0 - next phone conference
Thursday, June 17th, 4pm. Dag will send
information on
dial-in.
1 - status
of submissions, reviews, feedback, revisions on all the JGR
NLDAS papers.
Cosgrove et al., forcing
paper, accepted
Cosgrove et al., spin up
paper, in review
Robock et al., accepted
Luo et al., revision sent
out
Ming et al., rejected
Sheffield et al., accepted
Lohmann et al., under review
Mitchell et al., under review
Pinker et al., accepted
2 - reruns
and re-VALIDATION of changes to Noah, VIC, Mosaic and SAC for the 3-year
retrospective to reduce systematic model errors that the JGR Papers uncovered
(will
Mosaic: update ground heat
flux (heat capacity)
- parameter
calibration later this year
run - Randy's
traditional soil layer concept
Noah: work on snow
parameterization (snow fraction, albedo)
- run version
2.6
- work within
Mopex for baseflow parameters
VIC: latent heat
underestimated, problem was the use of calibrated parameters of the 50 year
retrospective runs from UW
Sac: time needed to
re-specify Sac model for re-runs
3 - Do
branching out:
common forward radiative transfer
model from
common Kalman filtering from NASA?
4 - How do we go about
re-processing the 3-year retrospective precip forcing
to include PRISM and Schaake analysis method? How do we eliminate the holes of
missing data
left behind by the Schaake analysis method?
fill in from old Cressman scheme
John and Yun
will email their ideas
NASA will post re-processed
forcing with PRISM based precip
Dag will contact Wei Shi about
the precip interpolation to be used for retrospective
runs.
5 - Inform the group of the
GOES solar insolation problems during Feb 02 thru Mar
03 due to snow cover problems, and then from 01 Apr 03 to around 01 Jul
03 owing to change in physical satellite of
GOES East and lack of split-window channels for skin temp on GOES East.
new technique is needed for new GOES satellites, quality
of new product will be seen.
6 - Addressing the high
morning bias in GOES-East solar insolation even during
the 3-year retro period when quality of latter was presumably at its best.
Lifeng's paper shows that that GOES vs
Mesonet shows high bias in the morning. Problem needs
to be brought to Rachel's attention.
Timeline for re-runs:
ARM/Mesonet
limited area re-runs (flexible dead-line, around October)
1) calibration/physics
improvements from JGR runs
2) runs
with unified parameters/parameterizations/model set-up
more reduced runs are necessary (with areas to be defined)
complete NLDAS re-runs
1) re-runs
need to be defined
2) groups
want to concentrate more on science -- complete re-runs
- have to
wait
We need a unified validation
package.