The volume of water flow should/would also vary at different times of the year, based on how much the water table is loaded.

Not sure how much of this would be complimentary to the acoustic effect of temperature. Either way, it’s not a simple single-solution explanation.