Model output statistics

{
  "id": "MLC-T-MET-001744",
  "term": "Model output statistics",
  "field": "Meteorology and atmospheric science",
  "definitions": [
    "A statistical technique used to improve the accuracy of numerical weather predictions by correcting systematic biases in model output. It involves developing statistical relationships between raw model forecasts and corresponding observed weather conditions. These relationships are then applied to subsequent model forecasts to produce more accurate predictions, often for specific locations.",
    "(abbrev. MOS) - the Hydrometeorological Center (HPC) produces a short range (6 to 60 hours) MOS (Model Output Statistics) guidance package generated from the NGM, GFS, and ETA models for over 300 individual stations in the continental United States. These alphanumeric messages are made available at approximately 0400 and 1600 UTC for the 0000 and 1200 UTC forecast cycles, respectively. Model Output Statistics are a set of statistical equations that use model output to forecast the probability of precipitation, high and low temperature, cloud cover, and precipitation amount for many cities across the USA. The statistical equations were specifically tailored for each location, taking into account factors such as each location's climate."
  ],
  "abbreviation": "MOS",
  "see_also": [
    "perfect prognosis method"
  ],
  "references": [
    "NWS Glossary"
  ],
  "url": "https://mlchart.com/terminology/meteorology/model-output-statistics/"
}