NOAA WW3 Model Setup Tehcnical Notes
NOAA WW3 Model Setup Tehcnical Notes
NOAA WW3 Model Setup Tehcnical Notes
Department of Commerce
National Oceanic and Atmospheric Administration
National Weather Service
National Centers for Environmental Prediction
5200 Auth Road Room 209
Camp Springs, MD 20746
Technical Note
This report contains the setup that we have developed to undertake a wave
hindcast study using the WAVEWATCH III R model. This model will be run
with the new NCEP Climate Forecast System Reanalysis Reforecast (CFSRR)
30-year homogeneous data set of hourly 1/2◦ spatial resolution winds, to generate
a wave climatology. This report contains detailed information on the grids that
have been developed, the products that are going to be generated and some initial
validation results.
DRAFT April 8, 2011 ii
http://polar.ncep.noaa.gov/waves
DRAFT April 8, 2011 iii
Contents
Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . i
Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ii
Table of contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iii
1 Introduction 1
2 Grids 3
3 Physics packages 9
4 Products 11
5 Initial validation 13
6 Conclusions 27
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
1 Introduction
A long-term global wave database is very useful to build wave climatologies, do
scenario studies as well as undertake model validation analysis across multiple
time scales. The wave modeling group at the National Center for Environmental
Prediction (NCEP) maintains a wave hindcast database that extends from 1999
to the present. This database uses the archived analysis winds from the GFS
atmospheric model to drive the waves. However, this database is inconsistent in
the sense that all the numerical and physical upgrades to the models (both wave
and atmosphere) are tied in with the variability in the underlying physics. This
can be countered by doing a reanalysis so that the same model can be used to
build a consistent multi-decadal database.
There is not enough available data to develop a traditional re-analysis of the
wave environment. Furthermore, wave dynamics are different from atmospheric
dynamics in the sense that they are more of a boundary value problem than
an initial value problem, with the wind forcing being the most dominant process
driving wave dynamics. It is more useful to do a hindcast re-run using a reanalysis
wind field. However, till now the reanalysis winds developed at NCEP were on
too coarse a grid to allow for the development of a meaningful wave hindcast
database.
A new NCEP Climate Forecast System Reanalysis Reforecast (CFSRR) sys-
tem has been recently developed and entails a coupled reanalysis of the atmo-
spheric, oceanic, sea-ice and land data from late 1979 through 2010, and a re-
forecast run with this reanalysis (Saha et al., 2010). This reanalysis has much
higher horizontal and vertical resolution of the atmosphere than the Global and
the North American Reanalysis, and can thus be used to develop a long-term
hindcast wave database.
The wave model used at NCEP is a third generation wind wave model WAVE-
WATCH III R (Tolman, 2009). In 2007, the model was expanded to run as a mo-
saic of two–way nested grids (Tolman, 2008). The nested grid driver is described
in Tolman (2007a,b), and the grid generation tools used to develop these grids
are described in Chawla and Tolman (2007, 2008). To drive the waves the wave
model requires two input fields: ice and winds (including the air-sea tempera-
ture difference). The high resolution winds used here are 10m above sea level
on an hourly temporal and 1/2◦ spatial resolution which cover the globe from
90◦ S–90◦ N. The reanalysis daily ice concentration fields are 1/2◦ spatial resolu-
tion, and are derived from passive microwave from the SMMR and SSMI using
the NASA Team algorithm. A companion report (Spindler et al., 2011) takes a
closer look at the CFSRR winds that are used in the development of this wave
database.
This hindcast database is foreseen to be developed in three stages. In the
first stage, the wave model shall be run (for the 30 year hindcast period from
DRAFT April 8, 2011 2
1979 to 2009) using the same physics packages that are currently used in NCEP
operations (with minor exceptions these are also the default settings described
in Tolman (2009)). This will set the baseline for the wave model. The database
will be regenerated in stage 2 and 3 with newer physics packages as they become
available, courtesy of a concurrent NOPP initiative to improve physics in oper-
ational wind wave models. However, the model setup and products will remain
unchanged at the different stages1 . The mosaic approach to wave modeling in
WAVEWATCH III allows us to develop a detailed modeling system with high
resolution grids in areas of interest.
The purpose of this report is to outline the model setup and different products
that are going to be generated as part of this database and some initial results.
In section 2 we outline the grids that have been developed for this database (with
further details in appendix B), the physics packages used at this stage are outlined
in section 3, the types of output products are listed in section 4, and some initial
validation results are shown in section 5.
