Please stop hour-averaging the rainfall rate
Can you explain to me, please, the reason for your practice of hour-averaging the rainfall rate? Why is rainfall rate treated differently than any other weather parameter being measured at the same time? It does not make any sense.
Imagine that you reported a rolling one-hour average for temperature or wind speed. That would obscure rapid changes in temperature or wind speed, just as you are currently obscuring sudden changes in rainfall rate. That would be stupid, wouldn't it? It is equally stupid to do that for rainfall rate.
Furthermore, if it rains for only 5 minutes, your hour-averaged data shows that it rained for an entire hour! Ridiculous!
Please stop hour-averaging the rainfall rate.
r75173
shared this idea
3 comments
-
r75173
commented
JEBTucson asked me to clarify how I believe that rainfall rate should be reported. I hope that the following explanation helps.
Rainfall rate should be reported based upon how fast it is raining at the time of the current observation, similar to temperature, dew point, wind speed, etc. The existing practice of reporting rainfall rate based upon how much it rained in the last hour is bogus and entirely different than how other weather measurements are reported.
The rainfall rate is reported as 'inches per hour,' so the formula that your software should use to calculate rainfall rate depends upon the time interval between observations. Some stations report observations to WU every 5 minutes; other stations report every 15 minutes; while other stations use a different interval between observations.
As a simple example that is easy to calculate, let's assume that you are reporting observations to WU every 15 minutes and it is now time to report another observation because 15 minutes have elapsed since the last report. If your rain gauge shows that 1 inch of rain fell in the last 15 minutes, then the rainfall rate that should be reported for the current observation is 4 inches per hour.
In general, if we assume that 'M' is the amount of time measured in minutes since the last observation and 'A' is the accumulated amount of rain measured in inches since the last observation, then the rainfall rate that should be reported for the current observation is equal to A multiplied by 60 and then divided by M. (The factor of 60 is included in the formula to convert minutes to hours.)
In the specific case of the simple example given above, A=1 inch and M=15 minutes, so the formula for rainfall rate should be A x 60/M = 1 x 60/15 = 4 inches/hour.
I believe that this would be a much better way of reporting rainfall rate than the existing practice, which is to add up all rain accumulation for the past hour. One of several deceptions introduced by the existing practice is that even if it rains for only one minute, WU continues to show a non-zero rainfall rate for every observation in the entire next hour! Ridiculous! This deception would be eliminated by my suggested way to calculate rainfall rate based upon the rain accumulated since the prior observation only.
I hope the above explanation helps to clarify my suggestion. If there are any more questions, please let me know.
-
JEBTucson
commented
How would you like to see WU report rainrate?
-
SpiderVV
commented
This actually depends on the WX station software too, I've seen stations with WeatherLink that output the real rainrate to WU temporarily and plot it on the graph correctly. It might have to do with settings, but yes, most WX station programs output it like that to WU.
