Region Southwest
The readings under the map are for the whole map, not only the part shown.
GFS, ECMWF · Run 06Z 1 Oct · +0 h · Valid Thu 06Z 1 Oct
Both models: 16.5% of the map is past its 90th percentile for 06Z · one model only: another 5.0%
Both models: 17.1% of the map is past its 90th percentile for 12Z · one model only: another 6.2%
Both models: 17.5% of the map is past its 90th percentile for 18Z · one model only: another 6.2%
Both models: 17.2% of the map is past its 90th percentile for 00Z · one model only: another 7.2%
Both models: 16.1% of the map is past its 90th percentile for 06Z · one model only: another 6.4%
Both models: 14.3% of the map is past its 90th percentile for 12Z · one model only: another 5.8%
Both models: 14.0% of the map is past its 90th percentile for 18Z · one model only: another 5.6%
Both models: 14.3% of the map is past its 90th percentile for 00Z · one model only: another 5.3%
Both models: 14.2% of the map is past its 90th percentile for 06Z · one model only: another 4.9%
Both models: 14.1% of the map is past its 90th percentile for 12Z · one model only: another 5.2%
Both models: 14.1% of the map is past its 90th percentile for 18Z · one model only: another 5.9%
Both models: 14.8% of the map is past its 90th percentile for 00Z · one model only: another 7.0%
Both models: 15.3% of the map is past its 90th percentile for 06Z · one model only: another 7.8%
Both models: 16.3% of the map is past its 90th percentile for 12Z · one model only: another 7.8%
Both models: 16.2% of the map is past its 90th percentile for 18Z · one model only: another 8.6%
Both models: 16.0% of the map is past its 90th percentile for 00Z · one model only: another 10.5%
Both models: 15.1% of the map is past its 90th percentile for 06Z · one model only: another 10.9%
Both models: 13.3% of the map is past its 90th percentile for 12Z · one model only: another 11.4%
Both models: 12.4% of the map is past its 90th percentile for 18Z · one model only: another 12.9%
Both models: 12.8% of the map is past its 90th percentile for 00Z · one model only: another 14.4%
Both models: 13.0% of the map is past its 90th percentile for 06Z · one model only: another 13.4%
GFS and ECMWF, each from its 06Z 1 Oct run — played forward 6 hours at a time, from the analysis to 120 hours out.
Methodology
PWAT — precipitable water — is how much water vapour the air overhead holds, as the depth of water it would make if it all fell at once. It is the moisture a storm has to work with, which is why forecasters look at it before heavy rain.
A column that is ordinary in Miami is extraordinary in Salt Lake City, so this map compares each place with itself: the column is ranked against that place's own columns at the same hour of the day, within 15 days of the date, over 2021–2026 of the ERA5 reanalysis. Past the 90th percentile means wetter than nine in ten of those. That is not the same number as a day's peak against peaks, which is how a heavy-rain threshold is usually set: an afternoon's column sits a few millimetres under that day's peak.
Two models, because only two file this field: GFS as PWAT and ECMWF as total column water vapour, the same quantity. ECMWF's AI model files only total column water, which counts cloud and rain as well and reads high exactly where it rains; GEM and NOAA's AI model file neither. Uncorrected model output, ranked against a reanalysis rather than against each model's own climate, so a model that runs moist will clear its 90th percentile more often than one day in ten.
Six-hourly to 120 hours. GFS and ECMWF both forecast further than 120 hours; this film stops there.
This map is drawn from model output we store ourselves, not from anyone's picture of it, and is based on data and products of NOAA, the European Centre for Medium-Range Weather Forecasts (ECMWF) and the Copernicus Climate Change Service. The per-city read of the same day — which cities the models are fighting about, and what the discussion makes of it — is on the national page.
Sources and licences
- NOAA: GFS, from NOAA's NOMADS server. US government data, not subject to copyright (NOAA/NWS terms of use).
- ECMWF: IFS, from ECMWF's open data. This service is based on data and products of the European Centre for Medium-Range Weather Forecasts (ECMWF). Source: www.ecmwf.int. This ECMWF data is published under a Creative Commons Attribution 4.0 International (CC BY 4.0). ECMWF does not accept any liability whatsoever for any error or omission in the data, their availability, or for any loss or damage arising from their use.
- Copernicus: ERA5 reanalysis, from NSF NCAR's copy. The PWAT normals are built from it. Contains modified Copernicus Climate Change Service information 2026. Neither the European Commission nor ECMWF is responsible for any use that may be made of the Copernicus information or data it contains. Published under CC BY 4.0.
What we change: each field is cropped to the frame these maps share when we download it, and each map then smooths, contours, differences or ranks those numbers as its Methodology says. These are our maps of the providers' numbers, not their products, and none of them has reviewed or endorsed this site.
Coastlines, borders, lakes and place names: Natural Earth, public domain.