1
This may change depending on developments in model capability and/or database genera-
tion
DRAFT April 8, 2011 3
2 Grids
The WAVEWATCH III model can be run as a mosaic of grids with two–way in-
teraction between the higher and lower resolution grids. This facilitates increased
computational efficiencies by restricting the higher resolution grids only in the
necessary areas.
Keeping in mind the requirements of our collaborative partners, a set of nested
grids was produced for the Global Ocean. In all the grids, the full resolution
ETOPO1 bathymetry was used as the reference grid. Three files were created
for each grid: a bathymetry, a mask, and an obstruction grid which accounts for
wave attenuation by unresolved islands. See Chawla and Tolman (2007, 2008)
for details on the software used for developing these grids.
Grids of three different resolutions were generated: low resolution (1/2 ◦ or
30 arc-minutes), mid resolution (1/6◦ or 10 arc-minutes), and high resolution
(1/15◦ or 4 arc-minutes). The grid generation software that has been used to
develop these grids provides the flexibility of closing some coastal features that
can only be resolved in high resolution grids (e.g. bays, harbors, estuaries etc.).
Appendix A provides a list of the water bodies that can be closed. With the
exception of the Bay of Fundy and Cook Inlet, only in the high resolution grids
were these coastal features kept open. Figures 2.1– 2.3 show the outlines of the
intermediate and high resolution grids, along with the masked areas.
The lowest resolution grids are the 30 arc-minute grids and cover the entire
globe (in longitude). In this current implementation, we are using regular spheri-
cal grids2 and as a result, model time steps are limited by the CFL limit near the
poles. For increased efficiency, the global domain was divided into three “bands”
(see Table 2.1). These three grids are primarily used for computation purposes,
and the output from these grids is stored in a single global grid (referred to as
glo 30m).
Table 2.1: 30 arc-minute grids: range and resolution. All output data
associated with a particular grid are identified by their grid labels.
2
both an unstructured and a curvilinear grid version of the model is currently under de-
velopment and depending upon progress may be involved in the later stages of the database
development
DRAFT April 8, 2011 4
Fig. 2.1 : East and West Coasts of USA, Alaska, and Hawaii Grids
Fig. 2.2 : North Sea / Baltic, Mediterranean, and NW Indian Ocean Grids
DRAFT April 8, 2011 5
Fig. 2.3 : Australia, Pacific Islands, and West Coast US (Hawaii 4 min) Grids
To avoid the singularity at the poles, all grid points beyond 82◦ N in the Arctic
grid are marked as inactive. This is not an issue in the Antarctic grid because of
land.
Nested inside the low resolution grids are the intermediate grids, which have
been masked so that the only active points are those within approximately 250
nautical miles from shore. See Table 2.2 for details on the individual grids. Note
that in the Alaska 10 arc-minute grid, Cook Inlet is open (see Figure B.3 in
Appendix B). In this resolution, Hawaii is part of the Pacific Islands grids.
The highest resolution grids are the 4 arc-minute coastal grids, masked so
that the only active points are those within approximately 100 km of shore. In
this resolution, Hawaii is part of the West Coast US grid. Since these grids can
be computationally very expensive, they have been limited to regions of highest
priority. See Table 2.3 for the details on the individual grids.
Overall, we have separated the global domain into sixteen computational grids
(Fig. 2.4) with the time step particulars given in Table 2.4. Detailed maps of
these grids can be seen in appendix B. The wave model uses four different time
steps internally for each grid, all of which are pre-defined in the grid setup and
have been outlined. Keep in mind that the wave model uses a dynamic time
step for source term integration and the time step in Table 2.4 is the minimum
time step allowed in the model. A discussion on the meaning of the different
time steps is beyond the scope of this report and the readers are referred to the
manual (Tolman, 2009) for more details. All the raw output from the wave model
is generated at these sixteen different grids. For practical reasons, the gridded
DRAFT April 8, 2011 6
Table 2.2: 10-minute grids: range and resolution.Note that in the higher lati-
tude grids of Alaska and the North Sea, the longitude resolution has been
increased to maintain the same resolution in rectilinear coordinates.
output from the three 30 arc-minute computational grids are combined into a
single global grid identified in Table 2.1, yielding fourteen output grids.
Table 2.4: Grid Time steps. All the times are in seconds.
