The Shelf Cloud at First Light

Local time is UTC+10, so every group needs ten hours added and the date rolled if it passes 24.

• Issued 231800Z = 0400 local on the 24th.
• Validity 2318/2412 UTC = 0420 local on the 24th to 2200 local on the 24th.
• TEMPO 2318/2402 UTC = 0420 to 1220 local on the 24th.

Your 0600–0700 window sits squarely inside the TEMPO period. That matters because TEMPO means the conditions are expected to occur for periods of 30 minutes or more - not 'might just brush past'. In that window the forecast allows visibility down to 3,000 m in thunderstorm rain, cloud at 1,000 ft, and cumulonimbus.

Now compare that with what you actually need. You are flying visually, keeping the aircraft in sight against a background of dark cloud and rain, at up to 120 m AGL. Three thousand metres of visibility in rain shafts is not a VLOS environment you can see a small RPA in, and CB cloud means convective turbulence and the possibility of a microburst or downdraught at the surface.

The forecast alone does not clear your window. It never did.

Start with the arithmetic, then add the meteorology.

• 60 km ≈ 32 nm.
• At 30 kt, time to the site = 32 / 30 hours ≈ 1.08 h ≈ 65 minutes, so the precipitation core reaches you around 0535 local - before first light.

But the echo is the last thing to arrive, not the first. A mature storm carries a cold pool beneath it, and the leading edge of that cold pool - the gust front - runs ahead of the precipitation, commonly 15–30 km and 30–60 minutes ahead. Radar will not show that boundary clearly.

The shelf cloud confirms it: that is the visible leading edge of the outflow, not the storm itself. It is already in your western sky at 0430, moving toward you at the speed of the boundary, not the speed of the cells. So the wind hazard can be at your site well before 0535, possibly before first light.

For a 6 kg-class multirotor on a fence line, a sudden shift from calm to 25–35 kt of cold outflow is not a 'come home carefully' event; it is a fly-away or a forced landing in the wrong place. The calm surface wind at 0430 is not reassurance, it is the signature of the moment before the outflow arrives.

Decision: no-go. The window is inside the TEMPO period, the outflow boundary is already visible, the cold pool will arrive around or before first light, and the only acceptable conditions for the task require clear sight of a small aircraft against the sky. Reschedule to the following morning and tell the client why in one sentence, with the radar picture as evidence rather than opinion.

If you are told to stay on site, define triggers before launch rather than deciding in the moment - deciding while a gust front hits is how people launch 'just for two minutes':

• Lightning within 20 km in the last 20 minutes - no flight, and no launch until 20 minutes after the last strike inside that radius.
• Gust spread over 15 kt, or any wind shift of more than 30° in ten minutes - no flight.
• Shelf cloud or dust front visible to the west and closing - no flight, regardless of the surface wind being calm.
• Visibility below 5 km in rain, or cloud base below 1,000 ft - no flight, because VLOS and see-and-avoid have both gone.
• Any aircraft already airborne when a trigger is met - land immediately at the nearest safe clear area, not at the launch point, if the launch point is downwind of the boundary.

Note that even in the best case the site is 25 km from the only TAF, so the local radar loop and your own eyes are the primary sources. That is a reason to be more conservative, not less.