Weather balloon
Follow your balloon
GPX / KMLImport a GPS track to compare the recorded flight with your forecast.
No GPS track imported.
Export from Find Me SPOT
In SPOT Maps, open Reports → Activity Report. Select one device and your flight’s date range, enable Show Details, then Run Report → Export → GPX (recommended) or KML. SPOT export guide ↗
Files stay in this tab and disappear on reload. No account connection or background tracking. GPX usually preserves timestamps; KML without timestamps is display only. Only absolute KML altitudes are used. Up to 5 MB / 20,000 points; export one flight at a time.
Update the forecast from a position
Predict a flight first, then update it from a GPS fix or timeline point.
Why estimate altitude?
Consumer trackers can omit, freeze, or limit altitude. SPOT lists a 21,320 ft (6,500 m) operating ceiling; this is a device rating, not a universal GPS cutoff. Receiver settings also matter: some GPS modes reject high-altitude fixes, and a 2D fix can hold the last known altitude.
We follow the original ascent, burst/float, and descent profile at the GPS timestamp. A consistent climb below 6,500 m can align its timing; missing, flat, or erratic readings do not. This estimate cannot detect an unexpected burst or guarantee better accuracy. Check launch time and equipment; use reported altitude for a validated high-altitude tracker, or enter a known altitude.
Only this fix and your flight settings are sent to SondeHub. The full track stays in this tab. Fixes older than 12 hours are display only.
Flight estimate
ForecastWhere will it land?
Choose a launch and check your equipment, then predict the flight.
Flight & wind
Band averages · explore the flight for wind at each point
Detailed winds · select to explore
How these bands work
Winds follow the balloon’s predicted location, altitude and time. Band boundaries stay fixed between ascent and descent; wind values follow the selected leg. Float winds follow the cursor at their actual altitude.
The overview combines nearby wind changes into a few readable layers. It considers speed and direction changes on both legs, keeping their wind averages separate. Each label is an average; the timeline and detailed list retain the smaller changes. Band heights stay on the true altitude scale.
Speeds marked ≈ are averages derived from movement between Tawhiri forecast points, not live observations or exact readings at every height. Band values average those intervals; the cursor shows its own interval. The calculation uses the flight’s weather run. Wind trails point downwind at an illustrative speed; they do not show vertical movement. Pause with Animate wind; reduced-motion preferences stop animation.
Winds over launch · reference
Predict a flight to load the launch-site wind profile.
Exact values · data table
| Height above sea level | Speed | From → toward |
|---|
NOAA GFS Global · Open-Meteo · launch-site snapshot
Forecast details & definitions
The date is the forecast hour nearest launch. Retrieved is when this browser received the data, not when the forecast was created. These are forecasts, not live readings; gusts are the preceding hour’s maximum.
Directions read from → toward: W → E means wind from the west carries the balloon east. Wind speed is horizontal; ascent and descent rates are vertical.
Aloft heights are forecast geopotential heights above sea level, not ground. hPa is air pressure; lower pressure generally means higher altitude. Only selected levels over the launch site are sampled—not every altitude or the balloon’s moving position. These pressure-level samples are launch-site reference data; they do not supply the route bands above.
The wind snapshot may use a different model run from the route. Open-Meteo does not provide its run timestamp here. The model run and age in the flight summary apply to the flight path only.
Assumptions & sensitivity
A forecast is not a guaranteed landing. Re-run near launch; wind and equipment differences can move the endpoint.
A sensitivity band around the active forecast (updated route when available). Tests rates ±10% and peak altitude ±1,000 m, one assumption at a time; descent-only tests descent rate, floating tests altitude ±250 m. Up to six extra forecasts on the same weather run.
This is not a measured margin of error or a confidence interval. It excludes wind forecast error and cannot promise the actual flight or landing will fall inside.
Predict a flight to estimate its sensitivity.
What if the launch changes?
Compare two alternatives using the same weather run. These show sensitivity, not a probability or landing zone.
Read the forecast
Start with the landing estimate. Use the profile to see how altitude and changing winds shape the route.
Why the balloon goes that way
A different wind at every height
A calm launch does not mean a short flight. Friction and terrain shape near-surface winds, while winds aloft can be much faster and blow in another direction. Changing speed or direction with height is called wind shear. The balloon drifts through those layers on ascent and descent; it does not simply follow the wind at the launch site.
What “prevailing winds” means
Prevailing winds are the usual wind direction over time. Across the middle latitudes, including most of the continental U.S., westerlies commonly move weather from west to east. Trade winds generally blow from the east in the tropics. These are broad patterns, not a promise for this launch: weather systems, mountains, season, and altitude can change the flow.
Where this forecast comes from
NOAA’s Global Forecast System (GFS) is a numerical weather model, run four times daily. CUSF’s Tawhiri predictor, hosted by SondeHub, uses forecast winds through the balloon’s changing position, altitude, and time to calculate the route. Each result shows the model run actually returned and its age.
The wind table adds a launch-site view from NOAA GFS through Open-Meteo. Both are model forecasts, not a nearby weather station or a live balloon sounding. The table is context for the route; it is not the data used to calculate it.
How this prediction works
Wind, altitude, time.
The latest available NOAA GFS forecast moves the balloon through changing winds. Tawhiri models a constant ascent rate, then a parachute descent that changes with air density. Float flights stop at the selected time; they do not predict a landing.
From hardware to flight.
Balloon mode estimates lift, expansion, and drag in the U.S. Standard Atmosphere. It assumes a spherical latex balloon, pure gas at air temperature, and no leakage. Ascent is approximated as a constant rate; parachute speed is referenced to sea level. Use direct rates for measured performance. Changing units preserves the same setup.
An estimate, not a landing zone.
Wind forecasts, burst height, and ascent rate can all shift the result. Terrain elevation is a 90 m grid estimate, not a survey. Re-run near launch and check recovery access and local flight requirements separately.
Open data. Light footprint.
Forecasts: CUSF / SondeHub and NOAA GFS. Terrain: Open-Meteo / Copernicus DEM. Search: GeoNames via Open-Meteo. Map: OpenStreetMap.
Free community services for noncommercial use. Your browser sends the selected coordinates and flight settings to these providers. After a successful prediction, one additional Open-Meteo request loads the launch-site wind profile. Saved sites stay in this browser; shared links contain the flight setup.