The water could allow electricity to be conducted between the battery supply to the LED gauge and the throttle signal wires or the connector pins on the harness connectors.
If you read my
Water activated throttle post you will see that the water did not conduct enough electricity between the +5V and the throttle signal to make the motor run.
On the main harness connector, the throttle signal pin is located next to the Battery + pin , so a tiny drop of water could easily conduct electricity between the pins and cause the motor to run.
The other possibility is water getting into the rubber junction unit on the motor harness. Although they appear to be nicely sealed, I did find that the wires are not actually bonded to the rubber housing and it could be possible for water to be drawn into the junction unit by capillary reaction
(also described as wicking) combined with pressure variation.
If you make sure that the junction unit is completely dry you can run some very thin cyanoacrylate glue around each of the cables entering the block and this should seal it up completely.
I use ZAP CA Thin, but you need to be very careful as it wicks really well and will go everywhere if you apply too much:
I know from past experience that rainwater can penetrate easier than tap water can. I once spent a lot of time with a hosepipe trying to recreate a water leak in a sunroof on one of my cars, but I could not get any water to leak inside the car.
Two days later it rained and the water soon found its way into the car and started dripping off the sunroof while the car was still parked in the same position that it was in when I ran the hosepipe all over it. I think the rainwater must have less surface tension than tap water does.
Alan