Precipitation Type in a Dammed Cold Pool

Aloft: complete melting Snow from the −17 °C cloud layer falls into a warm layer that crosses 0 °C near 880 hPa, peaks at +2.6 °C at 800 hPa, and returns below freezing near 745 hPa. That is roughly 1.4 km of above-freezing air with more than 2 °C of excess. As a rule of thumb, a warm layer at least 1–2 km deep with a peak above roughly +1 °C melts a snowflake completely; a shallow nose of a few hundred metres, or a peak barely above 0 °C, only partly melts it and the refrozen remnant falls as sleet. So the hydrometeor reaching the top of the cold pool overnight is a liquid drop.

Below: marginal refreezing The drop then falls through about 1 km of air at −0.6 to −4.2 °C. That is right at the classical boundary: a subfreezing layer shallower than about 500 m almost always gives freezing rain, while more than about 1 km with temperatures below −3 °C gives a good chance of refreezing into sleet. Expect a mix overnight, with freezing rain dominating where the cold layer is shallowest (east side of the metro) and sleet toward the gorge where easterly damming keeps it deepest.

Onset is different At 15 UTC the column is dry (Td −9.5 °C), so evaporative cooling is strong, the cold layer deepens and cools before saturation, and the warm nose is not yet fully established. The first hour or two is therefore most likely sleet, transitioning to freezing rain as the column saturates and warm advection strengthens the nose.

Wet-bulb estimate With T = −4.2 °C and Td = −9.5 °C, the wet-bulb temperature is roughly −6 to −6.5 °C. Precipitation falling into subsaturated air will drive the surface temperature toward the wet-bulb, not toward the model's forecast temperature, which typically assumes near-saturation too early. Plan on the first 1–2 hours running about 2 °C colder than the raw model surface temperature.

Three consequences

1. Type: a colder, deeper subfreezing layer means more sleet and more efficient refreezing of any liquid drops - the onset sleet signal is reinforced.
2. Geometry: the surface freezing line shifts 20–50 km south and east of where the model places it, which expands the area needing a headline. The advisory/warning boundary drawn from model 2 m temperature will be too far north.
3. Roads: road sensors already read −3 to −5 °C. Asphalt has enough thermal inertia that a thin glaze bonds immediately even if the air temperature later creeps to 0 °C or slightly above.

Decision rule Do not wait for the shelter air temperature to reach 0 °C before issuing. Issue on the wet-bulb and road-temperature evidence, and treat the first hour of observed precipitation type at the surface as the trigger to upgrade or downgrade.

Accretion arithmetic Glaze density is close to 0.9 g cm⁻³, so for moderate wind the radial ice depth is roughly comparable to the liquid depth that actually freezes on contact. If about 60% of the 18 mm falls as freezing rain, expect roughly 8–11 mm of radial ice; the 20 kt easterly wind increases accretion efficiency on north–south-oriented lines and branches.

That comfortably exceeds the 6.4 mm (0.25 in) ice-storm-warning threshold, and 8–11 mm is enough for scattered tree and wire damage.

Headline Issue an ice storm warning for the freezing-rain-dominant zone east of the gorge where easterly damming is weakest, and a winter weather advisory for the zone expected to stay sleet. State the two-type structure explicitly rather than averaging it into one number.

Failure mode The warning busts if the cold pool stays deep enough that most of the 18 mm arrives as sleet. Sleet has a much lower accretion efficiency and no glaze bond, so the ice totals collapse even though the QPF verifies. That is why the first hour of surface observations - precipitation type at the airport, plus road-sensor trends - is the deciding evidence, not the QPF. If sleet persists past 18 UTC, downgrade the warning zone quickly rather than waiting for the model to catch up.