; Vote for Climate Modeling week #3 lecture topics and labs
Learning Center
Plans & pricing Sign in
Sign Out
Your Federal Quarterly Tax Payments are due April 15th Get Help Now >>

Vote for Climate Modeling week #3 lecture topics and labs


Vote for Climate Modeling week #3 lecture topics and labs

More Info
  • pg 1
									Vote for Climate Modeling week #3 lecture topics and labs
In the final week we have 5 lectures and 5 tutorials only. Here are things that we think are important for you to learn/practice: In modeling and data analysis: • Building a many-box model • Building a Monte Carlo model • Looking at time-series data and in crucial graduate school skills: • Reading real scientific papers • Critical thinking • Public presentation of results But unfortunately we don’t have time to cover everything. Rather than us just choosing for you, we’d rather have your input on what topics are of most interest & use to you. Please read the options below and comment or vote on them. We will need to decide by Monday what we’re doing in the last week.

Many options to choose from. Right now we are tentatively planning to do the first three listed below. It does seem important to cover the structure of big climate models before Mark Hadross from UCT comes to speak. Please put a * next to any of these that you are interested in. More **s if you are very interested. Put an X (or many XXs) if you are definitely NOT interested.


1. Modeling horizontal flows. Equations of motion of air and oceans. How to model transport as a many-box model. 2. How big climate models are built up. Making a many-box model. Working out grid size, structure of model, computational speed, tradeoffs, paramaterizations. Representing processes that are below grid scale. 3. Structure of the atmosphere (and what that means for rain & circulations) Why air is thinner when you climb mountains; why there is a cloud over Table Mountain; why Muizenberg is sometimes cloudy and sometimes sunny; how the Hadley circulation works. 4. Paleoclimate and time series analysis. What do records say about past climates. How do you get that information? Paleoclimate records from Ghana and Tanzania. Time series techniques: trend analysis, Fourier transforms. 5. Probability and uncertainty in modeling. When is your model output good? What is a bad model? What do you do if you’re not certain about your parameters? Varying parameters to get a probability distribution of outcomes (“ensemble forecasting” used in weather prediction). Models actually based on probability distributions (Monte Carlo models). 6. Working with data. Review of basic data analysis techniques nearly all grad students need: smoothing, fiting, interpolating, evaluating statistics. “Goodness of fit” of model to data. 7. Climate change policy. Estimating how much it would cost to suffer climate change from too much CO2. Estimating cost to reduce CO2 emissions. Finding an optimal strategy. 8. Flows of water between atmosphere, land, and oceans. Making a proper multi-box model out of the land system. Predicting zones of precipitation deficit. Changes in future climate regimes. 9. Ocean circulation. How and why water moves in the ocean. Why we care: ocean moves heat to poles (Northern people would shiver without ocean circulations). Why is water temperature different on the 2 sides of the Cape? How the ocean is like an upside-down atmosphere. 10. Review of some lecture topic that was unclear. Your pick. 2

Labs and final project
Please put a * next to any of these that you are interested in. More **s if you are very interested. Put an X (or many XXs) if you are definitely NOT interested. 1. Many-box model: 2D model of horizontal motions. Moving material around in the atmosphere or ocean. This project would approximate an important part of what a big climate model does. This lab would require you to make a 2-D array of boxes in x-y space so it’s a slightly bigger programming project. 2. Many-box model: 1D model of population.. You can build a fun many-box model that does demographics of a population, and watch populations aging as birth rates change. This isn’t exactly climate, but population is at leat somewhat related to climate. This lab also goes along naturally with some lessons in probability. 3. Many-box model: hydrological cycle. So far we made a simple 2-box model of water in the oceans and atmospheres. We can build up a bigger model by including land and ice as separate boxes. 4. Big climate models: Looking at output How do we expect climate to change with increased CO2? We can show you results of a complex model simulation. The model used has atmosphere, land and oceans and thousands of grid points. You could look at how temperature, evaporation, precipitation, winds, cloud cover, albedo, and ice cover are expected to change with increasing CO2. 5. Time series analysis One of the tutors last year wrote a Python routine to calculate Fourier transforms, so we can look at frequencies in time series for understanding past climates. We’d also look at the fit residual from your first lab (data - fit). 6. Critical thinking project: identifying bad papers Many scientists estimate that 90% of everything that is published is either irrelevant or wrong. The remaining 10% moves science forward. When you’re a graduate student, how will you tell good science from bad? How will you know which papers to trust? Last year our final project involved reading modeling papers. The class was divided into groups, and each group received 2 papers, one that I thought was good and one that I 3

thought was bad . . . but they didn’t know which was which. The groups read the papers and made a formal presentation of their conclusions to all of AIMS. 7. Critical thinking project: evaluating a paper closely We can also take one famous paper that has some good aspects and some bad aspects and have a public debate on it. The class would be divided up into groups, and some would be assigned to argue that the paper is valuable and others to point out its weaknesses. 8. Reading, data interpretation and presentation Look at model output, investigate predictions for a particular country or region, and make a presentation to the class about it, discussing the forecast and uncertainties.


To top