Analysis Scheme and Data Inputs

 


Introduction

All forecast models have similar methods of analyzing the input data to produce a forecast. However, there are so many different observations that can be considered that each model chooses the values that are more important. Observations, however, are not collected in perfectly spaced intervals. This is why it is important to develop a grid to "fill in" missing information. There are two ways of "filling in" the missing information:

        1. using mathematical equations that smooth the observations, and
        2. using the results from the last run of the forecast model as the initial values in the new run.

The model is then put into motion to determine the "weather" of some future time for all grid points.

 Models are classified into three divisions:


Analysis Scheme

AVN (Aviation Model) and MRF (Medium-Range Forecast Model)

The AVN and the MRF also operate on the optimal interpolation technique. They both use the forecast from the final run as the initial guess of the status of the atmosphere. The analysis is preformed on:

The information is then entered into the 42 layers of the global spectral model where the wind and height fields are brought into balance. Even though the AVN and the MRF are used as two different models, they are actually just two different runs of the same global spectral model.

    AVN model forecasts

    MRF model forecasts

Recent updates to both the AVN and MRF models:

 

ECMWF (European Centre for Medium Range Weather Forecasts)

The objectives of the centre are:

The AVN, ECMWF, and MRF use variables such as geopotential height, temperature, u and v wind components, humidity, vorticity, and vertical velocity. The other models also make use of these variables, but they extend their use of parameters by including extra information.

All of this data is collected by various agencies including the National Weather Service (NWS), surface observations of the Federal Aviation Administration (FAA), World Meteorological Organization (WMO), and satellite products of the National Environmental Satellite, Data, Information Service (NESDIS).

 

ETA (Derived from the name of the model's vertical, step-mountain coordinate)

The ETA model extends its use of variables by including precipitation, moisture, cloud coverage, u and v momentum, pressure, surface energy budget, and sub-soil temperatures.  The model makes use of temperature, wind, moisture, ocean surface data, and precipitable water data as provided by rawinsondes, aircraft, wind profilers, and GOES satellite.

The ETA model operates its analysis on an optimal interpolation system. Optimal interpolation is a mathematical "smoothing" of the observed corrections for the observations, linearly at each grid point, with the first guess at the analysis grid point. The weights for the observed corrections are determined statistically and include the distance between observations, the distance between the observations and the analysis points, instrumentation accuracy, and first guess accuracy. This results in a smoother gradient for the initial spin-up cycle. The sharper the discrepancies between the initial values, the more the model will try to "correct" the problem, usually by smoothing the weather phenomenon out of existence (Hoke 1989)

Eta-32: Latest version of ETA

 

MM5 (Fifth Generation NCAR/Penn State Mesoscale Model)

It is the latest in a series that developed from a mesoscale model in the early 1970's. Since that time, it has undergone many changes designed to broaden its useage. These include:

 

MOS (Model Output Statistics)

The MOS interprets NWP models based on historical sampling and is used to predict events forced by synoptic-scale systems. This model is used in conjunction with other NWP models such as the AVN and MRF which take local effects and climate into consideration. In making the forecasts the MOS uses three predictors: NWP Model Forecasts, Prior Observations, and Geoclimatic Data. In addition, the MOS will correct some of the biases of the NWP models and compute uncertainty of the other model forecasts.

AVN based MOS guidance is available every three hours for short range projections of 6 to 72 hours in advance for:

MRF based guidance is available for medium range projections of 24-hours through 192 hours in advance for:

 

NGM (Nested Grid Model)

The NGM model extends its use of variables by including precipitation, moisture, cloud coverage, u and v momentum, pressure, surface energy budget, and sub-soil temperatures.

See ETA for how this model works.

 

RUC (Rapid Update Cycle)

The RUC also operates on an optimal interpolation system. It performs a multivariate height/wind analysis, which allows height observations to influence the wind analysis and wind observations to influence the height analysis. The height observations' real effect is to improve the temperature field. The biggest differences between the old 60 km RUC and the new 40 km RUC is the hourly assimilation of data from the three hour assimilation. The background information for each analysis is the previous one hour forecast. The new data sets included are the VAD wind profiles, boundary-layer profiler winds, Radio Acoustic Sounding System (RASS) temperatures, high-resolution ascent-descent aircraft reports, ship reports, and GOES integrated precipitable water retrievals (Benjamin and Brundage 1994). 

RUC-1

  • Updated every 3 hrs
  • Cutoff time +1 hr 20 min
  • 60-km Grid

RUC-2

  • Updated every hour
  • Cutoff time +0 hr 20 min
  • 50% larger grid than RUC-1
  • 40-km Grid
  • VAD wind profiles
  • Boundary-layer profiler winds, Radio Acoustic Sounding System (RASS) temperatures

The AVN, ECMWF, and MRF use variables such as geopotential height, temperature, u and v wind components, humidity, vorticity, and vertical velocity. The other models also make use of these variables, but they extend their use of parameters by including extra information.

All of this data is collected by various agencies including the National Weather Service (NWS), surface observations of the Federal Aviation Administration (FAA), World Meteorological Organization (WMO), and satellite products of the National Environmental Satellite, Data, Information Service (NESDIS).

