Flyfun Weather is an exploratory weather briefing tool for general aviation pilots. Its goal is to help you form an early picture of how weather patterns are coming together by comparing multiple numerical weather prediction (NWP) models. At its core, the briefing translates model data into aviation-focused advisories — icing risk, convective potential, cloud ceilings, turbulence, crosswinds — rated green/amber/red so you can quickly see what matters for your route. An altitude advisory even suggests the best flight level based on the combined conditions.
Beyond the advisories, Flyfun Weather lets you see where models agree and where they diverge, so you can gauge forecast confidence days before a planned flight. The briefing offers two display modes: compact mode for a quick overview of the essentials, and full detail mode that exposes all the raw data — soundings, cross-sections, thermodynamic indices, per-model values — with explanations, giving you full transparency into the numbers behind any summary.
Flyfun Weather is designed primarily for longer-range planning — helping you build a picture of weather trends days before a flight. As the flight date approaches, it adds real-time METAR/TAF observations for a final cross-check, but you should always verify against official sources (SIGMETs, AIRMETs, national aviation authority briefings) before flying.
Flyfun Weather is open-source and available on GitHub.
Flyfun Weather is updated frequently with new features and refinements. Check the What's New tab to see the latest changes.
The briefing is organized into collapsible sections. Here's what each one shows:
At the top of every briefing, the toolbar shows the route, gives you access to Refresh, PDF download, Email, Feedback, and the Tour button (the ?-icon). A Compact / Full Details toggle controls how much detail each section shows — compact is best for a daily glance, full for deep dives. The history dropdown lets you re-open any past snapshot to compare how the forecast has evolved. Flight owners can also toggle privacy (private flights are owner-only; public flights get a shareable short link via the Share icon) and configure a daily auto-refresh with email notification.
The banner at the top of the briefing is a colour-coded one-line verdict (GREEN, AMBER, or RED) derived from the worst-rated advisories. It's the headline read on your flight before you dig into the detail below.
For flights more than about a week out — beyond the high-resolution model horizon — the banner instead shows a dashed Early outlook badge (Trending Settled, Mixed Signals, or Trending Unsettled) rather than a green/amber/red verdict. This far ahead the models only agree on the broad pattern, so the briefing gives a confidence-led trend and tells you when more detailed analysis will become available.
Advisories are the heart of the briefing. They translate raw NWP model data into aviation-relevant assessments that help you make early decisions about flying, cancelling, or adjusting the timing of a flight.
Each advisory is rated GREEN, AMBER, or RED based on configurable thresholds. Advisories are evaluated per-model, so you can see whether models agree on a hazard or only one flags it. The summary rating shown for each advisory reflects the worst-case across all models.
Advisory categories include:
You can enable, disable, and tune the thresholds of each advisory in Settings, as well as choose which analysis methods are used for icing and cloud detection.
On the day of the flight, a METAR/TAF section is shown with current observations and TAF forecasts for your departure and arrival airports, fetched from official aviation weather sources. Useful for cross-referencing the NWP model output with real-time and near-term observations.
Alongside the airport observations, any SIGMETs that intersect your route corridor (turbulence, icing, thunderstorms, mountain wave, etc.) are listed with their hazard type, altitude band, and validity. When you refresh on the day of the flight, a banner highlights any conditions that have worsened since your last check — a degraded flight category, or a new or escalated SIGMET — so changes don't slip by while the rest of the briefing stays put.
As the flight gets close (about two days out), if your destination's forecast looks marginal, Flyfun Weather suggests nearby alternate airports that would fix the specific problem — a better flight category, lighter wind, or a kinder crosswind. Each candidate is shown as reachable before the destination (an early-decision diversion that avoids backtracking) or after it, with the extra distance each option costs. These are planning-grade divert ideas, not an operational alternate-minima calculation. Turn them on with the Alternates preference in Settings.
For the destination, the briefing also estimates whether the regulations would require you to file an alternate, computed two ways: FAA (14 CFR 91.169 — a binary Yes/No) and EASA Part-NCO (a Likely / Marginal / Unlikely confidence band). Because the published plate minima aren't known, the unknown value is expressed as a plausible range per approach type, which sets the width of the EASA band. Like the alternates above, this is advisory planning guidance, never a go/no-go verdict.
An LLM-generated plain-language overview of the weather situation along your route. It summarizes the key findings from the numerical analysis in a concise narrative. This section requires the AI Digest service to be enabled in Settings.
An optional synoptic weather discussion sourced from DWD (Deutscher Wetterdienst), providing a broader meteorological context for European flights.
Surface analysis and short-range forecast charts published by DWD — fronts, pressure centres, isobars across Europe. Shown alongside the synoptic text so you can see the broader pattern around the flight day. Click any chart to enlarge.
A vertical cross-section chart from autorouter.aero showing cloud, wind, temperature, and turbulence along your route. This requires the GRAMET service to be enabled and Autorouter credentials configured in Settings.
An interactive canvas-based vertical cross-section built from the NWP model data. Layers are organised into groups (clouds, icing, turbulence, convection, temperature lines, terrain) and you can toggle them individually in the control panel.
