Anomalous Propagation Versus Real Echo on a Coastal Radar
You are nowcasting for a coastal airport at 06 UTC in January. The 0.5° reflectivity product from the WSR-88D 45 km away shows an arc of 25–35 dBZ returns between 15 and 35 km from the radar, strongest on azimuths that cross a river valley. The echoes appear abruptly after 04 UTC, are quasi-stationary, and dual-polarization shows correlation coefficient 0.55–0.75, differential reflectivity near 0 dB, and near-zero radial velocity. The 00 UTC sounding shows a strong surface-based inversion: 6 °C at the surface, 0 °C at 950 hPa, and +9 °C at 850 hPa, with calm surface winds and 15 kt southerly flow at 925 hPa. Surface observations report visibility above 10 km, dewpoint 5 °C, and no precipitation. A second, weaker echo area lies 120–150 km southeast with 15–20 dBZ and correlation coefficient 0.97. Terminal forecast is due in 20 minutes, and the tower reports no virga.
1. Decide whether the near-radar echoes are precipitation and justify it from beam geometry, refractive environment, and dual-polarization signatures.
Read the worked answer
Verdict: these are not precipitation. They are anomalous propagation (ground clutter via superrefraction).
Refractive environment The sounding is a textbook ducting profile: 6 °C at the surface rising to +9 °C at 850 hPa. A surface-based inversion makes the refractive index fall off with height far faster than the standard atmosphere; when dN/dz becomes steeper than about −157 km⁻¹ (equivalently dM/dz < 0), the 0.5° beam is refracted downward and strikes terrain, trees, and buildings.
Geometry At 45 km the 0.5° beam under standard 4/3-earth refraction sits only about 0.5 km AGL (45·sin 0.5° ≈ 0.39 km plus ~0.12 km of curvature term). That is inside the shallow inversion layer, so the beam is steered into the ground within 15–35 km - exactly the observed range limit.
Signatures
- Morphology: abrupt onset after 04 UTC as the inversion formed, quasi-stationary arc, strongest across the river valley where the beam intersects terrain.
- ρhv of 0.55–0.75 and ZDR near 0 dB are non-meteorological; real hydrometeors give ρhv above about 0.95.
- Near-zero radial velocity confirms stationary targets.
- Surface observations show more than 10 km visibility with no precipitation, and the tower reports no virga.
The arc is a beam-steering artifact and must be discarded before any nowcast.
2. How would you confirm the AP diagnosis operationally, and what is your assessment of the 120–150 km echo?
Read the worked answer
Confirming AP
- Compare tilts. Ground clutter from ducting is confined to the lowest elevation slice; if the 1.5° tilt is clean while 0.5° is full of 25–35 dBZ, the return is range- and tilt-dependent, not precipitation.
- Compare across the polarization variables and against the radar's clutter map.
- Compare with a neighbouring radar. Because AP depends on beam geometry relative to the duct, a second radar at a different range and azimuth will not show the same arc.
- Cross-check satellite (no cloud), surface observations, and any disdrometer. Any automated precipitation algorithm fed only reflectivity will produce a false rain area here; the human ρhv check is the discriminator.
The distant echo is real 15–20 dBZ with ρhv 0.97 is genuine light precipitation. But at 0.5° and about 135 km, the beam height is roughly 2.3 km AGL (135·sin 0.5° ≈ 1.18 km plus ~1.07 km curvature term). With dewpoint 5 °C and dry sub-cloud air, much of the light echo is likely virga evaporating before the surface, and the tower confirms no virga overhead yet. So it is real aloft but not evidence of rain at the terminal.
3. Write the operational nowcast message for the next six hours, and say what the true forecast hazard is.
Read the worked answer
Message No precipitation at the terminal through 12 UTC. The near-radar echoes are anomalous propagation and are excluded from the nowcast; the 120–150 km echo is real but high-based and is not expected to reach the surface before dawn.
Real hazard: fog and stratus The ingredients are all present:
- a strong surface-based inversion with a moist southerly feed at 925 hPa (15 kt) overrunning a calm surface layer,
- dewpoint 5 °C with a small surface dewpoint depression,
- overnight radiational cooling.
Expect radiation fog or low stratus to develop after roughly 08 UTC, with visibility 1–3 SM and ceilings below 1,000 ft. That, not rain, is the terminal hazard.
Two things worth saying explicitly
- Track the distant echo with the 925 hPa southerly steering flow (~15 kt north-northwest) - arrival would take 4–5 hours if it survives, so it stays on the monitoring list but not in the TAF as precipitation.
- Warn downstream users that the AP artifact will intensify before dawn as the inversion strengthens and will vanish within an hour or two after sunrise. Automated nowcast systems will show spurious rain through that period, so the artefact itself needs to be in the briefing.