ASIS:Minutes 19 03 2008

Understanding contemporary change in the West Antarctic ice sheet Project meeting, 19th March 2008, UCL Notes

Present: 	Tony Payne, Anne Le Brocq, Andreas Vieli, Poul Christoffersen, Marion Bougamont. Copied to: 	Gwendolyn Leysinger-Vieli, Steve Price, Rupert Gladstone, Vicky Lee

Agreed actions for next meeting

Geometric data sets.
Create base data sets on 5-km projected grid (preferably as ASCII files). Data to include upper ice surface, bedrock and lower ice surface (i.e., base of ice shelves). Ice surface may also be available at 1 km.

Actions: a.	AJP to ask JLB about latest ice surface. b.	AGASEA and BEDMAP thickness data sets to be used with JLB ice surface. c.	ALeB to work on combining data sets in Arc (i.e., blending AGASEA into BEDMAP and finding bedrock/lower surface from thickness and upper surface). d.	tune firn density/depth model so that gridded grounding line agrees with ADD vector data.

Boundary condition data sets.
Data required on same 5-km projected grid for annual mean surface air temperature, snow accumulation rates and geothermal heat flux.

Actions: a.	GLV to check on accumulation and temperature data with BAS. b.	need to worry about ocean areas (i.e., integrating with ocean surface data sets).

Geology data set.
Template with geological domain. Probably 3 classes are sufficient: soft (i.e., Siple Coast marine sediment); intermediate (harder or more patchy till); and exposed hard rock. Geothermal heat flux could also be based on this mask. Are three categories sufficient?

Actions: a.	MB/PC to compile using base geometry data set. b.	Need to liaise with JD about ocean areas.

Basal processes model.
Coupling between basal processes model and higher-order flow model is now working.

Actions: a.	Could MB/PC send pdf of the model document and relevant code? (I will set up a private wiki so that we all have access to these docs) b.	Potential paper on dynamics of ice streams with coupled basal processes-higher order model in idealised geometry (requires thickness evolution, see below). c.	AJP to send information on previous attempts to model coupled system.

Ice-flow model.
3d, first order model is now more or less complete. SFP has been working on lateral and vertical boundary conditions.

Actions: a.	SFP to send details of Bueler thickness solver. b.	SFP to send details of basal traction implementation and AJP to write up ideas on tanh method. c.	Decision of whether to use GLAM or GLIMMER as the base model. This will depend on how nested grids are used in the project. Depending on decision either put higher-order stresses etc into GLIMMER or improve temperature etc in GLAM. d.	Decision on which nested grid library to use or whether to implant own version. e.	RG to look at using PETSc for matrix inversion and automatic differentiation. f.	Potential paper using 1-d flow model comparing nested grids with stretched coordinates. g.	RG/SFP to look into thickness solvers and decide whether Bueler is sufficient.

Data assimilation.
Actions: a.	Potential paper on data assimilation on whole PIG system using existing code and optimizing for basal traction (or beta-sq) and ice rheology with AGASEA data. b.	Need to decide who to approach for velocity data for PIG ice shelf and other Amundsen Sea ice shelves (also Ross?). c.	Work on parameterization of fracture in shear margins.

Longer-term objectives
1.	Model of real Siple Coast region ice stream from divide to ice shelf? For instance B/C interaction? 2.	Sort out stability issues and decide whether to allow water pressure to vary. Decide who to couple regional water flow with pore water. 3.	Data assimilation to include ice thickness. 4.	Model initialization procedure.

Next meeting in Cambridge during week beginning 2nd June 2008