NWP Model Biases

NWP systematic errors & biases

Numerical Weather Prediction Model Biases Tendencies and Strengths

( United Kingdom Meteorological Forecast Model ) The UKMET model is run every twelve hours (0 and 12 UTC) out to 144 hours for both North America and Europe. 

Known Biases & Tendencies:

• Tends to place weather features too far south

• Often too low with heights along the southern end of short wave troughs, which results in a storm track that is often too far south

• Has a problem grasping shallow cold air and tends to progress shorter wavelength features too quickly

• Tendency to break down amplified long wave patterns too quickly

• Tends to be very weak with surface cyclogenesis along fronts when a large short wave is digging into a trough

• Tends to be too weak with low-level temperature gradients and too broad with the forecasted baroclinic zone

• Poorest performance when it comes to forecasting polar vortices


• Does fairly well in forecasting the phasing of systems in the northern and southern branches of the jet stream across the northern hemisphere

• Does rather well on forecasting East Coast Cyclogenesis associated with cut-off lows during Fall and Spring

• Fairly good model in forecasting tropical cyclone motion.

GFS (Global Forecast System)  model is run every six hours (0, 6, 12, 18 UTC) out to 384 hours. 

Known Biases and Tendencies:

• Dry bias north of areas where the model has predicted over 2” of QPF in a six hour period. QPF produced from convective feedback blocks northward advection of moisture result of GFS Convective Parameterization Scheme (WARM SEASON ONLY)

• Often over estimates very light precipitation (< .10") in all regions

• Lacks consistency of QPF amounts between model runs

• Often overemphasizes 500 mb features (troughs, ridges, winds, etc.) further out than 3-5 days

• Sometimes referred to as the "GooFuS" model because of its "flip-flopping" tendency between model runs. Over longer range time periods model will tend to forecast cyclogenesis then back off circa day 5-7 before "re-forecasting" cyclogenesis day 3-5.

• Convective Feedback (warm season) - Primarily east of front range and west of Appalachians during wam season, any model run.
When specific thresholds in the mass fields are met, convective scheme is triggered and then dumps a large amount of QPF over a grid point - releasing so much latent heat over the grid point that the model is forced to adjust the mass fields by producing a local vertical motion max in the mid troposphere (~ 500mb), a corresponding upper level jet max over the vertical motion max - an intense and small scale vort max in the mid levels (MCV). The model scales up the mesoscale circulation at mid levels and holds onto it as a real feature for as long as 3 days. The model can produce precipitation in association with the feature as it tracks along in the flow. Probable reason: GFS Convective Parameterization Scheme
• *** HISTORICALLY GFS QPF verifies better than NAM. *** Better verification of GFS QPF > 36 hours

European Centre for Medium-Range Weather Forecasts The ECMWF model is run every 12 hours (0 and 12 UTC) out to 240 hours.

Known Biases & Tendencies:

• Overdevelops mid-level and upper-level cyclones across the southwestern United States.

• Has issues with lifting out of Cut-off lows over the Southwest U.S.

• Has a warm bias when it comes to forecasting mid-level and upper-level tropospheric heights and the resulting thickness calculations

• Has a tendency to “bomb-out” extratropical cyclones along the east coast of the United States too prematurely

• Tends to keep pressures too low on tropical cyclones moving poleward especially along the East Coast of the U.S. that may be undergoing XT or PTC transition.


• Has the best forecast verification score out of all of the medium and long range models

NAM (North American Mesoscale) The North American Mesoscale model is run every six hours (0, 6, 12, 18 UTC) out to 84 hours for North America. 

Known biases and tendencies:

• Develops cyclogenesis too quickly off of the east coast of the United States.

• Over predicts cold air near the ground in the northeast part of the United States when there is snow cover on the ground.

• Tendency to make precipitation overly convective during non-convective events, sometimes causing excessive QPF amounts. 

CMC (Canadian Model) Also known as the GGEM - Global Environmental Multiscale is run every 12 hours (0 and 12 UTC) out to 240 hours for North America. 

Known biases and tendencies:

• Tends to overdo cold air masses and troughs on the east coast during winter
    → 850 mb 0ยบ C Isotherm is often too far in front of precipitation, leading to unrealistic or unlikely snow amounts

• Spurious overforecasting of tropical cyclogenesis