LDAS Meeting Minutes
August 20th
2001
NCEP/EMC
I. STATUS OF REALTIME N-LDAS RUNS
Each of the four LDAS modeling groups reviewed the status of their realtime LDAS model executions at NCEP, using NCEP-produced hourly surface forcing on the common LDAS grid. The NOAH, SAC and VIC LSMs continue to execute in realtime to the present date, with NOAH and SAC having cycled continuously forward from 17 Apr 99 and VIC having cycled continuously forward from Oct 00. Owing to the concurrent major effort by NASA/GSFC of preparing 3 years of retrospective forcing for retrospective LDAS LSM runs (see below) for the period of Sep 96 to Sep 99, the MOSAIC realtime LDAS runs were suspended for a time in the past several months, but they have been resumed by NASA, which is in the process of "catching up" from the Apr 01 period of their last realtime executions. The NASA effort has been slowed temporarily by a major disk crash that recently occurred at their site.
II. STATUS OF 3-YEAR RETROSPECTIVE N-LDAS RUNS
In order to overcome initial spin-up behavior in LDAS LSM runs (which can span many months) and to overcome the various occasional glitches (such as interruption of an input observation stream) that unavoidably plague realtime LDAS forcing, the LDAS group decided some months back to execute a clean 3-year retrospective LDAS for the period Oct 96 to Sep 99 (spanning three complete "water" years). NASA/GSFC (Brian Cosgrove in Paul Houser's LDAS group) undertook the large substantial effort of constructing the retrospective forcing fields, highly leveraging NCEP LDAS forcing software and GCIP product archives at NCAR (Eta/EDAS and Stage IV), NCEP (Higgins retrospective precipitation analyses), and U.Maryland (Pinker retrospective GOES GEWEX products).
At the present meeting, Brian Cosgrove reported the completion of his generation of 3-year N-LDAS retrospective forcing. In doing some final QC of the latter forcing database, he recently uncovered a modest error in his generated solar insolation forcing at the very beginning and end of the solar day --referred to by Brian as the "solar edge" problem. He presented some examples of this solar edge problem, and after some discussion, the LDAS group members decided that the magnitude of the error was small and did not warrant a further costly and time consuming round of reprocessing.
ACTION: Hence, all four LDAS LSM modeling groups will commence with executing their respective LSMs in LDAS mode for the 3-year retrospective forcing period cited above and will report on these executions at the next monthly LDAS meeting.
III. REPORT BY WAYNE HIGGIN'S GROUP ON UNIFIED RETROSPECTIVE PRECIPITATION ANALYSIS
Wayne Higgins of NCEP/CPC reported on the data sources, QC, and objective analysis methodology options of his team's GCIP-sponsored, 50-year retrospective and realtime, daily, gage-only precipitation analysis over the CONUS. This daily precipitation analysis of Wayne's team represents the backbone of the realtime and retrospective precipitation forcing for the realtime and retrospective N-LDAS. The presentation at this meeting focused on the new objective analysis techniques and SNOTEL data sources that Wayne's team has acquired and tested from the work of John Schaake. Specifically, Wayne's team has tested, and presented results from, replacing their traditional Cressman objective analysis method with John Schaake's least-squares distance-weighting technique that also employs the PRIZM precipitation climatology to better define orographic precipitation patterns over the inter-mountain western U.S. This presentation sparked animated discussion among the LDAS group members over which objective analysis technique the LDAS group is inclined to ask Wayne's team to use in the precipitation analyses provided to the LDAS input product stream. A final resolution of this choice was hampered by the absence of John Schaake, so the issue will be taken up again at the next LDAS meeting that John attends. Wayne's team also discussed some recent changes in the buddy checks of their observation QC system that proved to be satisfactorily less heavy handed in rejecting very heavy convective precipitation reports.
IV. VALIDATION OF LDAS SOLAR RADIATION AGAINS SURFRAD OBSERVATIONS
Jesse Meng of the global LDAS group under Paul Houser at NASA/GSFC presented results from his emerging study to validate various surface solar insolation products against surface observations at the six CONUS stations of the SURFRAD network. To date Jesse has validated GOES-based retrievals from the U.Maryland/NESDIS GOES GEWEX product suite, EDAS- based solar insolation from NCEP, and solar insolation obtained following the "AGRMET" technique employed by the Air Force Weather Agency (AFWA) in their global LDAS. He illustrated the 1-2 hour phase error in the realtime LDAS solar insolation forcing that was present in the first 4-5 months of realtime LDAS forcing (pointed out by the Princeton VIC group at the annual GAPP/GCIP PI's meeting in May 01) and which was substantially fixed after August 1999, and fully fixed by May 2000. In follow-on work, Jesse will pinpoint the month between Dec 99 and May 00 when the phase error in realtime LDAS solar forcing was fully fixed. Jesse's surface radiation validation web site is http://www.emc.ncep.noaa.gov/mmb/gldas/ldas2.html.
V. SURFACE SOLAR RADIATION PRODUCTS BY NESDIS FROM ORBITING AVHRR
Istvan Laszlo of NESDIS/ORA presented results from pilot work in NESDIS to produce surface solar insolation retrievals globally at approximately 6-hourly frequency from the pair of NESDIS AVHRR polar orbiting platforms, using the cloud reflectance retrieval technique of Andy Heidinger of NESDIS/ORA, followed by a solar insolation retrieval algorithm following that used in geostationary retrievals of solar insolation by Rachel Pinker's group at U.Maryland. Istvan also discussed the problem of low-bias in the GOES retrievals of solar insolation over snow cover and presented a technique yielding some improvement, whereby the heretofore filtering out very shallow snow was dropped. Further work is needed on this low bias in solar insolation retrievals over snow cover, which fundamentally arises from the difficulty of distinguishing cloud cover from snow cover in the visible spectrum.
VI. LDAS VALIDATION OVER OKLAHOMA MESONET BY RUTGERS UNIVERSITY
Lifeng Luo of Rutgers University, under the GCIP project there of Alan Robock, presented plots comparing VIC-modeled LDAS soil moisture (provided by the VIC model runs of Justin Sheffield in Eric Wood's VIC group at Princeton) to OU Mesonet observations. The top layer VIC-modeled versus observed soil moisture features were impressively similar. The study will be expanded to soil temperature and other LDAS LSMs, for presentation at future monthly LDAS meetings. The web site showing Lifeng's LDAS VIC validation results over the OU Mesonet is http://climate.envsci.rutgers.edu/luo/research/LDAS/models.vic.php.