701-1252-00L Climate Change Uncertainty and Risk: from Probabilistic Forecasts to Economics of Climate Adaptation

Course: Monday 8-10 CHN E42

Exercise: Monday 10-12 NO D39

Lecturers: Prof. R. Knutti, IAC ETH, Dr. David Bresch, SwissRe

Assistants: Markus Huber (markus.huber[at]env.ethz.ch), Omar Bellprat (omar.bellprat[at]env.ethz.ch).

Details from the ETH Course Catalogue

Contents

The course introduces the concepts of predictability, probability, uncertainty and probabilistic risk modelling and their application to climate modeling and the economics of climate adaptation.

Students will acquire knowledge in uncertainty and risk quantification (probabilistic modelling) and an understanding of the economics of climate adaptation. They will become able to construct their own uncertainty and risk assessment models (MATLAB), hence basic understanding of scientific programming forms a prerequisite of the course.

The first part of the course covers methods to quantify uncertainty in detecting and attributing human influence on climate change and to generate probabilistic climate change projections on global to regional scales. Model evaluation, calibration and structural error are discussed. In the second part, quantification of risks associated with local climate impacts and the economics of different baskets of climate adaptation options are assessed – leading to informed decisions to optimally allocate resources. Such pre-emptive risk management allows evaluating a mix of prevention, preparation, response, recovery, and (financial) risk transfer actions, resulting in an optimal balance of public and private contributions to risk management, aiming at a more resilient society.
The course provides an introduction to the following themes:
1) basics of probabilistic modelling and quantification of uncertainty from global climate change to local impacts of extreme events
2) methods to optimize and constrain model parameters using observations
3) risk management from identification (perception) and understanding (assessment, modelling) to actions (prevention, preparation, response, recovery, risk transfer)
4) basics of economic evaluation, economic decision making in the presence of climate risks and pre-emptive risk management to optimally allocate resources.

Old Semester Outline 2010

Exercises every two weeks approximately. Powerpoint slides will be provided here.

• 1.3.2010:  Logistics, Introduction to probability, uncertaintyand risk management, introduction of toy model. Slides
  Exercises 1: Exercise 1 description, Matlab data file
  
• 8.3.2010: Model calibration, Bayesian methods for probabilistic projections. Slides
• 15.3.2010: Predictability of weather and climate, seasonal prediction, seamless prediction. Slides
  Exercises 2
  
• 22.3.2010: Detection/attribution, forced changes, natural variability, signal/noise, ensembles. Slides
  
• 29.3.2010: Model evaluation, multi model ensembles and structural error. Slides
  Exercises 3
  
• 12.4.2010: Presentations of toy model work, discussion
• 19.4.2010: Risk assessment model: from concept to concrete application.
  Slides Holland 1980 Landfall Holland Exercise 2 Burning_cost_rating.xls
  Natural catastrophes and reinsurance
  Exercises 4
  
• 26.4.2010: Climate change impact scenarios: How to represent climate impacts in the hazard module – and some insurance basics
  Slides SwissRE: Floods are insurable SwissRE: Storm Damage
  
• 3.5.2010: Basics of economic evaluation and economic decision making in the presence of climate risk
  Survey Slides
  Exercises 5 Exercise 2 Cost Benefit XLS
  
• 10.5.2010: The cost of adaptation - application of economic decision making to climate adaptation in developing and developed region
  Slides Exercise 2 SwissRE ECA report
  Exercises 6
  
• 17.5.2010: Shaping climate-resilient development – valuation of a basket of adaptation options
  Slides
  
• 31.5.2010: Presentations of full toy model, discussion

Credit points

Hands-on experience with probabilistic climate models and risk models will be acquired in the tutorials; hence basic understanding of scientific programming forms a prerequisite of the course. Basic understanding of the climate system, e.g. as covered in the course 'Klimasysteme' is required.

Examination: graded tutorials during the semester (benotete Semesterleistung) and short presentations

Presentations April 12, 2010

• Groups of max. three people
• 5 minutes (strict timing, 5 slides max), ideally presented by the whole group, Powerpoint or not
• Exciting presentation with a clear take home message (assume everyone is familiar with the toy model and with was discussed so far)
• Show us one interesting/new/unexpected/important/exciting aspect relevant to the problem.
• Criteria: Timing/contributions from each person (25%), delivery/presentation (25%), content (50%)

Possible topics

• Choice of a model or model structure
  (linear or not? Trend in variance? Statistical or dynamical)
• Parameter estimation/calibration (methods, uncertainty)
• Monte Carlo simulation, probabilistic models
  (sample size, consideration of parameter uncertainty)
• Damage functions (Nordhaus, Stern)
• Mitigation (cost, strategy, cost dependence on scale)
• Cost benefit analysis (appropriate or not, difficulties)
• Precautionary principle
• Learning with more data, robust decisions
• Overall uncertainty analysis
  (what are the most relevant choices and uncertainties) 

Presentations May 31, 2010
Topics and rules