AWIPS-I & AWIPS-II / WFO LAPS
The purpose of this page is to describe the major differences between the
WFO LAPS release and the more general LAPS that is described on ESRL's FAB
branch web pages. Progress on LAPS improvements and bug fixes relevant
to AWIPS are outlined. This page also provides specific links to other
information sources that may be of use to the WFO/AWIPS LAPS application.
AWIPS LAPS runs using NOAAPORT (nationally distributed) and LDAD (locally
distributed) data that has generally been translated into NetCDF on the
AWIPS platform for use by the LAPS ingest processes. First is a history
of LAPS in AWIPS-I following up with plans for LAPS & STMAS in AWIPS II.
Processes in the AWIPS sched.pl are
Note that these processes produce analysis products that
are fed into the "Big File" for display
on AWIPS. The default version of LAPS on AWIPS does not produce forecasts.
Forecasts using our "hot-start"
can be added if a number of additional components of LAPS are run.
This includes programs such as the balance package "qbalpe.exe", "lapsprep",
and "wfoprep" that create model initialization files. More details are in the
mainly under "Model Initialization and Post-Processing".
The lvd_sat_ingest.exe is run by laps_driver.pl under a seperate cron
What data are getting in LAPS? Check the LAPS GUI...
There is a graphical user interface (GUI) on AWIPS that can tell you what
data (obs & model background) are making it into the LAPS analyses. This can be different from the
obs that are plotted on the workstation due to timing, etc.
To check the GUI, click on "Tools", then
on "LAPS", and click on any of the 5 analyses of interest. In each case,
you should be
able to see the categories and lists of observations making it into the
If the GUI is unavailable the corresponding information can be seen in the LAPS log directory *wgi* files.
Verification statistics are also included for how well the
analysis (and initial background field) fits the observations.
Ingest Flow Chart
The blue colored data are what we are using in AWIPS LAPS. The other
data are being used in our "full-blown" LAPS and can potentially be
added to AWIPS/LAPS if the data becomes available.
An overall description of
is in our
We are using METARS for the surface observations.
LDAD and Maritime surface observations are also being used if available.
There may remain some questions about whether LDAD obs are being processed
by AWIPS in a timely enough fashion to make it into LAPS.
Surface observations from mesonet, wind profiler, and GPS stations should all
into the surface LDAD files.
The data paths for surface data are specified
There are three separate paths for METAR, LDAD, and Maritime files.
Surface observations are now used from a time window between 15 minutes
before and 15 minutes after the top of the hour.
In the past there have been some questions about bulls-eyes
in various surface analyses, particularly the CAPE field. The CAPE is
very sensitive to dewpoint and temperature, so some bulls-eyes are
normally to be expected. The key is to look at and assess
the dewpoint and temperature inputs to the CAPE. Some problems
in the dewpoint/temperature have been related to the spline routine and
erroneous observations making it past the QC.
We have replaced the spline with a more robust "Bratsethian" style
of Barnes successive correction. We are also improving the QC procedures
and thresholds. More details are listed below for each AWIPS build.
We are using the RUC model as the default background, with a failover to Eta
if RUC is not available. Some coastal sites may be better served by using
Eta for the default; this can probably be done on-site by editing the
'background.nl' namelist file.
Recently there have been intermittent LAPS
product outages reported at various AWIPS-I sites. This may be due to
some models have some missing fields. A new version of 'background.nl' is
being used at a few sites to address this (e.g. State College WFO).
The satellite interface for AWIPS depends on (8-bit) image data only, (those
acquired via the Satellite Broadcast Network). The satellite interface
in "regular" LAPS is capable of this plus interfacing to GVAR full resolution
image data (10-bit) and sounder radiance data (13-bit).
More information regarding AWIPS satellite application can
be found here.
In terms of the analyses, the IR image data feeds into the cloud analysis as
well as the surface temperature analysis. The VIS and 3.9u data
(used in clouds and snow cover)
is currently being developed over the next few builds (see below).
LAPS Satellite ingest processes the image closest to the top of the hour,
up to a range of H-15 to H+15 minutes.
The LAPS VAD wind interface (ingest_vad.exe) is currently inactive.
The radar reflectivity data is utilized by the cloud analyses
and QPE precipitation analysis.
We are currently utilizing
low-level radar reflectivity data (Level-III RPG) for input into LAPS.
The strategy is
to process low-level reflectivity data in polar form into a NetCDF file that
is readable by the LAPS ingest program 'remap_polar_netcdf.exe'.
Ideally we prefer full volume (or layer composite) radar data for LAPS,
though LAPS can function to a degree
with the two-dimensional, single-radar input.
