12917 Decision Support for Water Resources Management
Engelsk titel:
Decision Support for Water Resources
Management
Sprog:
Point( ECTS )
5
Kursustype:
Ph.d.
Skemaplacering:
May 2014: Preparatory reading (approx. 40 hours)
19-23 May 2014: Full time teaching block at DTU (40 hours)
May-June 2014: Final assignment (approx. 60 hours)
Water is the planet’s most critical natural resource and must be
managed carefully. Mathematical models are key scientific decision
support tools in water resources management. However, predictions
of water resources models are highly uncertain because of
insufficient knowledge of model structure, parameters and
time-variable inputs. Data collection can reduce model uncertainty.
For efficiency reasons, the cost of additional data collection
should be traded off against the benefit of reduced prediction
uncertainty. Moreover, many decision problems in water resources
management are staged: Today’s decisions affect the available
options tomorrow; however, today’s decisions must be taken while
tomorrow’s conditions are still uncertain.
This course will use groundwater flow and transport models as an
example, but the concepts are general and can be applied to all
water resources models. The course will start with a revision of
hydrogeological modelling theory and concepts. An example
subsurface model will be introduced with a purpose of prediction
and management. Model predictive uncertainty will be quantified and
parameter estimation techniques will be used to inform models with
measurements. Model uncertainty will be propagated all the way to
economic decision criteria used in practical management.
Optimization methods will be employed to find sets of
Pareto-optimal decisions. Sequential decision problems will be
addressed with dynamic optimization methods. The value of
additional measurements will be determined from their impact on
decision making. All concepts will be applied to example problems
which will be solved by the students during the course in hands-on
workshops.
Læringsmål:
En studerende, der fuldt ud har opfyldt kursets mål, vil kunne:
Evaluate model uncertainty and sensitivity to parameters and
inputs.
Analyze data worth, and apply parameter estimation to calibrate
models.
Synthesize uncertain model predictions with economic analyses
for decision making.
Develop tools to evaluate best (optimal) outcomes while
satisfying multiple objectives.
Analyze decisions in dynamic, or time varying problems.
Apply the modelling concept from the course to groundwater
models.
-
-
Kursusindhold:
• Principles of subsurface flow and transport modeling
• Model uncertainty
• Sensitivity analysis
• Parameter estimation methods
• Model predictions and prediction uncertainty
• Decisions under uncertainty
• Multi-objective optimization
• Dynamic optimization methods
• Data value analysis
Litteraturhenvisninger:
The course material includes lecture notes, book chapters, journal
articles and computer code. The preparatory reading assignment will
be communicated to the students via campusnet before May 1st,
2014.
Mulighed for GRØN DYST deltagelse:
Kontakt underviseren for information om hvorvidt dette kursus giver
den studerende mulighed for at lave eller forberede et projekt som
kan deltage i DTUs studenterkonference om bæredygtighed,
klimateknologi og miljø (GRØN DYST). Se mere på http://www.groendyst.dtu.dk