The spectral domain has been divided into 50 frequency and 36 directional
bins (directional resolution of 10◦ ). The minimum frequency has been set at
0.035 Hz and the frequency increment factor has been set at 1.07, providing a
frequency range of 0.035–0.963. A parametric tail is fitted beyond the highest
computed frequency.
DRAFT April 8, 2011 8
3 Physics packages
In WAVEWATCH III the packages used are defined by switches during the com-
pile stage of the wave model(Tolman, 2009). The switches that are being used at
this stage of the database development are as follows:
• NCEP2: NCEP GRIB2 package for IBM (being used in generating GRIB2
output)
• LRB4: 4 Byte words (this is the record length in direct access files)
• LLG: Spherical grid
• PR3: ULTIMATE QUICKEST propagation scheme with averaging tech-
nique for Garden Sprinkler alleviation
• FLX3: Friction velocity from Tolman and Chalikov input with a cap (see
manual for details)
• LN1: Cavaleri and Malanotte-Rizoli linear growth term with a filter
• ST2: Tolman and Chalikov source term package (Tolman and Chalikov,
1996)
• STAB2: Stability correction for Tolman and Chalikov package (see man-
ual)
• NL1: DIA approximation for non-linear interactions
• BT1: JONSWAP bottom friction formulation
• DB1: Battjes-Janssen shallow water depth breaking
• MLIM: Miche-style shallow water limiter for maximum energy (see manual
for details)
• TR0: No Triad interactions
• BS0: No bottom scattering
• WNT1: Linear interpolation of wind in time
• WNX1: Linear interpolation of wind speed in space
The key points are, that at this first stage of the database, DIA formulation
is used to account for non-linear interactions, shallow water wave breaking is
simulated using the Battjes-Janssen formulation and the Tolman-Chalikov source
term packages are used for wave generation and dissipation. In later stages of
the database development one or more of these packages will be changed.
DRAFT April 8, 2011 10
4 Products
Two types of output products are generated by the wave model—field output
that is produced on the model grid layout and point output that is produced at
select locations. For the database runs point output locations have been selected
that correspond to known buoy locations as well as additional locations that
were specifically requested by our collaborators. In all over two thousand output
points have been selected (see appendix C for the full list). The output products
generated on the grids are:
• Spectral partition data at all the grid output points. (Exception being the
30 arc-minute global grids where the output is stored at every other point).
Temporal resolution for the partition data is every hour.
• Bulk spectral parameters using the WMO format at each output point.
This includes the wind speed and direction, significant wave height and
peak period.
In the event that the output point is located in more than one grid, the en-
ergy spectrum is extracted from the finest grid that the point resides in. Linear
interpolation is used to generate the spectrum at the output point from the neigh-
boring grid points in the case that the output location does not correspond to a
computational point.
Apart from these two types of products wave output is also generated along
the altimeter tracks for later validation. This is done by interpolating in time and
space wave parameters (significant wave height and wind speed) from the hourly
field output files on to the altimeter tracks The altimeter tracks are obtained
from the quality controlled global altimeter data set that is maintained at the
French Research Institute for the Exploration of the Sea (IFREMER).
DRAFT April 8, 2011 12
5 Initial validation
As an initial validation, the model was run from June 2005 through December
2005. The model was started from rest condition in June, and run in month long
segments with restart files generated at the end of each run used to start the next
month’s runs. This is the format that shall be used also for the actual runs.
Some initial validation of the model runs has been done with the NDBC buoy
data as well as with IFREMER’s altimeter database. In this section results from
November 2005 (arbitrarily chosen) have been shown. Results from other months
show similar trends.
Figure 5.1 shows scatter plots from all the altimeters. Since just a plain scatter
plot does not provide enough information on the number of points, a probability
density function (PDF) was created by binning model and altimeter data into
bins and plotting that distribution (Tolman et al., 2006). The PDF is defined as
nbin
P DF =
nT ∆x∆y
where, nbin are the number of points in the bin, nT are the total number of points,
and ∆x, ∆y are the bin resolutions for model and altimeter data respectively.
Figure 5.1 shows that in all the altimeters the wave height has a positive
bias that increases for higher wave heights and a slightly negative bias for wind
speeds. Error maps corresponding to these four different altimeters are shown
in Figs. 5.2 to 5.5. Also a map of combined altimeter data was made (Fig 5.6).
The maps show that significant part of the biases are along the eastern edges of
the ocean basins, corresponding to insufficient dissipation in swell fields in the
model. This and the larger biases in the Southern Ocean shall be addressed in
the second stage of the wave database.
