Saturday, March 15, 2014

Prof. Inez Fung: "Anatomy of a Climate Model" the video


Climate models are constantly being misrepresented and the public has been sold impossible expectations by folks more interested in power-politics than learning about our planet.

Climate models are tools.  Tools to help scientists understand dynamics in action.  No one has ever pretended they could accurately prediction the future.  They inform us about dynamics and trends.  We human minds need to figure out the rest.  

One of the slickest, and sickest, tricks folks like McIntyre, SenInhofe, Morano, Heartland, and pals have pulled off, is misleading the public into believing that if climate models weren't perfect they deserve ridicule.  

Sadly the folks that attack climate models the most, are least interested in learning about them.  But, there must be people who do want to learn, 
I tell myself.

That is why I'm sharing this excellent talk by Professor Inez Fung.  


GoogleTechTalks  |  February 8, 2011  |  55:30 minutes


The first successful numerical weather forecast was made on the ENIAC (with fewer than 10 words memory) in 1950.  This talk traces the development of atmospheric General Circulation Models (GCMs) for weather forecasting, to Global Climate Models (GCMs) and Earth System Models (ESMs) to illustrate the guts and gore of the huge codes.  

If we cannot predict the weather beyond 2 weeks, what do we mean by climate projections for the next century?
~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~

Professor Inez Fung takes us on a fast paced tour of the complexities of climate models.  Along the way, she conveys the why and how of scientists figuring out the challenges of describing our global heat distribution engine mathematically.

Professor Inez Fung is speaking to a university audience so perhaps at moments she talks just at, or barely above, the heads of us regular folks.  Still, for those familiar with the basics of the science and a curiosity to learn more, you'll find this expert's talk fascinating and worth paying attention to.  

As part of my own learning process for especially good talks I make a rough time signature log for later reference and so I'll share that also.





~ ~ ~ ~ ~ ~ ~

{Video starts a little late and first words are chopped off - she's talking about the coordination between many science teams scattered around the globe...}

4:45  -  Charney's equations, describing our planet mathematically ...

5:30  -  "Reason for success in 1950"
"recognizing that you can do approximations to the equations of motions ..."

7:00  -  Review of equations ... and geophysical dynamics ...

8:30  -  "the bugaboo"  -  convective mixing and clouds ...

9:00  -  layers of complexity ...

10:00  -  and then there are the oceans, more challenges ...

11:50  -  "Numerical Weather Prediction" ...

14:00  -  models don't look at every "realization," they look at statistics.
Forecasting the mean and the variance ...  looking at boundary conditions ...

15:30  -  Improvements since 1950s
Modern climate models ...

16:30  -  Vertical resolution of modern climate models
50± layers in the lowest 10 to 15 kilometers
30± layers in the ocean
10± layers in the soil
New generation models will have horizontal resolution of 25km.
With very small time-steps between calculations. ...

{CC comment: Yes, these are tools, no one expects them to be crystal balls.}

17:45 -  Biological factors that are put in models
vegetation, growth, decomposition, carbon circulation, carbon in ocean, phytoplankton ...

18:30  -  Current codes contain on the order of a million lines of physics ...

18:45  -  "Variations on a Theme" ...

19:15  -  New approach ...  Dealing with geodesic gridding challenges ...

20:15  -  Dealing with wind ...

22:50  -  The physics of the Model
Explaining the details of Trenberth's Energy Balance chart
greenhouse gas physics ...

25:30  -  "When we look at climate change ..."

25:50  -  Mauna Loa Observatory, CO2 measurements 

27:00  -  ice cores ...

27:40  -  radiative forcing ...

30:30  -  aerosols ...

31:00  -  clouds ... uncertainties

32:50  -  climate forcing since 1800's ...

34:00  -  feedbacks and albedo ...

35:00  -  reviewing the Vostok record ...

36:45  -  climate sensitivity ...

38:15  -  cloud height changes forcing ...

39:00  -  coming back to the greenhouse gases ...

39:30  -  why experts are increasingly worried ...
"equilibrium warming"
equilibrium calculations ...

40:20  -  ocean dynamics...
feedback coupled with thermal inertia ...
heat that is in the pipeline

41:30  -  review ...
climate forcing, feedback, water cycle ...

41:40  -  looking at current climate models ...

43:20  -  IPCC Attribution graphs ...

44:10  -  Clouds and saturation pressure 
re. temp and height... migrating phase change ...

45:10  -  challenges in representing clouds ...

45:45  -  Precipitation ...
"Society doesn't care about half a degree changes in temperature.  Doesn't care about one degree change in temperature.  We care about changes in precipitation especially changes in precipitation extremes. ..."

46:10  -  when modeling, precipitation is a 2nd derivative of temperature...
wind is pressure gradient force, the 1st derivative of temperature,
convergence of lofting force ...

46:45  -  graph from IPCC report, "Projection of Precipitation" ...
agreement in the "sign" not the "magnitude" ...

47:15  -  Atmosphere's "Hadley Cell" is expanding...
"moving the sinking poleward branch towards us ..."
{inconsistent results among models ...  some show more expansion, some less, all show expansion!}

48:00  -  Soil Moisture ...
major challenge of climate modeling...
"How do we represent the water under our feet" ...

49:00  -  closing, "How can we predict the weather?"
equations...
"Deterministic Nonperiod Flow"


50:15  -  "Why should we be able to predict the weather? ..."
climate is the average of the weather ...

51:15  -  "Looking at the statistic of climate ..."
"Even with a chaos model, we can have a sense of predictability ..."

51:00  -  the trend...

52:20  -  "Challenges: observations, obs, obs... multiscale modeling, data assimilation, dynamical systems...

52:55  -  Keck HydroWatch Project ...

54:00  -  clouds and new observations ...
GoogleTechTalks  |  February 8, 2011  |  55:30 minutes

No comments: