My PhD advisor, Dr. Marla Sokolowski, at the University of Toronto always told us that she felt Drosophila adult or larval behavior was wonky in a thunderstorm. Since then I’ve paid close attention to the weather when running behavior experiments, and steered away from trying to do behavior experiments while there was a thunderstorm outside. In truth, I have noticed on those times that I do try to run behavior during the thunderstorm, the controls rarely work. Although I've never rigorously measured this, so it could easily be observational bias on my part, or something like a subconscious self-fulfilling prophecy type thing.
In any case I thought I’d pass this recommendation along. Don’t run behavior experiments in the lab when there is a massive thunderstorm (the thunder and lightning kind) outside, even when your room is temperature and humidity controlled.
There is some scientific evidence to suggest that this isn’t just superstition and biased observation on my part. Maybe the flies are sensitive to pressure and static in the air. Apparently Drosophila react to electric fields, and the amount they react is directly proportional to the field strength and inversely proportional to the infrasonic background (Chernyshev & Afonina, 1978). Atmospheric pressure apparently also affects flight (Chadwick & Williams 1949), mating (Austin et al 2014) and fertility (Andonyena et al, 2021). Atmospheric electric fields reduce locomotion (Edwards, 1960) and apparently flies avoid static electric fields (Ghamdi & Mesfer, 2012).
Overall, my recommendation is to skip doing behavior that day if you can. At the very least record the time the storm starts and stops and any weather metrics you can.