Southern Ontario Winter Flying Outlook

Southern Ontario winter flying outlook, 2026–27

Not for flight planning. This is seasonal context for the winter as a whole. It says nothing about a particular day. Use current NAV CANADA weather and NOTAMs for every flight.

This outlook cannot predict lake-effect precipitation. It can bring sudden whiteouts and very low ceilings to one area while nearby skies stay clear. Check current NAV CANADA forecasts and ECCC snow squall warnings. Why, and which airports are most exposed

Too small for these models: bands can be only a few kilometres to tens of kilometres wide. The forecast grids here are about 100 km, and the station network is sparse.
Too short-lived for a seasonal forecast: a band depends on wind direction, air and water temperatures and lake ice on that day.
A mild winter does not rule it out: open, ice-free water can feed intense bands whenever cold air arrives, and local totals could run well above these numbers. Near freezing, it can also fall as rain or a mix.
Most exposed airports:
- CNK4 Parry Sound: highest. On the east side of Georgian Bay, in the area usually hit hardest.
- CYKF CNC3 Waterloo and Brampton-Caledon: moderate. Inland and downwind of Lake Huron and Georgian Bay, so stronger bands can reach them.
- CYTZ CYOO CYQG CYOW Billy Bishop, Oshawa, Windsor and Ottawa: lower, but not zero. The main bands usually miss them.
This ranking is a judgment based on geography and past storm reports, not a measurement. Any of these airports can be affected.

Milder, with less snow

Don't expect more flyable days

More fog is plausible

Near-freezing weatherMeans closer to 0°C generally bring more slush, wet runways and days in the icing range. That is general meteorology, not a model result.

Your airport

Daytime ceilings under 1,000 ft AGL,

Visibility under 3 SM,

Share of hours: other winters → strong El Niño winters.

Under 1 SM:

Fog and snow,

Fog hours per : other → El Niño.

Snow hours:

Forecast temperature,

Forecast precipitation type,

Freezing rain or drizzle,

Hours per : other → El Niño. History only.

Forecast by month at this airport
Period	Normal mean	Departure, CanSIPS	Departure, SEAS5	Likely mean	Frozen share, normal	Frozen, CanSIPS	Frozen, SEAS5	Precip, CanSIPS	Precip, SEAS5

See it on the map

Past flying weather

Winter 2026–27 forecast

Model

Period, for the whole page

All airports,

Click a row to open that airport above. History compares the two strong El Niño winters, 2016 and 2024, with other winters from 2011 to 2026. Winters are named by their January, so December of winter 2016 is December 2015.

Airport	Name	Ceilings <1,000 ft, other winters	2016	2024	Under 3 SM, other	Under 3 SM, El Niño	Fog hrs, other	Fog hrs, El Niño	Snow hrs, other	Snow hrs, El Niño	Forecast departure	Likely mean	Frozen share, normal	Frozen share, forecast

Ceilings are the share of daytime hours, 08:00–17:59 EST. Forecast departure and frozen share average SEAS5 and CanSIPS.

Accuracy/Skill

Past forecast skill, Dec–Feb

Model reruns of winters 1994–2017 against ECCC stations, trends removed. A correlation of 0 means no year-to-year skill and 1 means perfect.

Variable	Model	Skill	Correlation

Do ECCC and ECMWF agree? Dec–Feb, share of map

Forecast	Agree	No clear signal	Disagree

Agree: both favour the same outcome. Disagree: they favour opposite outcomes, such as wetter and drier. No clear signal: at least one favours neither, so they neither agree nor disagree. Agreement is not proof, because both can be wrong in the same way.

What these numbers can and can't tell you

Temperature: the most reliable part, for direction more than exact size.
Precipitation amount: little past skill. Treat it as near normal.
Ceilings, visibility and fog: no seasonal forecast exists here. These numbers are history, and the El Niño comparison rests on two winters.
2024 drives the seasonal ceiling signal: over Dec–Feb it set new highs at most airports, while 2016 stayed in the normal range. A low-ceiling winter is possible, not certain.
December carries the pattern: both El Niño Decembers had more low ceilings than usual at every airport, and the El Niño average for fog and low visibility was higher at every station. February mostly went the other way. Monthly figures rest on just two Decembers, two Januaries and two Februaries, so they are noisier than the seasonal ones.
This event may be stronger than either: CanSIPS has December Niño 3.4 at , beyond anything in its 1991–2020 reruns.
Freezing rain is not forecast here: a near-freezing precipitation index was built from SEAS5 and tested against winters 1994–2017 at stations. Its year-to-year correlation with observed freezing rain was , and its map pattern ran opposite to where freezing rain actually occurs (). It showed no skill, so it was left out. It also can't see the warm air layer aloft that freezing rain needs.
Frozen share likely runs high: in reruns of 1998 and 2016, SEAS5's frozen share fell about points where stations saw about .
Lake-effect precipitation is not predicted: forecast grids are about 100 km, so snowbelts and individual bands are not resolved. Valley fog and other local effects are also missing.

Background

What past strong El Niño winters did across the region

Median of ECCC stations against 1991–2020. Winters are named by their January. Older winters also ran colder because of long-term warming.

Winter	Strength (ONI)	Temperature	Precipitation	Snowfall

Regional forecast by period

Period	Temp, SEAS5	Temp, CanSIPS	Precip, SEAS5	Precip, CanSIPS	Snowfall, SEAS5	ECCC chance of above-normal temp

How this was made

Forecasts: CanSIPS, started 1 November 2026, and ECMWF SEAS5, started 1 October 2026. November's SEAS5 run was not yet available. Each is compared with its own past reruns: 1991–2020 for CanSIPS and 1993–2016 for SEAS5.
Frozen share: ECCC 1991–2020 station normals, adjusted by SEAS5's modelled snowfall or by a temperature relationship for CanSIPS. The CanSIPS version is an estimate, not an ECCC product.
Ceilings: routine hourly METARs, from 08:00 to 17:59 EST. A ceiling is the lowest broken, overcast or vertical-visibility layer. Records mix human and automated reports, with some reporting gaps.
Maps: forecast fields are interpolated smoothly from grids of about 100 km and cut to the Ontario outline. Tables use the original values and each station's own normals, so a map reading can differ from a table, by several points of frozen share near the lakeshore, as at Oshawa.
Periods: a monthly figure uses only that calendar month from each winter, and needs the same completeness as a seasonal one.
Visibility, fog, snow and freezing precipitation: ECCC hourly observations. Very low freezing counts at some airports likely reflect under-reporting.
Freezing rain history: ECCC hourly weather codes for freezing rain or drizzle. Stations need consistent codes in at least 25 of 29 winters. Where an airport has both staffed and automatic records, the one reporting freezing rain is used, because automatic records can show zero.
Borrowed stations: CNC3 uses Toronto Pearson and CNK4 uses Muskoka, about 50 km away, for history. CNC3 and CNK4 normals, and frozen shares at those two plus CYKF and CYTZ, are fitted from nearby stations.