Why some contrails persist and spread

One newly formed trail, two very different fates

Immediately after formation, a contrail contains many small ice crystals. In air below ice saturation, those crystals sublimate—solid ice changes directly to water vapor—and the line may disappear in seconds or minutes. In air above ice saturation, vapor deposits onto the crystals. The trail remains, accumulates ambient water, and can spread into contrail cirrus.

Ice-supersaturated regions: the invisible fuel for persistence

Relative humidity depends on the reference surface. Weather apps usually report near-surface relative humidity with respect to liquid water. Persistent contrail analysis instead asks about relative humidity with respect to ice (RHi) near flight altitude. When RHi exceeds 100 percent, the air contains more water vapor than is in equilibrium with a flat ice surface, so exposed ice crystals can grow.

Ice-supersaturated regions (ISSRs) are often shallow and have sharp edges. A flight can enter and leave one in seconds, producing a trail that starts or stops abruptly. Multiple aircraft crossing the same region can produce parallel or intersecting trails, while nearby flights at another altitude make no persistent trail.

From narrow line to cirrus-like cloud

1. Jet and vortex phase

   Within seconds, the plume is entrained into the aircraft wake. Counter-rotating vortices can pull much of the young trail downward while a secondary wake remains nearer flight level.

2. Dissipation phase

   After the organized wake breaks down, turbulence mixes the ice with the environment. Dry layers erode it; supersaturated layers sustain it.

3. Diffusion and spreading

   Vertical wind shear moves the top and bottom of the cloud at different speeds or directions. Falling ice crystals can enter adjacent layers, creating fall streaks and broadening the feature.

4. Loss of a unique signature

   With time, the straight centerline bends, fragments, and blends with surrounding cirrus. At this stage a single visible image may no longer distinguish aircraft-induced from natural ice cloud.

NASA-supported analysis of a persistent SUCCESS campaign case modeled 15–180 minutes of evolution and found that high ambient ice supersaturation and moderate shear could explain observed spreading and ice growth. That result is a well-instrumented case, not a universal lifetime formula.

Why persistence is difficult to forecast precisely

Numerical weather prediction divides the atmosphere into grid boxes. A thin ISSR may occupy only part of a box, and routine observations provide fewer direct upper-tropospheric humidity constraints than surface variables. Humidity sensors and models also have biases in very cold air. A grid-cell average can therefore look too dry while a narrow layer is supersaturated—or too moist while the flight path is dry.

Contrail Research adds four additional mismatches that users should keep in mind:

• an aircraft’s pressure altitude is not always its exact geometric height;
• the model uses selected upper-air samples rather than a complete vertical profile;
• weather is hourly while aircraft reports arrive at irregular, faster intervals;
• the displayed grid center is not the boundary of the humid layer.

Consequently, a “dry” model miss does not disprove an observed persistent trail, and a supportive forecast does not guarantee that one will appear. Forecast skill must be reported statistically across many labeled cases.

How to document persistence from the ground

1. Use a time series. Photograph the same region every one to two minutes for at least 15 minutes, preserving original timestamps.
2. Include orientation. Record compass direction, approximate elevation angle, and stable landmarks when safe and practical.
3. Watch the oldest section. A fresh line continuously created behind an aircraft is not yet evidence that individual ice crystals persist. Track a fixed segment after the aircraft leaves.
4. Separate movement from growth. Translation with the wind changes location; increasing width or optical depth suggests evolution.
5. Retrieve context promptly. Save flight and weather sources with UTC times before provider windows expire, but treat a nearby flight as a candidate rather than a confirmed source.

Four interpretation rules

Many trails at once

Usually indicates overlapping traffic with a shared supportive upper-air region. It does not show that every aircraft or altitude had identical conditions.

Sudden gaps

Consistent with horizontal or vertical humidity variability, though incomplete observation, occlusion, and perspective remain alternatives.

Broad haze later

Can be consistent with aged contrail cirrus, natural cirrus, or a mixture. Attribution confidence normally decreases as linear identity is lost.

No model cloud

Could mean genuinely unfavorable air, a weather-model miss, a pressure-level mismatch, or unavailable data. Check the source/status before interpreting absence.