[CCSM-LandWG] Draft Agenda - CCSM Land Model Working Group

[Lydia Shiver]

CCSM Land Model Working Group Meeting
Monday, 14 March, 2005 and Tuesday, 15 March 2005
NCAR, Mesa Lab, Chapman Room

Agenda

Purpose – The hydrologic cycle in the Amazon Basin is poorly simulated 
in CAM3/CLM3 (prescribed SSTs and sea-ice) as well as in CCSM3 (active 
ocean/sea-ice). Precipitation is severely underestimated. The 
partitioning of this precipitation into evaporation, transpiration, soil 
moisture, and runoff is also deficient.  Several model developers have 
proposed specific changes to CLM3 that address some or all of these 
biases and improve the hydrologic cycle in the Amazon. This includes, 
e.g., subgrid-scale distribution of precipitation, hydraulic lift of 
soil water by roots, runoff parameterizations, soil water effects on 
transpiration, enhanced infiltration or macropores, and canopy 
processes. The goal of this workshop is to review the proposed changes 
to CLM3 and consider additional changes, identify promising 
modifications, and to develop a common pathway to systematically and 
sequentially test the changes so as to determine a suite of 
modifications that will be the basis for a future version of CLM. It is 
hoped that this testing can be accomplished by CLM support staff housed 
at NCAR during April-June so that a report on the model testing and a 
proposal for a modified CLM can be presented at the annual CCSM meeting, 
21-23 June 2005.

The first day of the meeting will provide a general overview of methods 
and results and establish the basis to proceed with model testing. 
Invited speakers will have an opportunity to talk about their proposed 
changes, insights to the hydrology of the Amazon basin they have gained 
from their work, or datasets and procedures for testing the model (e.g., 
analyses of the hydrologic cycle). Additional speakers can contribute 
other proposed modifications in brief comments. Significant time is set 
aside in the afternoon for discussion of how to proceed on a common 
pathway. Criteria that must be satisfied to be included in the common 
pathway are: a) it solves an existing serious deficiency; b) it is easy 
to code or sample code has been tested; and c) it has already been 
tested and shown to work.  New ideas with no code or testing will be 
recognized, but will be left for further individual development.

On the second day, model developers will be expected to describe in 
detail the proposed changes to CLM3. As an outcome of this meeting, NCAR 
staff expect to have the necessary information to implement the proposed 
changes in a standard version of CLM3 so that the changes can be 
systematically and sequentially tested and documented. Model developers 
are expected to guide the CLM support staff through the proposed 
changes, highlighting lines of code that have changed, new lines of 
code, and summarizing this in an “implementation guide.” Access to this 
implementation guide is expected prior to the meeting to ensure a 
high-level of software engineering discussion on the second day. Each 
speaker’s implementation guide will be distributed among the invited 
model developers prior to the meeting.


Monday, 14 March 2005

08:30 – 09:00 Continental breakfast
09:00 – 09:15 Opening remarks and introduction
0:915 – 09:45 Zong-Liang Yang
09:45 – 10:15 Jung-Eun Lee
10:15 – 10:30 Break
10:30 – 11:00 Bob Dickinson
11:00 – 11:30 Peter Thornton
11:30 – 12:00 David Lawrence
12:00 – 13:00 Lunch
13:00 – 13:30 Randy Koster
13:30 – 14:00 Aiguo Dai
14:00 – 14:30 Ed Beighley/Ken Eggert
14:30 – 14:45 Break
14:45 – 17:00 Discussion
Parameterizations
Additional changes to CLM – 5 minute presentations by attendees
What parameterizations look promising to implement?
How might they interact with other changes to the model?
Do they each solve the same “problem” but with a different approach?
What processes have not been discussed?
Testing protocol
		How do we systematically evaluate the agreed upon
		changes?
		What experiments need to be run?
			e.g., offline tower flux, offline global
			simulation, CAM3/CLM3
		What datasets are available to test the model?
			Tower flux, large-basin to continental scale
How do we ensure improvements to the Amazon do not degrade the 
simulation elsewhere?

Tuesday, 15 March 2005

08:30 – 09:00 Continental breakfast
09:00 – 09:30 Unresolved business from yesterday afternoon
09:30 – 11:00 Code show and tell
11:00 – 11:30 Timeline for experiments
11:30 – 12:00 Other CLM development
		Prognostic canopy air space
12:00 		Adjourn

Gordon Bonan