The Graphical Turbulence Guidance, the Clear Air Turbulence Forecast Nobody Told You About, and the Free NOAA Tool That Predicts the Invisible Before You Fly Into It

The Graphical Turbulence Guidance (GTG) is a free, publicly available forecast product from NOAA’s Aviation Weather Center that predicts turbulence probability across the entire continental United States from the surface to flight level 450, updated every hour. It covers the one type of turbulence your radar will never show you: clear air turbulence with no clouds, no convective activity, and no SIGMET on your route. Two extra minutes in the preflight brief is all it takes to check it.

What Is the Graphical Turbulence Guidance?

GTG was developed through a collaboration between NOAA’s Aviation Weather Center in Kansas City and researchers at the National Center for Atmospheric Research (NCAR). It has been operational in some form since the early 2000s and lives on the Aviation Weather Center’s public website at aviationweather.gov.

The Aviation Weather Center is the same government shop that produces the aviation weather products pilots already use daily: SIGMETs, convective outlooks, and prog charts. GTG sits alongside those products and addresses a specific gap none of them fill well.

Why Don’t Radar and SIGMETs Cover Clear Air Turbulence?

There are two broad categories of turbulence that send passengers to hospitals and put aircraft through out-of-envelope structural inspections. The first is convective turbulence - the bumps inside and near thunderstorms. Onboard radar shows the cells. Convective SIGMETs give the regulatory picture. You can see it and route around it.

The second category is clear air turbulence (CAT). No clouds. No radar return. No visual cue. You’re in blue sky at FL350 and the atmosphere delivers a fifty-foot upset with zero warning.

CAT forms at the boundaries between airmasses moving at very different speeds and directions. The jet stream is the classic source - where the high-speed core meets slower air to either side, intense vertical wind shear generates chaotic rolling motion in apparently calm air. Mountain wave turbulence is another major contributor: air flowing over significant terrain sets up wave patterns on the downstream side that can extend hundreds of miles. Frontally induced turbulence, tropopause folds, and gravity waves propagating outward from distant convection all produce turbulence in radar-transparent air.

The consequences are documented and ongoing. In May 2024, Singapore Airlines Flight 321 encountered severe clear air turbulence over the Andaman Sea during a thunderstorm avoidance maneuver. One passenger died of a suspected cardiac event triggered by the upset. Several dozen were seriously injured - meals were on trays, some passengers were unbelted. Research published over the past decade indicates that severe CAT encounters at cruise altitude are trending upward as climate change intensifies jet stream variability.

PIREPs offer the most direct information but have structural limits. Coverage is inconsistent across lower altitudes, the Gulf, and large portions of general aviation airspace. Even where PIREPs exist, they are snapshots from one aircraft at one point in time - they tell you what was there, not what’s building an hour ahead on your route.

Turbulence SIGMETs operate at a coarse scale by design. A turbulence SIGMET might cover a geographic box the size of Nebraska. The forecaster draws conservatively to encompass the area of potential encounter. Inside that box, you might find rough air in one corridor and completely smooth air one hundred miles south. SIGMETs don’t carry the precision needed to route around them meaningfully.

GTG was built to close that gap.

How Does GTG Predict Turbulence No One Can See?

GTG doesn’t directly observe turbulence. It runs a suite of diagnostic mathematical algorithms against the output of high-resolution numerical weather prediction models - specifically two models pilots should know by name.

The Rapid Refresh (RAP) runs every hour across the continental United States at 13-kilometer resolution. The High-Resolution Rapid Refresh (HRRR) runs at 3-kilometer resolution and is particularly effective at capturing convective weather and terrain-driven phenomena. Both models ingest continuous observational data from radiosondes, aircraft reports, surface stations, and satellite imagery, then compute a three-dimensional picture of the atmosphere updated hourly.

The turbulence algorithms scan the model output fields for specific physical signatures:

• Vertical wind shear: how quickly wind speed and direction change with altitude
• Deformation field: how an airmass is being stretched and distorted horizontally
• Divergence and convergence patterns
• The Richardson number: a ratio atmospheric scientists use to assess whether buoyant forces or shear forces dominate a given layer - when the Richardson number falls below a critical threshold, that layer is likely to become turbulent

Any single diagnostic in isolation produces a noisy picture. What distinguishes GTG is the ensemble approach: roughly 10 to 12 separate diagnostic algorithms run independently against the same model data, each answering the same fundamental question from a different physical perspective. Their results are combined using a statistical weighting scheme calibrated against a large historical database of real-world turbulence encounters.

The output is expressed in energy dissipation rate (EDR), the aviation industry’s standard unit for turbulence intensity. Major carriers equip their aircraft to automatically log and transmit EDR data through their ACARS systems whenever acceleration thresholds are crossed. That automated data feeds directly back to the Aviation Weather Center as real-time validation.

