A mesoscale snow band is a narrow, elongated zone of much heavier snow embedded within a larger winter storm. Research on Northeast U.S. cyclones defines a band on radar as a linear feature 20 to 100 km (about 12 to 60 miles) wide and more than 250 km (about 155 miles) long, with reflectivity above 30 dBZ lasting at least 2 hours.
Where bands form
Many of the best-studied bands form in the comma head of a developing low, the comma-shaped part of the cloud and precipitation shield that wraps around the storm center.
What drives a band
The key forcing in these studies is frontogenesis, the strengthening of a temperature gradient, which drives a narrow zone of rising air along the frontal zone. In a 2010 study by Novak and colleagues, lower-tropospheric frontogenesis nearly doubled in the 6 hours before bands formed.
Stability matters too. The same study found weakly stable conditions or conditional instability most likely during band formation, with layers of conditional instability present nearly three times as often as layers of conditional symmetric instability.
Why totals jump over short distances
A band is narrow, and it can last for hours. That combination produces intense snowfall and extreme snowfall gradients, so one community can measure far more snow than another a short drive away.
How forecasters look for them
Forecasters look for strong frontogenesis paired with weak stability, then check convection-allowing models for banded structures. The Weather Prediction Center (WPC) runs a prototype snowband probability page that identifies band objects in hourly precipitation from three model ensembles, masked to areas forecast to have snow.
Band placement remains the hard part. Research describes these bands as difficult to forecast, so a band in model output is best read as a possibility within a zone rather than a precise track.