Time series comparisons at some select buoys are also shown in Figs. 5.7–
5.13. The presented locations were chosen arbitrarily to correspond to points in
the Atlantic Ocean, Pacific Ocean and in the Southern Hemisphere. The biases
in the time series reflect what was observed in the error maps in the altimeters.
DRAFT April 8, 2011 14
Fig. 5.2 : Error map of Significant wave height from the GFO satellite. To
build these maps, model and data wave heights along altimeter tracks
are binned into 2◦ bins. Error statistics have been computed inside each
bin and a 3 point running average is done across the longitudes and
latitudes to remove track signature. All values are in m (except for SI
which is non-dimensional). The Hs map on the top left panel is made
with altimeter data.
DRAFT April 8, 2011 16
Fig. 5.3 : Idem. Fig 5.2 but from the ENVISAT satellite.
DRAFT April 8, 2011 17
Fig. 5.4 : Idem. Fig 5.2 but from the JASON-1 satellite (white boxes indicate
not enough data).
DRAFT April 8, 2011 18
Fig. 5.5 : Idem. Fig 5.2 but from the ERS2 satellite. Note the limited data
availability.
DRAFT April 8, 2011 19
Fig. 5.6 : Idem. Fig 5.2 but for combined altimeter data from all the satel-
lites. Maps for the combined data was made by binning the data in 1◦
bins (as opposed to 2◦ bins in the other figures).
DRAFT April 8, 2011 20
6 Conclusions
This report highlights the setup of the wave model to develop the 30-year hindcast
based on the CFSRR reanalysis winds. Grids were designed such that the major
water bodies are all well resolved. Higher resolution grids were developed in areas
that are of interest to our collaborative partners as well as places where buoy data
are available for validation studies. At this stage of the study, we are still limited
to using regular spherical grids. This may change for the later stages depending
on the progress made in developing alternative numerical algorithms. To make
sure that reasonable output is being generated a validation exercise was carried
out with output from a short test run. The errors were found to be on the same
order as our hindcast validation study based on analysis winds from the GFS
model(Chawla et al., 2009). A detailed analysis of the output will be done once
the first stage of the database is complete.
DRAFT April 8, 2011 28
References
Chawla, A. and H. L. Tolman, 2007: Automated grid generation for WAVE-
WATCH III. Tech. Note 254, NOAA/NWS/NCEP/MMAB, 71 pp.
Chawla, A. and H. L. Tolman, 2008: Obstruction grids for spectral wave models.
Ocean Mod., 22, 12–25.
Chawla, A., H. L. Tolman, J. L. Hanson, E.-M. Devaliere and V. M. Gerald,
2009: Validation of a multi-grid WAVEWATCH III modelling system. in 11th
international workshop on wave hindcasting and forecasting & coastal hazards
symposium.
Saha, S., S. Moorthi, H. Pan, X. Wu, J. Wang, S. Nadiga, P. Tripp, R. Kistler,
J. Wollen, D. Behringer, H. Liu, D. Stokes, R. Grumbine, G. Gayno, J. Wang,
Y. Hou, H. Chuang, H. Juang, J. Sela, M. Iredell, R. Treadon, D. Kleist,
P. V. Delst, D. Keyser, J. Derber, M. Ek, J. Meng, H. Wei, R. Yang, S. Lord,
H. van den Dool, A. Kumar, W. Wang, C. Long, M. Chelliah, Y. Xue,
B. Huang, J. Schemm, W. Ebisuzaki, R. Lin, P. Xie, M. Chen, S. Zhou, W. Hig-
gins, C. Zou, Q. Liu, Y. Chen, Y. Han, L. Cucurull, R. Reynolds, G. Rutledge
and M. Goldberg, 2010: The NCEP climate forecast system reanalysis. Bull.
Am. Meteor. Soc., pp. 1015–1057.
Spindler, D. M., A. Chawla and H. L. Tolman, 2011: An initial look
at the CFSR Reanalysis winds for wave modeling. Tech. Note nnn,
NOAA/NWS/NCEP/MMAB, nn pp.
Tolman, H. L., 2007a: The 2007 release of WAVEWATCH III. in 10th interna-
tional workshop on wave hindcasting and forecasting & coastal hazards sympo-
sium. Paper Q4.
Tolman, H. L., 2007b: Development of a multi-grid version of WAVEWATCH
III. Tech. Note 256, NOAA/NWS/NCEP/MMAB, 88 pp. + Appendices.
Tolman, H. L., 2008: A mosaic approach to wind wave modeling. Ocean Mod-
elling, 25, 35–47.
Tolman, H. L., 2009: User manual and system documentation of WAVEWATCH
III version 3.14. Tech. Note 276, NOAA/NWS/NCEP/MMAB, 220 pp.
Tolman, H. L., D. Cao and V. . M. Gerald, 2006: Altimeter data for use in wave
models at ncep. Tech Note 252, NOAA/NWS/NCEP/MMAB.
Tolman, H. L. and D. Chalikov, 1996: Source terms in a thrid generation wind
wave model. Journal of Physical Oceanography, 26, 2497–2518.
This page is intentionally left blank.
APPENDICES
DRAFT April 8, 2011 A.1
B Appendix B: Grids
$ input.
$ - The first three data columns are used by the forecats code, the other
$ are used by postprocessing scripts.
$
$ NE Pacific deep ocean
$
$ LON LAT NAME AH TYPE SOURCE SCALE
$ ---------------------------------------------------------
-148.02 56.31 ’46001 ’ 5.0 DAT NDBC 360
-130.27 42.60 ’46002 ’ 5.0 DAT NDBC 360
-136.10 50.93 ’46004 ’ 5.0 DAT ENCAN 360
-131.02 46.05 ’46005 ’ 5.0 DAT NDBC 360
-137.48 40.80 ’46006 ’ 5.0 DAT NDBC 360
-177.58 57.05 ’46035 ’ 10.0 DAT NDBC 360
-133.94 48.35 ’46036 ’ 5.0 DAT ENCAN 360
-130.00 37.98 ’46059 ’ 5.0 DAT NDBC 360
-154.98 52.70 ’46066 ’ 5.0 DAT NDBC 360
175.28 55.00 ’46070 ’ 5.0 DAT NDBC 360
-172.03 54.94 ’46073 ’ 10.0 DAT NDBC 360
-138.85 53.91 ’46184 ’ 5.0 DAT ENCAN 360
$
$ NE Pacific coastal
$
$ LON LAT NAME AH TYPE SOURCE SCALE
$ ---------------------------------------------------------
$ Alaska
$
-146.83 60.22 ’46061 ’ 5.0 DAT NDBC 90
179.05 51.16 ’46071 ’ 5.0 DAT NDBC 360
-171.73 52.25 ’46072 ’ 5.0 DAT NDBC 360
-160.81 53.93 ’46075 ’ 5.0 DAT NDBC 360
-148.00 59.50 ’46076 ’ 5.0 DAT NDBC 360
-152.45 56.05 ’46078 ’ 5.0 DAT NDBC 360
-152.09 59.76 ’46106 ’ 999 DAT NDBC 75
-150.00 58.00 ’46080 ’ 5.0 DAT NDBC 360
-143.42 59.69 ’46082 ’ 5.0 DAT NDBC 360
-138.00 58.25 ’46083 ’ 5.0 DAT NDBC 360
-136.16 56.59 ’46084 ’ 5.0 DAT NDBC 360
-142.56 56.85 ’46085 ’ 5.0 DAT NDBC 360
$
$ Canada
$
DRAFT April 8, 2011 C.11
$
-9.26 68.48 ’64071 ’ 999. DAT UNKNOWN 60
-23.10 64.05 ’TFGSK ’ 999. DAT UNKNOWN 60
-25.00 65.69 ’TFBLK ’ 999. DAT UNKNOWN 60
$ -23.36 66.44 ’TFSTD ’ 999. DAT UNKNOWN 60
$ -21.12 65.76 ’TFDRN ’ 999. DAT UNKNOWN 60
-18.20 66.50 ’TFGRS ’ 999. DAT UNKNOWN 60
-13.50 65.65 ’TFKGR ’ 999. DAT UNKNOWN 60
-15.20 64.00 ’TFHFN ’ 999. DAT UNKNOWN 60
-20.35 63.00 ’TFSRT ’ 999. DAT UNKNOWN 60
$ -22.46 63.82 ’TFGRV ’ 999. DAT UNKNOWN 60
$
$ Norwegian Sea
$
7.80 64.30 ’LF3F ’ 999. DAT UNKNOWN 360
7.30 65.30 ’LF3N ’ 999. DAT UNKNOWN 60
8.10 66.00 ’LF5T ’ 999. DAT UNKNOWN 360
