North American Mesoscale Model

North American Mesoscale Model
Name	North American Mesoscale Model
Developer	National Centers for Environmental Prediction / National Oceanic and Atmospheric Administration
Initial release	1991
Latest release	2020s
Operating system	Cross-platform
Genre	Numerical weather prediction
License	Public domain

North American Mesoscale Model The North American Mesoscale Model is a regional numerical weather prediction system used for short- to medium-range forecasting across North America, including United States, Canada, and parts of Mexico. It provides deterministic and ensemble guidance for operational centers such as the National Weather Service, Environment and Climate Change Canada, and private forecast providers. The model is closely integrated with observational networks from agencies like the National Oceanic and Atmospheric Administration, National Aeronautics and Space Administration, and international partners including European Centre for Medium-Range Weather Forecasts.

Overview

The system produces gridded forecasts of atmospheric state variables—temperature, wind, humidity, and precipitation—on scales suited to mesoscale phenomena such as convective systems, frontal passages, and orographic precipitation. Users range from national agencies like the Federal Aviation Administration and Department of Defense to research institutions such as National Center for Atmospheric Research and universities including Massachusetts Institute of Technology and University of Oklahoma. Model output informs aviation planning at Federal Aviation Administration Air Traffic Organization centers and hydrologic modeling by agencies like the United States Geological Survey. International collaborations involve networks such as the World Meteorological Organization and regional centers like the Canadian Meteorological Centre.

Development and History

Development traces to mesoscale research from groups including National Severe Storms Laboratory, Forecast Systems Laboratory, and academic programs at Penn State University and Colorado State University. The model evolved through architectures influenced by numerical schemes from Geophysical Fluid Dynamics Laboratory and data assimilation advances pioneered at European Centre for Medium-Range Weather Forecasts. Major upgrades corresponded with deployments of observing systems such as the GOES satellite series, Doppler radar networks like NEXRAD, and global models like Global Forecast System. Institutional milestones include cooperative efforts between NOAA, NCEP, and Air Force Weather Agency (now Air Force Weather Agency (USAF) reorganizations), and adoption in operational suites used by National Weather Service forecast offices.

Model Description and Configuration

The model uses a nonhydrostatic dynamical core with physical parameterizations for boundary layer, microphysics, radiation, and land-surface processes. Grid configurations include multiple resolutions (e.g., 3 km, 12 km) and nested domains covering continental-scale and regional domains. Key parameterizations trace lineage to schemes developed at University College London, Rutgers University, and labs like Los Alamos National Laboratory. Model coupling integrates surface datasets from United States Geological Survey and sea-surface conditions from National Oceanic and Atmospheric Administration National Centers for Environmental Information. Operational configurations align with computing environments at supercomputing centers such as Oak Ridge National Laboratory, National Center for Atmospheric Research supercomputers, and NOAA National Centers for Environmental Prediction parallel systems.

Data Assimilation and Input Sources

Data assimilation ingests observations from satellite missions including GOES-R Series, MetOp, and Suomi NPP, aircraft reports such as Aircraft Meteorological Data Relay, radiosonde networks coordinated by World Meteorological Organization, and radar assimilations from NEXRAD and Canadian weather radar systems. Surface networks include Automated Surface Observing System stations and ocean buoys maintained by National Data Buoy Center. Reanalysis products from NCEP/NCAR Reanalysis and global models like ECMWF and UK Met Office Unified Model provide boundary conditions. Advanced assimilation methods incorporate ensemble data assimilation techniques developed in collaborations with European Centre for Medium-Range Weather Forecasts and research from Data Assimilation Research Testbed groups.

Operational Use and Forecast Products

Operational output covers short-term convective outlooks, severe weather guidance, aviation forecasts, and probabilistic fields for temperature, precipitation, and wind. Forecast products feed into services such as National Weather Service Storm Prediction Center outlooks, Hydrometeorological Prediction Center analyses, and aviation products for Federal Aviation Administration planning. Private-sector users include weather firms like The Weather Company and AccuWeather, while emergency management agencies like FEMA use the model for incident planning. Visualization and distribution occur via portals run by NOAA National Weather Service, regional meteorological services including Environment and Climate Change Canada, and academic data centers like Unidata.

Verification and Performance

Verification studies compare model output against observations from Surface Aviation Observations (METAR), radiosonde launches, and radar-derived precipitation estimates. Performance assessments reference metrics established by groups such as National Weather Service verification branches and research from American Meteorological Society conferences. Strengths include improved convective-scale depiction relative to earlier global models from Global Forecast System; weaknesses emerge in complex orography regions like the Rocky Mountains and cold-season precipitation type forecasting impacting regions such as New England. Intercomparisons with ensembles like the Global Ensemble Forecast System and models from ECMWF inform bias corrections and postprocessing.

Applications and Limitations

Applications span aviation, marine forecasts for the Atlantic Ocean and Pacific Ocean, hydrology for river basins like the Mississippi River Basin, and renewable energy siting analyses involving wind and solar forecasts for projects in Texas and California. Limitations include sensitivity to initial condition errors in convective regimes, boundary condition dependence on global models such as GFS, and parameterization uncertainties over heterogeneous surfaces like the Great Plains. Ongoing research partnerships with institutions like Scripps Institution of Oceanography and NOAA Research aim to reduce biases through improved physics, higher resolution, and enhanced data assimilation using observations from platforms such as CubeSats and Unmanned Aerial Systems.

Category:Numerical weather prediction models