Data Cutoffs and Scheduling

Description

    Models use atmospheric observations to analyze and predict the future state of the atmosphere. Most models use information from other models to initialize the data. The output that is produced by a model is distributed worldwide in a series of "runs." Although these models have many things in common, they differ in purpose and content. One thing that all models have in common is the establishment of analysis and initial conditions that models use to describe the current state of the atmosphere. Also, all models strive to forecast the future state, and all models disseminate the analyses and present a useful forecast to a wide variety of users.
    Models are unable to continously take in new data. For this reason a stopping point must be established. The term "data cutoff" is used to refer to that stopping point. Models assimilate all the data that was taken in and create a "run". Each model has its own data cutoff and schedule.
    Scheduling refers to how often a model is run. Some models run more frequently than others (for example, every hour instead of every three hours...). This allows them to update and include new observations. The more recent the meteorological data is, the more precise the model is able to depict the current weather situation.
   Data cutoffs and scheduling are integral parts of a model's input and output. The data cutoff stops the input of current information. The schedule denotes the time when the model uses the data to make its runs.

Easy Reference Chart

Model	Scheduling	Data Cutoff
MESOETA	00Z, 06Z, 12Z, 18Z	H + 0:50
ETA	0Z, 03Z, 12Z, and 18Z	H + 1:10
RUC RUC-2	Every 3 Hours Hourly	H + 1:20 H + 0:20
NGM	0Z, 12Z	~3 hrs after 0Z, 12Z
ECMWF	0Z, 6Z, 12Z, 18Z, and global 10-day forecast based on 12Z data	~20 minutes before each run for initialization
MRF	0Z only	H + 6:00
AVN	0Z, 6Z, 12Z, 18Z	H + 2:45
MM5	00Z, 06Z, 12Z, 18Z
WRF	00Z, 12Z	N/A
NOGAPS	00Z, 12Z	H + 3:30

Note: H = most recent observation time

Comparison of Models

MESOETA

The MESOETA model runs at 00Z, 06Z, 12Z, and 18Z to make use of the most current lateral boundary conditions. This model provides forecasts over the United States at a very high resolution and runs out to 48 hours.

ETA

The ETA model also provides a regional forecast over the United States soon after the 0Z, 3Z, 12Z, and 18Z hours. The 0Z and 12Z extent to 84 hours. The National Weather Service and meteorological community are able to access and use this model.

RUC

The RUC model provides short-term forecasts over the United States and portions of Canada. An analysis and hourly forecast files out to 12 hours are produced every three hours. The data is from sources such as profilers and aircraft taken at asynoptic times. Surface data cutoff is 35 minutes after the analysis time. Before each forecast the RUC is run adiabatically for 60 minutes in both the forward and backward directions.

RUC-2

The RUC-2 model is an updated version of the RUC model. It provides hourly forecasts and has a shorter data cutoff time. It does 3 hour forecasts every hour and every third hour it does a 12 hour forecast.

NGM

The NGM model generates 48 hour forecasts twice a day for MOS guidance. Although the data cutoff is 2 hours after 0Z and 12Z, observations received at the National Meteorological Center by cutoff time within 3 hours of 00Z and 12Z are considered for the analysis.

ECMWF

The ECMWF produces routine global analysis every 6 hours and a global 10-day forecasts based on 12Z data. The model uses the 6 hour forecast as its initial guess in performing an objective analysis every 6 hours.

MRF

The purpose of the MRF model is to generate a global forecast for the medium range time scale, generally understood to mean the 3-10 day range. It uses Global Data Assimilation System (GDAS) - delay the final run as long as possible so as to pick up late arriving conventional and satellite data..

AVN

The AVN model is another global forecast model. It is used to create short range forecasts, up to 3 days in the future. A difference between the AVN and the MRF is the data cutoff times (AVN does not utilize GDAS). The purpose of the 0Z forecast is to prepare guidance in support of NCEP's international aviation responsibilities. Also,it speeds up the reception of aviation guidance over the GTS. The data cutoff time was advanced to 2:45 in 1989. The model is run four times daily at 0Z, 6Z, 12Z, and 18Z and extends out to 78 hours.

MM5

The MM5 model is a regional model that requires an initial condition along with a lateral boundary condition. It is only used by the US Air Force and utilized by some universites such as Penn State.

WRF

This model is still being developed and is not in complete operational use. However, the overall goal of the WRF program is to develop a next-generation mesoscale forecast model. This model will advance both the understanding and prediction of important mesoscale precipitation systems. The model runs twice a day at 00Z and 12Z.

NOGAPS

This model is only used by the US Navy. It runs twice a day at 00Z and 12Z and goes out to 10 days. The model concentrates on how surface heat and momentum fluxes interact with clouds and radiation. In addition, this model has acted to improve prediction of cyclogenisis.

Critical evaluation

Ideally a model would be very sophisticated, running many times a day and gathering a great deal current meteorological observations. Unfortunately, these sophisticated models would take a long time to run. There must be a trade off in which the data cutoffs and scheduling are involved.

Each of these models serves a different purpose, therefore, it is difficult to choose one model over another. When choosing a model, the user must decide what he is looking for. Before choosing a model one may want to decide if a small time and spatial scale will be sufficient or if one wants to look at the big picture. Some models are able to pick up on the mesoscale features while others concentrate on synoptic features. The synoptic scale models may overlook the smaller scale features. The RUC-2 and the MESOETA are examples of models that continually improve the forecast of smaller scale features. With increased technology, models are able to predict features with increasing accuracy. Faster computers will soon be able to produce results more rapidly.

References

Hoke, James H., Norman A. Phillips, Geoffrey J. DiMego, James J. Tucillo, Joseph G. Sela, 1989: The Regional Analysis and Forecast System of the National Meteorological Center. Weather and Forecasting, 4, 323-334.

Caplan, Peter, 1998: Global Model. http://sgi62.wwb.noaa.gov:8080/STATS/COMET/html/pres2.html

ECMWF Data Coverage Charts, 1998: http://www.ecmwf.int/services/dcover/

Kanamitsu, Masao, 1989: Description of the NMC Global Data Assimilation and Forecast System. Weather and Forecasting, 4, 335-342.

Nielsen-Gammon, John, 1996-98: Computer Models. http://www.met.tamu.edu/class/Metr151/tut/models/model7.html

NOAA, 2001: Information on Operational Models. http://205.156.54.206/er/bgm/models

NOAA/NWS, 1996: Annual Numerical Model Research Report. http://sgi62.wwb.noaa.gov:8080/research/ap/web3/annul.html

RUC/MAPS, 1998: Explanation of Observational Data Input to RUC-2. http://maps.fsl.noaa.gov/

USA Today Weather, 1999: National Models. http://www.usatoday.com/weather/wmodlist.htm

Weather Research and Forecasting (WRF) Model, http://www.wrf-model.org

Jeff Jamison, Julie Jasien, and Greg Ostermeier
April 30, 1998
Updated by:
Andy Brown, and Andy Wilkinson
February 3, 1999
Sarah Libby,andLaura Beeson
February 29, 2000
Justin Palmer, Mindy Chavez andToniTreadway
February 14 2001
Ed Fortner, Brooke Stellato andKirsten Murray
February 14 2002

Click here to return to the main modeling page.