For pilots, GTG translates EDR values into the familiar categories: light, light to moderate, moderate, moderate to severe, and severe.

What Do GTG’s Altitude Bands and Forecast Times Cover?

GTG organizes its output into three altitude bands matched to where pilots actually fly:

• Surface to 10,000 feet: the low-altitude environment for most general aviation
• 10,000 feet to FL180: the mid-altitude environment
• FL180 and above: the high-altitude jet and turboprop world

Select the band that matches your planned cruise altitude and you get a turbulence probability picture specific to that slice of atmosphere.

GTG also produces forecasts, not just current conditions. Forecast valid times extend from current conditions out to 18 hours. For a long cross-country, select the forecast valid time closest to your estimated time over each stretch of your route. You get a spatially specific probability picture for when you will actually be there - something PIREPs structurally cannot provide.

How Do I Read GTG Before a Cross-Country Flight?

The product is accessible at aviationweather.gov under the weather products menu. Select the turbulence section, then GTG. Choose your altitude band and the forecast valid time that brackets your planned en route time over the areas you want to evaluate.

The color scale is intuitive:

• Green: low or negligible turbulence probability
• Yellow: elevated probability for light to moderate turbulence
• Orange and red: conditions where rerouting or altitude adjustment is warranted in planning

Build a composite picture rather than relying on GTG alone. Compare it against PIREPs for the same area and check the SIGMETs for context.

When GTG shows elevated probability and PIREPs confirm it, that’s a high-confidence forecast. When GTG flags an area with no confirming PIREPs, treat it as a real risk that hasn’t been observed yet - the turbulence may be there and simply not yet reported. When PIREPs show bumps that GTG missed, trust the PIREPs. Observed reality beats model output every time.

Where Does GTG Fall Short?

Mountain wave turbulence remains the hardest problem. The algorithms have improved with the 3-kilometer HRRR data, but complex terrain-induced turbulence over the Rockies, Cascades, Sierras, and Appalachians remains difficult to model at a granular level. Near significant terrain on a day with strong winds aloft, GTG is your starting point, not your answer. Cross-check it against PIREPs from aircraft in the area, consult the wave SIGMETs, and respect what the mountains can do on a clear-sky day.

Low-level turbulence below roughly 3,000 feet AGL is also challenging. Mechanical turbulence from surface winds over ridges, thermal convection over sun-heated terrain - GTG can show elevated risk in those areas, but fine-scale detail at low altitude is limited by model grid size.

GTG should be read as probability, not certainty. An area showing moderate turbulence probability means the atmospheric conditions that produce moderate turbulence are present and the diagnostic algorithms agree. It does not mean every aircraft transiting that area will encounter it. Turbulence is patchy and intermittent by nature. The smart call is to plan around elevated probability areas when routing flexibility exists.

How Is GTG Getting More Accurate Over Time?

GTG’s accuracy is bounded by how well its algorithms can be calibrated, and calibration requires large volumes of real-world validation data. That data historically came from manual PIREPs - sporadic, inconsistently formatted, spotty in coverage.

Automated turbulence reporting from airline fleets has changed the calibration picture significantly. When a major carrier’s aircraft logs an EDR-based encounter automatically, that observation returns to the Aviation Weather Center in near real-time. The volume of automated reports dwarfs what manual PIREPs ever provided. More data means better-calibrated statistical weights and more accurate forecasts.

The next frontier is extending that reporting network. Automated turbulence data currently comes predominantly from major airline aircraft. Regional carriers, business aviation, and general aviation are largely not contributing. Active research is exploring whether vertical acceleration signatures derivable from existing ADS-B transponder data could extract turbulence observations from the broader fleet. If that matures into an operational capability, the validation dataset expands significantly and the whole system improves with every flight that passes through rough air.

Machine learning approaches are also being tested against the traditional ensemble methodology. Early published results from the Aviation Weather Center and NCAR researchers suggest accuracy improvements, particularly for clear air turbulence over oceanic areas where observational data is thinnest. The trajectory of capability improvement over the last twenty years is real, documented, and ongoing.

Key Takeaways

• The Graphical Turbulence Guidance (GTG) is a free, hourly-updated forecast at aviationweather.gov that predicts clear air turbulence across the continental US from the surface to FL450
• GTG uses an ensemble of 10–12 diagnostic algorithms run against the RAP and HRRR models, calibrated against real-world pilot reports - it is meaningfully more than a single-model product
• The product covers three altitude bands (surface–10,000 ft; 10,000 ft–FL180; FL180+) and forecasts out to 18 hours, letting you evaluate conditions for your actual time over a given stretch of route
• Orange and red areas warrant rerouting or altitude adjustment in planning; cross-reference against PIREPs and SIGMETs to build a composite picture
• Mountain wave turbulence and low-level mechanical turbulence remain the product’s weakest areas - treat GTG as the starting point near significant terrain, not the final word