|
||||
Section Contents
FluxNet Land surface modelling meeting
The meeting report has been published in FLUXLETTER (Vol. 1 No. 4) and is available here.Paper published online: Improving land surface models with FLUXNET data, M Williams et al., Biogeosciences Discussions. The paper is now accessible and open for Interactive Public Discussion until 02 May 2009 at: http://www.biogeosciences-discuss.net/6/2785/2009/
Revised document
Comments document
CarbonFusion also provided sponsorship for the FLUXNET Asilomar modelling workshop in February 2009.
University of Edinburgh, School of GeoSciences, Crew Bldg, Room 304
4-6 June 2008
Objective
To resolve the details of a coordinated approach to calibrating Land surface models from FluxNet data using data assimilation (DA).
Outputs
- A workplan for coordinating international resources for an LSM calibration exercise using FluxNet data
- A publication demonstrating the utility of FluxNet data for model calibration
Key questions
- How many models should be calibrated? One or many?
- What calibration approach should be used? Do different approaches have different strengths and weaknesses?
- How should observation and model error be determined and accounted for?
- How do we deal with observational or model bias?
- Is the objective to retrieve a full quantification of land surface exchange for the observed period (state estimation), or to generate a forecast model (parameter estimation)?
- How do we address the initial condition problem?
- How do we detect and address model flaws?
- How many and which towers should be used?
- How should the flux data be processed?
- What ancillary data (including EO) can and should be used?
Presentations
Welcome and introduction, Mathew WilliamsUsing FluxNet data to evaluate land surface models, Ray Leuning
Upscaling from eddy tower to continent: how to treat uncertainties in model parameters, Enrico Tomelleri
Some thoughts on errors and uncertainties in modeling and measuring surface-atmosphere exchanges, Andrew Richardson
Biases in land surface models’, Ying Ping Wang
Can we use flux data for reducing Land Surface model uncertainties?, Markus Reichstein
Building from the bottom-up and learning as we go: data requirements for upscaling ecosystem function, Paul Stoy
Some challenges of data-model integration, Martin Jung
Parameter identifiability, constraints, and equifinality in data assimilation with ecosystem models, Yiqi Luo
Steady state impacts in inverse model optimization, Nuno Carvalhais
Optimising ORCHIDEE simulations at tropical sites, Hans Verbeeck
Calibration of land surface models, Cathy Trudinger
Flux data to detect land surface model deficiencies and their complementarity with space-borne observations, Phillipe Peylin
Results from the Reflex experiment, Mathew Williams
Can satellite observations of biomass improve land surface models? Shaun Quegan
Location
Our meeting starts at 9 am on Wednesday 4 June.
Our venue is Room 304, Crew Bldg, Kings Buildings Campus, University of Edinburgh.
Hopefully this is within walking distance of your accommodation.
You can find a range of maps at this link
Crew Bldg is number 5 in the map here
If you have any problems you can call me at
0131 667 8631 (home), 0131 650 7776 (work) or 0775 360 3349 (mobile).
You can find travel information here
If you are arriving at the airport you may want to take a taxi to your hotel/b&b. Please keep all receipts.
Attendees
Mathew Williams (Edinburgh) Organiser
Ray Leuning (CSIRO)
Andrew Richardson (UNH)
Hans Verbeeck (Antwerp/LSCE)
Yiqi Luo (Oklahoma)
Nuno Carvalhais (Lisbon)
Ying Ping Wang (CSIRO)
Markus Reichstein (MPI)
Paul Stoy (Edinburgh)
Cathy Trudinger (CSIRO)
Philippe Peylin (LSCE)
David Pearson (UKMO)
Enrico Tomalleri (MPI)
Shaun Quegan (Sheffield)
Martin Jung (MPI)
Relevant papers
Abramovitz GRL 2005Santarem GBC 2007
Jung GBC 2007
Williams GCB 2005