2.00 66.00 ’LDWR ’ 999. DAT UNKNOWN 360
$
$ North Sea
$
1.10 55.30 ’62026 ’ 999. DAT UNKNOWN 360
0.00 57.00 ’62109 ’ 999. DAT UNKNOWN 25
0.40 58.10 ’62111 ’ 999. DAT UNKNOWN 25
1.30 58.70 ’62112 ’ 999. DAT UNKNOWN 25
1.40 57.70 ’62116 ’ 999. DAT UNKNOWN 360
0.00 57.90 ’62117 ’ 999. DAT UNKNOWN 15
0.90 57.70 ’62118 ’ 999. DAT UNKNOWN 15
2.00 57.00 ’62119 ’ 999. DAT UNKNOWN 25
$ -3.50 53.80 ’62125 ’ 999. DAT PRIV 25
$ -3.60 53.90 ’62126 ’ 999. DAT UNKNOWN 25
$ -3.80 54.00 ’62135 ’ 999. DAT UNKNOWN 25
1.40 58.70 ’62128 ’ 999. DAT UNKNOWN 25
2.00 56.40 ’62132 ’ 999. DAT UNKNOWN 25
1.00 57.10 ’62133 ’ 999. DAT UNKNOWN 15
2.10 53.00 ’62142 ’ 999. DAT PRIV 30
1.80 57.70 ’62143 ’ 999. DAT UNKNOWN 25
1.70 53.40 ’62144 ’ 999. DAT PRIV 45
2.80 53.10 ’62145 ’ 999. DAT PRIV 360
2.10 57.10 ’62146 ’ 999. DAT UNKNOWN 25
1.80 57.00 ’62152 ’ 999. DAT UNKNOWN 25
0.50 57.40 ’62162 ’ 999. DAT UNKNOWN 25
DRAFT April 8, 2011 C.20
$
$ LON LAT NAME AH TYPE SOURCE SCALE
$ ---------------------------------------------------------
-48.13 -27.70 ’31201 ’ 999. DAT SCRIPPS 180
-48.75 -32.00 ’RIO_GRANDE’ 999. VBY NCEP 360
-46.25 -28.00 ’FLORIPA ’ 999. VBY NCEP 360
-43.75 -25.00 ’SANTOS ’ 999. VBY NCEP 360
-38.75 -21.00 ’CAMPOS ’ 999. VBY NCEP 360
-36.25 -13.00 ’SALVADOR ’ 999. VBY NCEP 360
-32.50 -8.00 ’RECIFE ’ 999. VBY NCEP 360
-36.25 -3.00 ’FORTALEZA ’ 999. VBY NCEP 360
-47.50 3.00 ’AMAZON ’ 999. VBY NCEP 360
-30.00 1.00 ’PETER_PAUL’ 999. VBY NCEP 360
$
$ Peru/Chile Basin
$
$ LON LAT NAME AH TYPE SOURCE SCALE
$ ---------------------------------------------------------
-85.38 -19.62 ’32012’ 999. DAT WHOI 360
$
$ South Africa
$
$ LON LAT NAME AH TYPE SOURCE SCALE
$ ---------------------------------------------------------
22.17 -34.97 ’AGULHAS_FA’ 10.0 DAT PRIV 360
$
$ Australia
$
$ LON LAT NAME AH TYPE SOURCE SCALE
$ ---------------------------------------------------------
141.75 -12.68 ’52121 ’ 999. DAT UNKNOWN 50
150.34 -35.71 ’55014 ’ 999. DAT UNKNOWN 50
153.73 -28.69 ’55017 ’ 999. DAT UNKNOWN 10
153.27 -30.35 ’55018 ’ 999. DAT UNKNOWN 10
152.86 -31.83 ’55019 ’ 999. DAT UNKNOWN 50
150.18 -37.29 ’55020 ’ 999. DAT UNKNOWN 50
151.03 -34.48 ’55022 ’ 999. DAT UNKNOWN 50
151.42 -33.77 ’55024 ’ 999. DAT UNKNOWN 50
151.32 -33.90 ’55025 ’ 999. DAT UNKNOWN 10
145.01 -42.08 ’55026 ’ 999. DAT UNKNOWN 50
145.71 -16.73 ’55028 ’ 999. DAT UNKNOWN 50
147.06 -19.16 ’55029 ’ 999. DAT UNKNOWN 50
DRAFT April 8, 2011 C.22
$ End of list
$
0.00 0.00 ’STOPSTRING’ 999. XXX NCEP 0