as shown by the radar last night around midnight, a broad area of precipitation extended over the region.


What made this situation particularly disturbing, was that our primary tool for spotting weather prediction outliers.....ensemble prediction running the model many times, each slightly differently--did not effectively pick up on this wet solution.  

The figure below shows the predicted precipitation at SeaTac for the many ensemle members (the various lines) and actual precipitation is shown by the purple circles.  Only one of two dozen ensembles got it right.   The average of all the ensembles, the ensemble mean, shown in black, predicted less than half of the observed precipitation.


This is a rather large and unusual failure for such a short forecast.   In any case, one can learn from failures and I will be examing this one carefully.

Finally, a fascinating aspect of this event was the extreme variability of the snow, driven by temperature variations with height.

 Locations in Seattle near sea level had only a light dusting, while just a few miles away there was 4-5 inches of heavy, wet snow.   

Sobered by this failure I drove around Seattle this morning.   Near sea level on Sand Point Way (NE 90th), there was nothing.  


But as a drove to Lake City , a few hundred ft above SL, it was all white with a few inches of slush.  By  Aurora avenue, around 4 inches.


Sunday's Snow Event

Clouds and precipitation are starting to swing northward around the low-pressure area off of California (see latest infrared satellite image).


The latest model runs brings some light snow into Oregon and southern Washington tomorrow, as illustrated by the total predicted snow accumulation through 1 AM Monday.


I am starting to get tired of these marginal snow situations......



How can we predict the climate 50 years from now if we can't forecast the weather next week?

This is a question asked by a commenter, Skeptic517, on my blog yesterday. It is a good question , one that I suspect most climate activist...