LAPS uses the low-level tilt scanned closest to the top of the hour,
up to a range of H-20 to H+20 minutes.
In the future we hope to use multiple radars. Towards this end
a mosaicing program is being implemented as part of the LAPS
AWIPS suite to combine
individual 'vrc' files generated for each radar
using 'remap_polar_netcdf.exe' into a composite 'vrc'.
To improve ground clutter rejection, we hope to use either the DHS (Digital
Hybrid Scan), or multiple low-level tilts.
In the more distant future, wideband (Level-II) reflectivity and velocity
data would be attractive for
use in AWIPS/LAPS.
The additon of Doppler velocity data would improve the quality of the
wind analysis (3-D and surface).
We are currently interfacing to the WFO profiler data
(WPDN 404 MHz network) via 'ingest_pro.exe'. Profiler data will be read in
from the current hour's NetCDF file, if that file is created by :25 past the
hour. If the file arrives later than :25 after then no profiler data gets in.
The boundary layer 915 MHz profiler network interface is inactive.
The RASS ingest also is inactive (ingest_lrs.exe).
can be activated if appropriate
NetCDF data becomes available in the AWIPS database.
ACARS should be functioning in build OB1
with voice Pireps following in OB2.
The interface to RAOB data (ingest_raob.exe) should be functioning
as of version OB1.
RAOBs, even if available, are sporadic so most
hours LAPS runs would not see them being used within the +/- 1 hour
LAPS uses GPS precipitable water data in the humidity analysis
that is stored in NIMBUS/NetCDF files. There is a second local
data (MADIS) NetCDF file that also carries GPS TPW data, though
it has too long of a data latency for real-time use.
The NIMBUS/NetCDF files are available at the ESRL FTP site
and can be used to supply real-time GPS data if a feed of
the data can be set up at the local WFO. The 'moisture_switch.nl'
namelist file in LAPS can be edited to point to this data directory.
An overall description of analysis processes is in Section 3.3 of our
LAPS helicity is calculated as a mean value through
the column, (i.e. "integrated helicity") over
0-3 km AGL. This is what most of the community uses.
For additional information, see our
file in the section on "deriv.exe".
The README document discusses
the newest version of helicity, while the paper on the LAPS Wind Analysis
mentioned therein discusses an older version.
For OB3 we are changing the LAPS storm motion algorithm to use the
Bunkers et. al. method to help ensure consistency with
other helicity calculations in AWIPS.
This has a better accounting for right-moving storms, especially since we
previously had the shear
term turned off and were using a simple mean-wind vector.
There are two LI selections on the workstation.
LAPS LI is a regular LAPS product (in the LAPS big file). It
is computed from a surface parcel taken from the LAPS T and Td sfc
analyses. Surface (terrain following) pressure and 500mb temps are
also taken from LAPS.
Computed LI is an intrinsic on the fly workstation function that
calculates LI from the state LAPS variables.
Overview of AWIPS-I OB3 LAPS improvements:
- Improved coastline gradients of surface variables utilizing land fraction data
- New (
Bunkers et. al.
) method for computing storm motions feeding to
Overview of AWIPS-I OB4 (2005) LAPS improvements:
(Current version at most WFOs)
- Moving of LAPS to the AWIPS/Linux platform
- Improved use of radar data
- Multiple radar mosaicing (narrowband reflectivity)
Improvements being implemented or considered for AWIPS-II:
(Deployed at ~5 WFOs)
- Graphical User Interface (under LAPS Tools on the workstation)
- Domain Resizability (e.g. 4km resolution)
- More frequent analyses (e.g. 15 minutes)
- Graphical Product Monitor
- STMAS (Space-Time variational analysis of variables)
- Surface Obs QC
- Blacklist of station providers (in addition to individual stations)
- Reading in and application of LDAD/MADIS QC flags
- Handling of surface stations with known bias
- Surface Analysis
- Allow adjustment (on workstation) of reduced pressure level
instead of having it fixed at a default value
- Additional background models such as NAM/WRF or HRRR
- Improved use of radar data
- Wide-band full volume scans
- Improved accounting for sub-cloud precipitation evaporation occuring
below the radar horizon. This is most important in western mountainous areas.
- Use of Doppler velocities.
- Wind Profiler
- add observations just outside the domain (implies restructuring the wind
- Use of visible satellite in cloud analysis (cloud addition step)
- GPS total precipitable water data processing via MADIS.
- LI/CAPE/CIN with different parcels in boundary layer
- Fire weather products
- Run and display LAPS forecasts (
initialized via hot start
LAPS home page (with 2010 workshop presentations)
AWIPS LAPS manual
AWIPS LAPS Information
LAPS "Bigfile" Generation
Updated September 2012