Thunderstorm Build-Up on a Long Pipeline Patrol
You are planning a 42 km linear pipeline patrol with a fixed-wing UAS of 4.5 kg, cruise 22 m/s, endurance 95 minutes at 400 ft AGL. The route runs west to east. The morning briefing shows a surface trough to the west, convective available potential energy of 1 800 J/kg, and a forecast of scattered thunderstorms developing from 14:00. At 12:30 the automated weather station at the western end reports 31 degrees Celsius, dew point 22 degrees Celsius, wind 210 degrees at 8 kt, and a cumulus field with tops reported at 9 000 ft. Your launch site is at the eastern end, 42 km from the western end. You have one aircraft and one battery set, and the next crew is available at 16:00.
1. Assess whether the 12:30 launch should proceed, and justify your decision with the meteorological reasoning.
Read the worked answer
The temperature-dewpoint spread is 9 degrees Celsius, which is moist but not saturated at the surface. More important is the trend: CAPE of 1 800 J/kg with surface heating to 31 degrees Celsius and a trough to the west is a classic afternoon convective setup, and the cumulus tops at 9 000 ft already show deep vertical development. The 14:00 trigger time is 90 minutes away, and my sortie is 42 km at 22 m/s, which is about 32 minutes one way, or roughly 64 minutes for a there-and-back patrol, leaving little margin against a 95-minute endurance.
The critical issue is not the average weather but the tail of the distribution. Thunderstorms produce gust fronts, outflow, hail, lightning and severe turbulence, and a 4.5 kg fixed-wing aircraft has no meaningful tolerance for any of them. The storm motion is likely to be eastward or northeastward with the upper flow, which means the storms will move toward my launch site and along my route.
I would not launch at 12:30. The forecast trigger is 14:00, but convection often initiates earlier than forecast when a trough is present, and the cumulus tops already indicate that the cap is weakening. The safe decision is to wait, reassess at 14:00 with updated soundings and radar, and plan the sortie for the 16:00 crew window if the storms have cleared. The patrol is routine and repeatable; the weather risk is not.
2. Suppose you launch at 12:30 anyway and, at 13:20, 18 km from the launch site, you observe a developing cumulonimbus with an anvil 12 km to the west and a sudden wind shift at the aircraft. What do you do?
Read the worked answer
The anvil and the wind shift are both signs that the storm is mature and that its outflow has reached me. The wind shift is the gust front. I would immediately turn east toward the launch site, climb only if terrain or obstacles require it, and increase speed to the maximum continuous setting to shorten exposure. I would not attempt to fly around or under the storm.
I would compute the return time: 18 km at, say, 28 m/s is about 11 minutes, which is within the remaining endurance if the battery is healthy. I would monitor the wind shift's effect on ground speed; a headwind on the return leg could double the time. If the headwind makes the launch site unreachable with reserve, I would select the nearest pre-identified emergency landing area to the east, land, and recover the aircraft later.
I would notify the operator and any coordinating agency of the diversion, and I would not resume flight until the storm has passed and the gust front has cleared. After landing I would inspect the airframe for hail or debris damage and download the telemetry.
3. What pre-flight planning products and decision thresholds would have made this situation less likely?
Read the worked answer
I would use a convective forecast product with explicit initiation timing and coverage, a sounding or instability index, radar and satellite imagery, and a lightning detection feed. I would set hard thresholds in the operations manual: no launch if thunderstorms are forecast within 20 km of the route during the sortie window plus a 30-minute buffer; no launch if the temperature-dewpoint spread is below 8 degrees Celsius with CAPE above 1 000 J/kg and daytime heating ahead; abort if lightning is detected within 20 km; abort if a gust front or wind shift exceeding 15 kt is observed.
I would also plan the sortie to finish before the forecast trigger time, not to start before it. For a 64-minute patrol with a 14:00 trigger, the launch must be before about 12:30 only if the forecast is reliable, and the forecast is not reliable enough to cut it that fine. I would plan the patrol for the morning or the 16:00 window.
Finally, I would brief the crew on the abort criteria and the emergency landing areas along the route, so the decision is pre-made rather than improvised in flight.