WeatherNext 3 Hurricane Forecast: Tracking Storms Hour by Hour
Tropical systems are where forecast detail matters most. WeatherNext 3 outputs cyclone tracks natively and refreshes every hour, and Google DeepMind has published a case study describing how the technology helped the National Hurricane Center predict a historic landfall. This page explains what hourly AI guidance adds to storm tracking, where it stops, and how to follow a storm's next 24 hours in this workspace. New to the model? Start with what WeatherNext 3 changed or how to read hourly forecasts.
7-day free trial, then $19/mo · cancel anytime · compare plans
Why hourly matters when a storm is closing in
A landfall window can tighten by hours between one model run and the next. WeatherNext 3 changes how often that information refreshes and how much detail comes attached.
A new forecast every hour
Older cycles refreshed every six hours with a data lag on top. Hourly updates let you watch a track or intensity shift settle instead of waiting half a day to learn what changed.
Direct satellite input
The model learns from live geostationary satellite mosaics rather than lagged numerical analyses - the change that makes an hourly cadence possible at all.
Native cyclone tracks
WeatherNext 3 outputs discrete cyclone tracks alongside its gridded fields, so storm positions come from the model itself rather than a separate post-processing step.
Sharper landfall detail
Temperature and humidity at 5 km and wind at 10 km give coastal communities, valleys and islands more usable detail as a storm approaches land.
How to follow a storm hour by hour
A repeatable routine for hurricane and typhoon tracking once a system is inside the forecast window.
Search the storm region
Open the forecast for the coastal area in the storm's path rather than the open ocean, and set the time window to the landfall period you care about.
Switch to wind and precipitation
Wind speed shows the extent of damaging conditions; precipitation shows where the core is organizing. Read both before trusting a single map.
Step through the hours
Move through the hourly sequence instead of one timestamp, and note the arrival time of the threshold that matters for your decision.
Re-check after each update
Compare the newest run with the previous one. Small steady shifts are normal; a persistent drift across several runs is the signal worth acting on.
The National Hurricane Center case, read carefully
In its launch materials, Google DeepMind published a case study describing how WeatherNext helped the National Hurricane Center better predict Hurricane Melissa's historic landfall in Jamaica. Separate research describes the WeatherNext Cyclones line running operationally during the 2025 Atlantic season; in October 2026, news coverage described the center drawing on Google's weather AI as Hurricane Isaias approached the Gulf Coast.
AI guidance is in operational hands
These are forecasts being evaluated inside real forecast operations, not lab demos - which is also why the model keeps changing under live evaluation.
It aids forecasters, it does not replace them
Official warnings, evacuation orders and advisories still come from national meteorological services and hurricane centers. Nothing on this site changes that.
The track record is young
WeatherNext 3 launched in September 2026. The case studies are specific and promising; the honest reading is that the operational evidence is early and growing.
WeatherNext 3 hurricane questions
Short answers about storms, tracks and official warnings.
Preparing for a season rather than a storm? Reading the hourly forecast covers the basics.
Follow the next storm hour by hour
Start the 7-day free trial and track a real system as it develops, with hourly updates for the exact coast you care about. For background, see what the model changed or the data channels behind the tracks.
Related guides
The pages that go with this one, chosen by what each one answers.
