Summer School on Global Sustainability-Projects & Working Groups
From Santa Fe Institute Events Wiki
|Summer School on Global Sustainability|
ENGLAND & NAKICENOVIC: Bush, Cullenward, Frisch, Nguyen, Nkem, Robinson
07/15/09 - Climate models
An overview was first given about the definition, classification, properties and benefits of a climate model. The process of building a model extends to verification, validation and assessment. Computer models are often preferred for their advantages in terms of expense, comprehensiveness, logic, accessibility, flexibility (capable of examining, easy for sensitivity analysis, parameter tuning etc).
When it comes to modeling climate change we talk of models of the ocean and atmosphere. There are 6 governing equations: for ocean there are 3 of momentum, conservation for heat, salt, and mass and the same is for the atmosphere with humidity instead of salt. Forcing conditions are added to represent only processes at the top of the atmosphere for full blown climate models. Starting with initial conditions we solve the governing equations over a discrete grid using observations in the forcing conditions and then integrate the solutions forward in time. Running submodels or subcomponents of the Earth requires many more forcing conditions that need to be addressed. Hence it is harder to run submodels rather than a whole model.
What kind of grid is employed determines what kind of dynamics it will represent effectively. In the late 80s, the resolution is (4 degrees by 4 degrees) 400km by 400km. Now it is about 1 degree by 1 degree. When we can't get what we want, we go down to a finer resolution by nesting a sub-grid of finer scale on top of the coarse grid. Sometimes we have to change the model or some models give us the option of where to zoom in. The option of nesting different resolutions allows for adding the affects of cities, etc that are quite small but can effect local weather/climate significantly.
The vertical vs. horizontal resolution is another issue. A spectral grid is such that the lines are parallel along the y axis but bend along the x axis. 1D e.g. that about an air column in the atmosphere and 2D models e.g. Henry Stommel's 2-boxes models still have good uses on their own. 3D models employ spectral grids which minimize numerical erros in atmospheric models.
The development of grids for use in climate modeling exemplifies only part of the progress since the 1970s. It would be interesting to track down the progress, quantitatively and qualitatively, of the scientific modeling community. What they have in common is for the most part models are run on FORTRAN90. The language is better in picking out errors in floating points. It is worth noting that the process of running a model to get its output and spatial visualization are separate. Having models is one thing, having interfaces where researchers will simply walk in and use is another option and there are some good ones. There are good models and fake ones that look like real. This is not a parade. Open access is important for refining models and sharing ideas in modeling climate changes.
Model assessment makes use of CFCs which invade the ocean in very specific way and hence can act as tracers in ocean circulation experiment. As these were introduced, the presence/level of CFCs in deep water provides an excellent sense of movement and mixing. This can be used to double-check the accuracy of ocean models, being more precise than temperature and salinity measurements. Among uncertainties in existing models, Greenland ice is a big uncertainty (1.6m is likely sea level rise, 85cm from land ice). Another is sulfite areosols–a pollutant that has helped to keep things cooler. Our effort to decarbonization will reduce their presence in the atmosphere, the impact of which is again unknown.
RUSSELL & GRÜBLER: Bamutaze, Borgeson, Engler, Ona, Pasqualini, Zellner
Media:Russel_summary.doc Here is the summary of Joellen Russel's climate modeling talks.
Here is the summary of Arnulf Grubler's talks.
EDENHOFER & PAUL: Bottrill, Brelsford, Doshi, Geddes, Gong, Zaks
DASGUPTA & LOVINS: Clewlow, Dangerman, Hunt von Herbing, Kane, Sammeth, Wolf
HARGADON & MEADOWS: Gourdji, G. Jones, McInnis, Morgan, Mueller
RUBBIA & KUTSCHER: Gonzales, Hagerman, A. Jones, Krakauer, LaCerva, Lacroix
Active, assertive participation by students is an intrinsic part of this event. Attendance at all program sessions is mandatory. Participants have been divided into six groups; each group designed to be as topically and geographically diverse as possible. Chairperson and rapporteur roles may rotate within each group.
Groups have several tasks. One function is to discuss each day’s lectures among themselves and to be prepared to actively participate in the group discussion sessions each day. In addition, each group will be tasked to write a summary of the main points of the two lecturers, one from the first week and one from the second week. Responsibilities may be divided (that is, each of the six students might be responsible for producing a lecture transcription, editing duties could be divided, etc). The lecture summaries should include comments by the group on the broader implications of the lectures, critical analysis of the research area covered, and resonance to other presentations. Each speaker will be available to meet with the groups covering his/her lectures. This material may be edited and produced in book form.
Additionally, each group will be asked to produce a research agenda as a product of the school. This would not be limited to the lectures each group covered, but rather focus on the entire two-week agenda and identify synergies, areas of disagreement, and gaps in our knowledge that can be resolved by future research. The final day of the school will include a plenary session during which each group’s rapporteur presents the research agenda. Finally, we will draw on these agendas to produce an open letter to President Obama or Science Advisor John Holdren.