Papabravo said:
So your suggestion is to measure volume per unit time in units such as milliliters per minute. OK so we need a way to measure liquid volume, and we need a way to flush the measurement volume so we can repeat the measurement in the next time interval.
I would suggest that it might be easier to measure weight. Since we know the density of water we can compute the volume. Flushing the container then becomes more of a problem.
Rain is generally measured in depth, so you could say "we had three inches of rain last week".
Obviously all you need to do this is a container, and measure the depth of rain that collects in it. However, this is rather insensitive, so you would usually pour the water into a narrow measuring jug, which gives much greater resolution on the scale - obviously there's a conversion factor involved. Next 'improvement' is to collect the water in a funnel (for the large collecting area), and have the funnel feed directly into the measuring jug.
This is still manual though, and you have to empty the jug as well.
An electronic method is to use a 'see-saw' that works a switch (usually a reed, so there's no mechanical force required). The 'see-saw' accepts water from the funnel on one side, once it gets too heavy the see-saw changes over, the water is tipped out, and the funnel feeds the second side, until that gets too heavy, and so on. This gives a pulse from the reed relay, the speed of which is dependent on the level of rain fall.
You can count the number of pulses for a long term measure of the rainfall, or you can measure the time between them for a fairly real time measure of the amount of rain falling. Obviously it depends on how big the see-saw is, and how much water it requires to tip over.
EDIT: WELL!! - while I was typing someone posted a picture!!