A scalar graph below the cross-section, sharing the same x-axis (route distance). Pick any two metrics for the left and right Y-axes — wind speed, ceiling, CAPE, crosswind, temperature, and more. Useful for seeing how a single quantity evolves along the route alongside the vertical layers above.
An interactive map showing your route with segments colored by a selectable weather metric (wind, icing risk, cloud cover, humidity, etc.). Use the altitude slider to see how conditions change at different flight levels. Hover over a segment to see the value at that point.
Skew-T / Log-P diagrams for any point on your route — click a point on the cross-section to load its sounding. Three view modes are available:
A table of thermodynamic indices derived from each sounding: CAPE, CIN, lifted index, freezing level, cloud layers, icing zones, and more. Values are shown per-model so you can spot where models agree or diverge.
A divergence analysis that scores how well the models agree on key variables at each waypoint. High agreement gives confidence in the forecast; poor agreement signals uncertainty.
Throughout the briefing you'll see small (i) buttons next to metrics and layer names. Clicking one opens a popup explaining what the metric measures, what the thresholds mean, and how to interpret the values in a flying context.
Flyfun Weather pulls each NWP model from multiple sources and blends them. Where available, raw GRIB2 from the originating forecast centre (ECMWF, DWD, NOAA) provides the high-resolution upper-air sounding and cloud microphysics; outside those coverage areas (or beyond the direct horizon) Open-Meteo fills in. The table below is generated live from the server's source registry, so what you see is exactly what the briefing engine has available right now — including the most recent run, how far it covers, and when the next update is expected.
Full per-metric attribution lives in the Analysis Metrics Reference; the engine-level details are in Weather Engine Specs.
Open the full Data Sources tab → for every column we track (model family, resolution, cycles, latest run, covers until, next update).
When you enter a waypoint that isn't an airport (a 5-letter fix, a navaid identifier, or a free-route point), Flyfun Weather looks it up in a local navigation database built from four free, public sources. The database is rebuilt on the 28-day AIRAC/NASR cycle and currently contains ~95,900 waypoints after deduplication.
| Source | Coverage | Format | Link |
|---|---|---|---|
| Eurocontrol FRA | ~8,100 ECAC waypoints (FRA-significant 5LNC + NAVAIDs), per-AIRAC | Excel .xlsx | eurocontrol.int |
| OpenNav | ~16,000 waypoints across 32 European countries (broader than FRA) | HTML scrape | opennav.com |
| OurAirports NAVAIDs | ~9,900 worldwide NAVAIDs (VOR/DME/NDB/TACAN/VORTAC) | CSV | ourairports-data |
| FAA NASR Fixes | ~70,200 US named fixes (intersections, RNAV, reporting points) | Zipped CSV | faa.gov NASR |
All four sources are free, require no authentication, and refresh on the 28-day AIRAC/NASR cycle. When the same identifier appears in multiple sources, the deduplication priority is Eurocontrol FRA → OpenNav → OurAirports → FAA NASR — FRA wins ties because it carries the richest metadata (FIR codes, level availability).
| Data | Provider | Notes |
|---|---|---|
| METAR / TAF | NOAA Aviation Weather Center (aviationweather.gov API) | Primary source for live observations and terminal forecasts. For a manual cross-check (especially in Europe), see Ogimet. |
| GRAMET cross-section | autorouter.aero | Requires you to connect your personal autorouter account in Settings. The GRAMET image is fetched on your behalf using your credentials. |
| DWD synoptic discussion | Deutscher Wetterdienst (opendata.dwd.de) | Short-range and medium-range synoptic text bulletins for European weather context. |
| NWS Area Forecast Discussion | NOAA / NWS (aviationweather.gov, api.weather.gov) | Local forecaster discussions from the relevant NWS office, used when the route includes US airports. |
| AI synopsis & translations | Anthropic Claude | Plain-language briefing summary and DWD German-to-English translation. Optional — toggle in Settings. |
Flyfun Weather runs an MCP server, so you can work with your briefings from inside an AI assistant such as Claude. Instead of clicking through the site, you simply ask in plain language.
Every answer links back to the full interactive briefing here on the site for cross-sections, Skew-T diagrams, and route maps.
In claude.ai:
https://mcp.flyfun.aero/weather
You'll be asked to sign in with your Flyfun Weather account the first time, so the assistant only ever sees your own flights and briefings.
We also have a FlyFun Weather custom GPT that gets the same information as the MCP server above. It's newer and its behaviour is less tested than the Claude integration, so treat it as experimental. As with Claude, you'll sign in with your Flyfun Weather account the first time, so it only ever sees your own flights and briefings.
Flyfun Weather is currently in early preview. To request access, sign in and wait for an email confirmation that your account has been enabled.
To suggest improvements, report bugs, or give any feedback, please create an issue on GitHub.
Flyfun Weather is designed to help and support your flight planning and decision-making — giving you an early, multi-model picture of the weather and surfacing the factors that matter for your route. It relies on automated analysis and AI, which can make mistakes or miss things, so treat its output as one input among many rather than the final word.
It is not a substitute for official weather briefings from your national aviation authority. Always cross-check with official sources (METARs, TAFs, SIGMETs, AIRMETs) before flying. The pilot in command remains the sole decision-maker and is responsible for the safe conduct of the flight.
This application was built with AI assistance (Claude, Anthropic).
Loading…