Daily snow extent and melt-out dates, basin by basin.
Snow/no-snow classification every clear day and a melt-out time series for each basin you define, fed to a console, API or CSV.
Spec sheet
- GSD
- Wide / 10–30 m
- Cadence
- Daily, on clear passes
- Platform
- Wide-swath satellite
- Spectral
- Multispectral
- Delivery
- Console, API or CSV
- Unit
- One basin at a time, as you define it
The files
- Deliverable
Daily snow extent classification
- Format
Console or API- What's in it
- Snow or no snow, per pixel, for each basin, on every day with a usable clear pass. Days with no usable pass are marked as gaps and left empty. Snow-covered area for the basin comes with it.
- Deliverable
Melt-out time series
- Format
CSV- What's in it
- One row per day per basin: date, snow-covered area, and a flag for a gap day. The date snow cover drops off is marked in the series. It loads straight into the snowmelt or streamflow model you already run.
- Deliverable
Melt-out by elevation band
- Format
CSV- What's in it
- Where you hold a DEM for the basin, the series is split by elevation band so you can watch the snowline climb and see when each band goes bare.
- Deliverable
Console feed
- Format
API or CSV- What's in it
- The same daily record pulled into your own forecast tools, by API or as a CSV drop. It sits beside your gauge and precipitation data as one more input.
How it works
01You supply
Define the basins
Send the catchment polygons you forecast for, as a
shapefileorGeoPackage. Add theDEMif you want elevation bands.02We do
Daily pass
A wide-swath multispectral satellite image over each basin, each day one is available.
03We do
Classify snow cover
Each pixel is classified as snow or no snow from its spectral response. Cloud-covered pixels are masked out and carried as gaps, never filled.
04Comes back
Delivery
Extent and time series go to your console or feed each day there is a usable pass, in the format you chose.
Limits
These are the conditions that change what the series can show.
- Condition
Cloud sets the gaps
- Effect on the output
- Optical imagery needs a clear daytime view. Persistent cloud leaves gaps in the daily series, and a stormy week can leave a long one. Daily is the target, tied to the clear passes that exist.
- Condition
Extent, not snowpack
- Effect on the output
- This classifies where snow lies. It does not measure depth, density or snow water equivalent, so it cannot say how much water is left in the snow.
- Condition
10–30 m is the floor
- Effect on the output
- A patch of snow smaller than a pixel or two does not show, and a pixel half covered gets called one way or the other. Fringe snow at the melt edge is the least certain part of any day's map.
- Condition
Trees and shadow
- Effect on the output
- Dense canopy can hide snow lying under it, and deep terrain shadow limits what the sensor sees on steep slopes. Both are reasons a basin's classified area can differ from what a ground observer would report.
- Condition
Observed, not forecast
- Effect on the output
- The series reports melt-out as it happened. Turning that into a streamflow or hazard forecast is the job of your own model.
Why this exists
Your snow picture comes from survey courses, pillows and stations, which are points, plus a satellite look taken once a season or at a scale too coarse for a small headwater basin. When the freshet starts and someone asks where the snowline sits today, the answer is several days old or inferred from gauges that have already moved. This replaces waiting on that to know where melt stands now.
Who it's for
For
- Hydrological forecasters at river authorities, water utilities and flood warning agencies who need snowline position between surveys.
- Water-supply forecast desks that want a daily snow-extent input beside gauge and precipitation data.
- Avalanche and flood hazard desks that want observed melt-out timing as an input to their own assessment.
Not for
- Teams that need snow water equivalent, depth or volume. That takes snow pillows, survey courses or airborne surveys.
- Teams looking for a flood, streamflow or avalanche forecast. This is an input to one.
- Hourly or real-time needs. The cadence is daily at best, and only on clear passes.
- Single-slope or point-scale snow questions, where a field observer or a drone will see more.
Questions from forecast desks
Does this measure snow water equivalent?
No. It classifies snow extent from optical imagery. It does not measure depth or water content.
Does it see through cloud?
No. Optical imagery needs a clear daytime pass. Persistent cloud leaves gaps in the daily series, and the gaps stay visible.
Can it predict flood timing or avalanche risk?
No. It reports observed melt-out timing. Forecasting streamflow or hazard from it is your model's and your desk's job.
How do we get the data into our tools?
By console, API or CSV. Tell us what your forecast models load and we set the fields and file layout to match at scoping.
How do we check it against our own snow observations?
Compare the classified extent with your station and survey readings and with any camera or field observations on the same dates. Give us where your ground truth sits when the basin is scoped and we can set it beside the series.
How many basins can you cover?
As many as you define. Each basin is scoped from the polygons you send.
Who owns the delivered data?
Use and ownership terms for what we deliver are written into the contract before work starts. The imagery providers keep their own licences on the source imagery.
How is it priced?
Each project is scoped and quoted on its own, from the number and size of basins and the delivery format. You get a written scope to take through procurement.
Send the basin polygons and put snow extent on your forecast desk.
Or .