|
NOGAPS
Last Updated:
Tuesday, 15-Jul-2008 11:05:08 PDT
|
|

The acronym NOGAPS has been used since 1982
to represent the Navys global numerical weather prediction system,
which includes components for data quality control, data assimilation,
model initialization, and model forecasts. While the nature of the components
has changed considerably over the years, NOGAPS has been and remains the
central engine that is the heart of the Navys environmental prediction
capability. NOGAPS provides forcing fields for mesoscale weather prediction;
tropical cyclone prediction; aerosol prediction; ocean, wave, and ice
prediction; and aircraft and ship routing applications. NOGAPS also forms
the backbone of the Navys ensemble prediction system, which provides
global forecasts out to 10 days.

Continually improving the components of
NOGAPS has been an ongoing research and development project over the years
and many significant changes have been made to the system since it was
first implemented. This historical development is documented in two short
articles, one from 1998
and one from 2002. NOGAPS
has also proven itself to be a valuable research tool, and the system
is actively being used in basic research to increase our understanding
of global dynamics, air/sea interaction, and atmospheric predictability.
History of NRL Global
Model Resolution
Early modeling efforts
1959- 500 mb Quasi-geostrophic barotropic model for thickness and height
1970- N. Hemisphere Primitive Equation Model (63x63 grid points/5 levels)
1975- Added S. Hemisphere PE Model (63 x 63 grid points/5 levels) for
first global coverage
Implementation of global UCLA-based model with
a Successive Corrections Analysis scheme
1982- NOGAPS 1.0 (4.0º x 5.0º/L6)
1983- NOGAPS 2.0 (2.4º x 3.0º/L6)
1983- NOGAPS 2.1 (2.4º x 3.0º/L9)
1987- NOGAPS 2.2 (2.4º x 3.0º/L9)
Implementation of global Spectral model with
a Multivariate Optimum Interpolation Analysis scheme
1988- NOGAPS 3.0 (T47/L18) ~ 2.50º horizontally
1989- NOGAPS 3.1 (T47/L18) ~ 2.50º horizontally
1989- NOGAPS 3.2 (T79/L18) ~ 1.50º horizontally
1992- NOGAPS 3.3 (T79/L18) ~ 1.50º horizontally
1994- NOGAPS 3.4 (T159/L18) ~ 0.75º horizontally
1998- NOGAPS 4.0 (T159/L24) ~ 0.75º horizontally
2002- NOGAPS 4.1 (T239/L30) ~ 0.50º horizontally
History of NOGAPS
Skill
A common metric used by operational forecast centers to measure global
model forecast skill is the anomaly correlation; a correlation of 0.6
typically represents a forecast that has some practical skill. As you
can see from the chart below, while some years are clearly more predictable
than others, the NOGAPS forecasts show an overall upward trend in skill
at all forecast times, with the most dramatic improvements seen in the
5-day forecasts.
The Future of NOGAPS
Several exciting developments are on the horizon for NOGAPS, including
a new variational data assimilation scheme planned for 2003 implementation
and continued improvements to the models physics. More revolutionary
changes include working with the NRL Space Science and Remote Sensing
Divisions to extend NOGAPS much farther up into the atmosphere, to include
all of the stratosphere and the mesosphere. We are also extending downward,
by coupling NOGAPS to a global ocean data assimilation system. Finally,
we are experimenting with finite element grids to examine their potential
to form the basis of a new dynamical core for global modeling and possibly
even a unified global/mesoscale prediction system.
|
NOGAPS 2002
MVOI analysis of heights and winds
.
Complex QC/aircraft QC
Cressman moisture analysis
...
Normal Model Initialization
Spectral T239 (720 x 360 grid)
30 vertical levels/FD
Central Finite Difference
...
12 Minutes/forecast-day
Model top 1 mb (45 km)
.
Palmers et al. Gravity Wave Drag
Louiss vertical mixing
..
Shallow cumulus mixing
Emanuel cumulus convection
Harshvardhans radiation
Teixeiras cloud scheme
...
Large-scale precipitation
Bucket hydrology/Skin temperature
.
Fixed Sea Surface Temperature
Breeder method for ensembles
..
|
NOGAPS 2004+
NAVDAS 3DVAR analysis of t and winds
NAVDAS advanced QC
NAVDAS 3DVAR analysis of moisture
Normal Model Initialization
Spectral/FE, Reduced grid T479
48-60 vertical levels
Semi-Lagrangian
12 Minutes/forecast-day
Model top 0.09 mb (67 km)
Advanced Gravity Wave Drag
TKE based vertical mixing
Mixing off strato-cumulus tops
Emanuel cumulus convection
NASA (Chous) radiation
Explicit cloud scheme
Liquid water prediction
Land Surface Model
Coupled ocean/ice models
Errico/Baumhefner method for ensembles
|
|
|