NOTE: The nws-changes mailing list is no longer active. The list archives are made available for historical reasons.
Posted: Mon, Nov 16, 1992 3:18 PM EST Msg: JGJC-5452-1222
K.MIELKE(REC)
CC: G.HUFFORD(REC), G.TRAPP(REC), L.MILLER.UCAR(REC),
T.WHITTAKER(REC)
Subj: CHANGES RELATED TO FRHT
I sent the following message this afternoon on AFOS, NWWS, Domestic Data
Service, International Data Service, Honolulu X.25, WMSC (FAA) - who forwards
it to Alaska, Canadian AES, and ISPAN.
---------------------------------------------------------------------------
WSHPNSNMC
NOFS10 KWBC 161900
NATIONAL TECHNICAL INFORMATION MESSAGE 92-11
NATIONAL WEATHER SERVICE HEADQUARTERS
200 PM EST MON NOV 16 1992
TO: ALL NWS OFFICES
DOMESTIC DATA SERVICE
FROM: MARY M. GLACKIN
CHIEF, SERVICES DEVELOPMENT BRANCH
SUBJECT: CHANGE TO THE HORIZONTAL INTERPOLATION USED
TO PRODUCE NGM FOUS PRODUCTS
(The following information was provided by Dr. Ken Mitchell of
the Regional and Mesoscale Modeling Branch of NMC.)
On November 18, 1992 a change is being implemented in the method
for deriving the NGM forecast values provided in the NGM direct
model output products (known on AFOS as "FRHT"). The change
involves an improved procedure for the horizontal interpolation
from NGM grid points to the FRHT station locations.
This change affects the method by which the forecast values in
the following bulletins are derived:
AFOS PILs:
cccFRHT60 - 78, cccFRHT80 - 84, and cccFRHT 86 - 96.
WMO headers:
FOUE60-64, FOUM65-71, FOUW72-73, FOCN74-76, FOGX77, FOUW78,
FOUE80, FOUM81-82, FOUW83, FOUM84, and FOUS86-96
The change does not affect the format of the bulletins listed
above, but only the method by which the forecasts are derived.
Presently, all NGM direct model output values (except
precipitation) are derived by two successive horizontal
interpolations as follows: first, an interpolation from the
NGM's C-grid to the LFM grid, and second, an interpolation from
the LFM grid to the FRHT station locations. The intermediate
interpolation to the coarse LFM grid (190.5 km resolution at 60N)
is undesirable, in that it smooths out the higher resolution
detail present in the finer NGM C-grid (90.75 km resolution at
60N).
Beginning on November 18, 1992, the intermediate interpolation to
the LFM grid is being eliminated, and all NGM direct model output
values (including precipitation) will be derived by a horizontal
interpolation directly from the NGM C-grid to the FRHT station
locations. This change will retain higher horizontal resolution
information in the FRHT values. This impact should be realized
the most in those fields (and synoptic situations) with the
strongest horizontal gradients, including the low level
temperature and humidity fields (T1, T3, R1), surface pressure
(PS), lifted index (LI), and the 700 mb vertical velocity (VVV).
A significant motivation for the change was the adoption of a
final C-grid in the final NGM implementation of August 6, 1991.
The final C-grid has dimensions of 147 x 161 and a resolution of
83.026 km at 45 N, compared to the former C-grid, which was 113 x
91 with a resolution of 83.664 km at 45 N. Though the new and
former grids have very similar resolution, the final C-grid
covers a substantially larger area. Also, the grid point
locations of the two grids are spatially offset by about half a
grid interval. This grid change plus other final NGM changes are
described further in the March 1992 issue of "Weather and
Forecasting" (published by the American Meteorological Society)
in the NMC Notes (written by DiMego et al.) and in the March 1991
issue of "Weather and Forecasting" in the NMC Notes (written by
Petersen et al.).
The impact of the direct model output interpolation change on the
FRHT precipitation values will be minor in the great majority of
cases. From early 1988 until August 6, 1991, the FRHT
precipitation value was obtained directly from the operational
113 x 91 C-grid. Since August 6, 1991, the precipitation has
been interpolated from the new 147 x 161 C-grid to the 113 x 91
C-grid (retained as an output grid) and then to the FRHT
location. However, since the two C-grids are so similar, this
intermediate interpolation had non-trivial effects only
occasionally (and then only modest) in large local precipitation
maximums. With the November 18 change, the precipitation will be
interpolated directly from the 147 x 161 C-grid to the FRHT
location, as with all other fields.
Since AFOS and FAX charts of NGM fields are generated from NGM
output interpolated to the LFM grid (exception: vertical velocity
and precipitation are generated from the C-grid), there can be
some differences between FRHT values and corresponding values
inferred from the FAX or AFOS charts.
Users are reminded that the FRHT output is valid for NGM model
sigma layers. To convert NGM FRHT output to pressure
coordinates, one must use NGM station pressures. Though the NGM
sigma layer structure has not been changed, one of the final NGM
changes was the introduction of a new NGM terrain field, having
somewhat higher resolution (see Petersen et al., 1991). This
change and the change of grid point locations in the final C-grid
affects the NGM FRHT station heights/pressures. Hence, in
Technical Procedures Bulletin (TPB) 351 ("FOUS Messages from the
RAFS"), Table 2, which lists model terrain height and normal
surface pressure at FRHT stations, is no longer correct. In the
revised table, the FOUS station terrain heights are obtained from
a single interpolation from the new (final) NGM terrain heights
on the final 147 x 161 C-grid, exactly consistent with the
interpolation that provides the surface pressure value (PS) in
the FRHT message. Of the 114 stations listed in Table 2 of TBP
351, the terrain height change falls in the range of plus or
minus 0-50, 50-100, 100-150, 150-200, 200-300, and 300+ meters
for 69, 21, 11, 6, 6, and 1 stations, respectively.
A revision of TPB 351 is being prepared. The revised TPB will
include a current FRHT station list and a revised Table 2.
END
nws-changes archives: