Modelling protocols Marcel van Oijen (CEH-Edinburgh) With input from participants Garmisch Mar & May...
-
Upload
alexandra-cobb -
Category
Documents
-
view
213 -
download
0
Transcript of Modelling protocols Marcel van Oijen (CEH-Edinburgh) With input from participants Garmisch Mar & May...
Modelling protocolsModelling protocolsModelling protocolsModelling protocols
Marcel van Oijen (CEH-Edinburgh)
With input from participants Garmisch Mar & May 2006, Protocol-GMP team, Protocol-UQ/UA team
The protocols of NEUThe protocols of NEUThe protocols of NEUThe protocols of NEU
Common Modelling Protocol Common Measurement Protocol
UQ/UA protocol
NEU planned to produce 4 protocols:
Month 7 Measurement
Month 9 Modelling
Month 10 Uncertainty quantification & analysis in regional Nr and GHG models
Month 18 Uncertainty quantification & analysis in plot scale models
Discussion Garmisch (Mar & May 2006): The two uncertainty protocols will be combined.
The protocols of NEUThe protocols of NEUThe protocols of NEUThe protocols of NEU
Protocol-discussion:
- Garmisch Mar 2006
- Garmisch May 2006
- Documents sent for discussion (minutes from meetings, protocol-outlines 4 Aug 2006)
- Protocol-GMP team (Marcel van Oijen, Reiner Wassman, Pete Smith, GertJan Reinds, Peter Verweij)
- Protocol-UQ/UA team (Marcel van Oijen, Gerard Heuvelink, Hans Kros, Kairsty Topp, Benoit Gabrielle, Mette L.)
The 2 modelling protocols of NEUThe 2 modelling protocols of NEUThe 2 modelling protocols of NEUThe 2 modelling protocols of NEU
Conclusions:
- Protocols should not be too prescriptive
- Protocols should be short and refer to existing protocols
- Protocols should make it easier for people to work to a high standard
Protocol-discussion:
- Garmisch Mar 2006
- Garmisch May 2006
- Documents sent for discussion (minutes from meetings, protocol-outlines 4 Aug 2006)
- Protocol-GMP team (Marcel van Oijen, Reiner Wassman, Pete Smith, GertJan Reinds, Peter Verweij)
- Protocol-UQ/UA team (Marcel van Oijen, Gerard Heuvelink, Hans Kros, Kairsty Topp, Benoit Gabrielle, Mette L.)
The 2 modelling protocols of NEUThe 2 modelling protocols of NEUThe 2 modelling protocols of NEUThe 2 modelling protocols of NEU
Conclusions:
- Protocols should not be too prescriptive
- Protocols should be short and refer to existing protocols
- Protocols should make it easier for people to work to a high standard
Draft protocols for modelling 18 Sep 2006Draft protocols for modelling 18 Sep 2006Draft protocols for modelling 18 Sep 2006Draft protocols for modelling 18 Sep 2006
Protocol-GMP (3 pp.) Protocol-UQ/UA (5 pp.)
Protocol-UQ/UA – page 1Protocol-UQ/UA – page 1Protocol-UQ/UA – page 1Protocol-UQ/UA – page 1
1.1 Introduction
1.2 Web links to established protocols
Protocol-UQ/UA – page 2Protocol-UQ/UA – page 2Protocol-UQ/UA – page 2Protocol-UQ/UA – page 2
2. Relevant NEU-documents
3. Glossary
4. Protocol summary
4.1 Commitments
Protocol-UQ/UA – page 2Protocol-UQ/UA – page 2Protocol-UQ/UA – page 2Protocol-UQ/UA – page 2
4.1 Commitments
In NEU, all modellers are committed to:Carry out UQ of their model and report: (1) The chosen UQ-method, (2) Output uncertainty. The UQ-method is chosen by the modeller but can be one of the recommended methods described in the following sections.UQ is carried out at least twice: near beginning (year 1 for plot-scale modellers, year 2 for regional scale modellers) and end of the NEU-project, but it is recommended to do this whenever significant new data have become available and whenever the model has changed.
Protocol-UQ/UA – page 3Protocol-UQ/UA – page 3Protocol-UQ/UA – page 3Protocol-UQ/UA – page 3
4.2 Recommended work
In NEU, all modellers are recommended to:- Carry out UA for their model and report: (1) The chosen UA-method, (2) The inputs that account for most of output uncertainty. It is recommended to carry out UA after each UQ and use the results for improvement of the modelling as well as to guide the collection of data.- Carry out a model comparison, possibly in collaboration with other groups, and report: (1) The method used for model comparison, (2) The relative probabilities of different models, model versions and/or model algorithms (i.e. submodels for specific processes or groups of processes) of being correct. Systematic comparison of completely different models is recommended for near the end of the project, systematic comparison of different model versions is recommended to be repeated throughout the period of model development in the first years of NEU.
Protocol-UQ/UA – page 3Protocol-UQ/UA – page 3Protocol-UQ/UA – page 3Protocol-UQ/UA – page 3
5 UQ model inputs
5.1 Use of pdf’s
5.2 Expert judgment
5.3 From data to pdf
5.4 Bayesian approach
6 Uncertainty propagation
6.1 Monte Carlo (MC)
Protocol-UQ/UA – page 4Protocol-UQ/UA – page 4Protocol-UQ/UA – page 4Protocol-UQ/UA – page 4
6.2 Latin Hypercube sampling
6.3 Screening
7 When new data come in
7.1 Repeat UQ
7.2 Bayesian Calibration (BC) and MCMC
7.3 BC: universal vs site-specific inputs
8 Uncertainty about model structure
8.1 Bayesian Model Comparison (BMC)
Protocol-UQ/UA – page 5Protocol-UQ/UA – page 5Protocol-UQ/UA – page 5Protocol-UQ/UA – page 5
8.2 MODEVAL.XLS
10. References
9 Uncertainty Analysis (UA)
Garmisch, “Uncertainty workshop” May Garmisch, “Uncertainty workshop” May 20062006
Garmisch, “Uncertainty workshop” May Garmisch, “Uncertainty workshop” May 20062006
Model output uncertainty
Data uncertainty
Input uncertainty
Structural uncertainty
Measurement acc. & prec.
Sampling / representation
• Parameters (initial constants, process coefficients)
• Environmental scenarios (weather, N-dep., CO2, land-use
6.5 NEU C3: work on uncertainty6.5 NEU C3: work on uncertainty6.5 NEU C3: work on uncertainty6.5 NEU C3: work on uncertainty