I drive a lot, so I ask myself: How is it that we can know to the minute when we’ll arrive at our destination after an hour’s drive but not know something as basic as rainfall? The answer really comes down to the extreme variability of precipitation in time and space, along with the practical challenge of measurement. To measure is to know. In the case of rain, the challenge of measurement impedes our ability to know. Rainfall measurements must balance the wide extremes between drizzle and downpour, liquid rain and frozen hail, not to mention subtleties like wind blowing rain up and out of the gauge, rain coming down too hard for the bucket to keep up, sunlight evaporating fog and drizzle (more relevant for disease), leaves and bugs clogging the base of the funnel. It’s not hard to see why folks get so fed up with their tipping buckets even as they they have an insatiable desire for better rain data.
There are basically three types of rain gauges out there,
the “tipping bucket” type that is familiar to most of us, the “accumulation” type that you can get down at the hardware store, and a grab bag of “drop-counting” types that strive to measure, surprise, individual drops. If you’ve used RainWise or Davis or Pessl, you’ve used a tipping bucket. It’s a funnel and as the funnel catches water, it routes it into a little seesaw that counts up how many times it has tipped. The main problem with these is they fill up with schmutz, they lose their calibration, they lose water from too little rain (evaporation), too much rain (tipper can’t keep up) and wind, which blows rain out due to the aerodynamics.
If you’ve used a wedge gauge, that is an accumulator. The National Weather Service actually relies on accumulation gauges
, not digital, because they can be designed with no wind losses, some oil on top so they don’t evaporate, and they just fill up so there is no such thing as too heavy a downpour. Downside is that they are quite large. And manual! A person has to look at it and write it down. Various methods have been developed to allow for automation, including pressure transducers (wires that vibrate at a depth-dependent frequency), but these introduce as many errors as they solve.
Drop-counting rain gauges are quite interesting to us at Arable, not least because we have a couple of patents on them. But more relevant is that drop counters enable a link to the NEXRAD Doppler Radar Network
. Radar backscatter is, oddly enough, proportional to drop diameter raised to the sixth power, times the number of such drops. Back in the 1940s a couple of university professors
at McGill developed a relationship between the rain rate, the size, and the number of drops, which allowed for there to be a simple relationship between the radar signal and the rainfall. There’s some more details than that, but it remains a shortcoming of radar that it can’t really distinguish between a cloud with lots of small drops and a smaller amount of larger drops, which has immense consequences for estimating rainfall.