Engines
Weather Trends
From the minute to the season, on our own model.
Storm nowcasting, a high-resolution 10-day forecast and a 6-month seasonal trend — from one engine.
Platforms
Weather Trends is São Tomás’s meteorological engine — the same one that feeds Risk360, licensed separately for teams that already have their own dashboard. The difference is not having a forecast; it is covering three decision horizons in one engine, each at the resolution that decision demands.
- minute by minute
- storm nowcasting
- 10 days
- operational forecast
- 6 months
- seasonal anomaly trend
- 9 → 2 km
- grid, convection resolved
Three horizons, three decisions
1
Minutes · the storm
Real-time cell nowcasting: stop the switching operation, pull the crew back, trigger the alert.
2
Days · the working week
A ten-day forecast on grids down to 2 km, with convection explicitly resolved.
3
Months · the season and the build
Seasonal anomaly trend out to six months, with three models side by side and the disagreement stated.
What the engine delivers
The rainfall that matters is not the daily one
Alongside the daily accumulation comes the rainfall of the most intense hour — that is what floods a city, not the 24-hour total. With it: minimum and maximum temperature, morning and afternoon humidity kept apart, and maximum wind.
Storms via GDI, with the caveat attached
Convective potential comes from the Galvez–Davison Index, in four bands. And we say what most omit: it is storm potential, not probability of rain. Confusing the two is the most common misreading of severe-weather forecasts.
Delivered over API
Data as a Service: clean data lands in your system, ready to feed a dashboard, an alert or your own model.
A Risk360 engine, also licensed separately. Coverage and integration scope are defined per project.
Who it is for
Technology teams that already have a dashboard, operations that want the forecast inside the system they already use, and projects combining meteorology with their own asset registry.
The rest of the suite
Risk360
From storm nowcasting to the six-month trend: risk on every critical asset, action prioritised in the right windows, and response sustained through the crisis.
ExploreClimate Edge
Climate scenarios translated into physical risk on each asset — to decide where to invest, what to reinforce and what inaction will cost, with documented methodology and stated uncertainty.
ExploreMapsAI
An AI agent over your maps: ask in Portuguese, English or Spanish, get back a map and a number — no GIS, no spreadsheet, no queue.
ExploreWildfire Sentinel
Detection, risk and spread cross-referenced against your registry: the alarm carries the asset name, the distance in metres and an estimate of the hours until the fire reaches it.
ExploreBring us your challenge
If it involves climate, territory and data, it is our ground. The conversation starts from a real slice of your operation — not from a deck.