GOAL: a strategy for actually initializing a
seasonal coupled forecast using a product from a
"LDAS" observed forcing/data assimilation system.
•NSIPP wants action on this, and is willing to
cooperate on trying a limited set of land initialization experiments
with their coupled model.
•This activity is central to the mission of the
LDAS team, so we should respond with a plan and some
action.
The current limits seem to be:
1.NSIPP will be doing forecasts using a 2x2.5 degree, tiled Mosaic coupled
model for at least the next year. So, this is the system we must provide
initial conditions for.
2.A long-term (~10 years or more) land state archive is required for
obtaining climatology for the initial land state. There is also a
desire to initialize in near-real time with the newest observed forcing and assimilated states. These two
desires are in obvious conflict, so we may need to do two sets of experiments to provide both a long-term archive and real-time ICs.
3.The desire is to have a reasonable land initial condition for use in S-I
initialization by fall 2001.
4.
-THIS WILL REQUIRE A REALISTIC, SIMPLE, Doable
APPROACH using EXISTING TOOLS.