 

WRF (Weather Research and Forecasting model)

(The model is not yet finished)

The overall goal of the WRF Model project is to develop a next-generation mesoscale forecast model and assimilation system that will advance both the understanding and prediction of important mesoscale precipitation systems, and promote closer ties between the research and operational forecasting communites. The model is being developed as a collaborative effort among several organizations including the NSF, Dept. of Commerce, and the Dept. of Defense, together with the participation of a number of university scientists.


References

All models website http://www.unidata.ucar.edu/data/models.html

MRF/AVN changes - http://sgi62.wwb.noaa.gov:8080/research/model_changes.html

http://box.mmm.ucar.edu/mm5/overview.html

http://www.wrf-model.org

http://205.156.54.206/tdl/synop/mos2000.htm

http://www.wrh.noaa.gov/wrhq/98TAs/9803/

Benjamin, Stanley G. and Kevin J. Brundage, NOAA/ERL Forecast Systems Laboratory and Lauren L. Monrone, National
    Meteorological Center, Development Division NWS/NOAA. June 1994: "Implementation of the Rapid Update Cycle."
    <http://maps.fsl.noaa.gov/tpbruc.cgi#>

Benjamin, Stanley G., John M. Brown, Kevin J. Brundage, Barry E. Schwartz, Tatiane G. Smirnova, and Tracy L. Smith.
    "RUC-2-Rapid Update Cycle Version 2 Technical Procedures Bulletin."  February 2, 1998:
    <http://maps.fsl.noaa.gov/ruc2.tpb.html>

Black, Thomas L.  June 1994:  "NMC Notes: The New NMC Mesoscale ETA Model: Description  and Forecast Examples."
    Weather and Forecasting.  (265-278).

Caplan, Peter, and Hua Lu Pan. January 24, 2000: "Changes to the 1999 NCEP Operational MRF/AVN Global Analysis/Forecast
    System." <http://sgi62.wwb.noaa.gov:8080/tpb97/T170/html/T170.html>

Hoke, James E., Norman A. Phillips, Geoffrey J. Dimego, James J. Tuccillo, and Joseph G. Sela. September 1989: "The Regional
    Analysis and Forecast System of the National Meteorological Center."  Weather and Forecasting:Volume 4.  (323-334).

Mittelstadt, Jon, WRH-SSD. January 1998: "Western Regional Technical Attachment No. 98-03, January 27, 1998: The Eta-32
    Model."  <http://nimbo.wrh.noaa.gov/wrhq/98TAs/9803/index.html>

Nelson, Jr., James A. July 27, 1999: "Western Region Technical Attachment NO. 99-14."
    <http://www.wrh.noaa.gov/wrhq/99TAs/9914/index.html>

Nielsen-Gammon, John, Texas A&M Meteorology Department. 1998: "The Global Spectral Model."
    <http://www.met.tamu.edu/class/Metr151/tut/models/model6.html>

Nielsen-Gammon, John, Texas A&M Meteorology Department. 1998: "THandy-Dandy MOS Decoder."
    <http://www.met.tamu.edu/class/Metr151/tut/models/model12.html.>

Palmer, Chad, USA Today. 1997: "Some Operational Forecast Models."  <http://www.usatoday.com/weather/wmodlist.htm>

Pan, Hua Lu, John Derber, David Parish,  William Gemmill, Song-You Hong, and Peter Caplan. October 1996: "Changes to the
    1995 NCEP Operational MRF Model Analysis/Forecast System."  National Weather Service: Technical Procedures.  28.

Scialdone, John, CEOS-IDN. January 1998: "Aviation Global Analyses and Forecasts from the Spectral Forecast Model."
    <http://www.neonet.nl/ceos-idn/datasets/NMC_AVN.html>

Scialdone, John, CEOS-IDN. January 1998: "Nested Grid Model (NGM) Data, Analyses, and Forecasts for North America
    Using the Regional Analysis and Forecast System (RAFS)."
    <http://www.neonet.nl/ceos-idn/datasets/NMC_695_NWS0001.html>

Scialdone, John, CEOS-IDN. January 1998: "Medium Range Global Analyses and Forecasts from the Spectral Forecast Model."
    <http://www.neonet.nl/ceos-idn/datasets/NMC_MRF.html>

Staudenmaier, Mike, NWSO. February 1996: "Western Regional Technical Attachment No. 96-06, February 27, 1996: "A
    Description of the Meso-ETA Model."

Woods, Austin.  December 1997: "ECMWF-Forecasting by Computer." <http://www.ecmwf.int/research/fc_by_computer.html>
 


Authors:

Ida Casas
Brooke Cornell
Mary Chacon

May 2, 1998

Updated 02/02/99 By:

Sam Shamburger
Donna Sharp
Rick Scott

Updated 02/29/00 By:

Anita Rapp
Maria Shelley

Updated 02/20/01 By:

Lisa Carlisle
Marek Siwiak

Updated 02/18/02 By:

Suzanne Hill
Mandy Kellner
Richard Penshorn

http://www.met.tamu.edu/class/metr452/models/group2/webpage.dir/analysis.html