Measuring rain the easy way: using an automatic logging rain gauge

I used to have one of those rain gages, but my uncle “borrowed” it and turned it into a smoker. You’d wind it up and tape a piece of paper on a drum that would take a week (I think) to make a whole revolution. The scale would depress an ink stylus which would make a continous trace of the weeks cumulative precipitation. I also have a recording barometer that works the same way. I hide it away out of site when my uncle visits.

Rainfall levels are probably more difficult to get accurate than surface temperature. One flash flood I have looked at, Boscastle, Cornwall, UK. there were two gauges in the catchment differing by 5ins over the flood period, these were 2 miles apart. So getting an area rainfall figure is fraught with problems.
Still looks a nice instrument.

My father has a similar device as part of a radio linked home weather station. He has it mounted next to a conventional plastic calibrated jug gauge, and the two have been remarkably comparable to each other in readings. Occasionally very light rain will cause discrepancies, and the spider issue mentioned above is certainly one we have encountered, along with small parts of leaves & debris which get past the screen. This may be more relevant in our typical domestic setting – at a more remote location it may not be a problem. In our example I would not expect such a device to work reliably for a year without some regular checks and cleaning.

I work for the NWS. The government already uses tipping bucket rain gauges at all “official” observation sites. They are not without problems. We spend time having to quality control their data.
1) There is a known and verified bias of underestimating rainfall when heavy rain is occurring…estimated to be as much as 5% depending on how hard it is raining. (Arguably, when rain fall reports are most significant or critical.)
2) Dew tips. These occur when enough dew forms to cause a tipping of the bucket. Dew by definition is not precipitation.
3) Wind tips. This occurs if the bucket is almost full, but hasn’t quite tipped over, but a gust of wind occurring well after the precipitation has ended finally pushes it over.
4) Bird poop. Loves to clog the rain gauge up and has to be manually cleaned out. A partial clog will cause a slow drip over hours if not days and can often go undetected if someone isn’t paying close attention to actual start/end times of precipitation.
5) And my personal favorite. The story of a field mouse nesting inside the gauge and running back and forth on the bucket causing sporadic reports of precipitation. That was a fun one to figure out.
And of course snow, sleet, and ice reports are terrible if not impossible.
When it comes to measuring precipitation, nothing beats a standard 8-inch diameter gauge with a dedicated human to measure it. There is a push of automating all cooperative observation stations with tipping bucket gauges. The data quality will plummet!

I also work for the NWS, at a River Forecast Center. Most river gage sites have tipping buckets connected to data collection platforms that telemeter via the GOES satellites hourly. Part of my job is to QC over a thousand tipping bucket gages hourly for input into radar precipitation estimation corrections. As mentioned earlier, clogs and partial clogs are a constant problem. Dew will form on the funnels first, and if next to a dusty environment such as a gravel road or a farm field, the dust interacts with the dew and slides down to the funnel neck, creating a very effective adobe plug. Spiders build nests inside, which interfere with the tipper movement. Blowing leaves collect to create partial plugs, so as mentioned above, a one hour rain storm may dribble out for hours or days after the event. Dirt collecting in the buckets can cause over-estimation. Mast-mounted gages can be knocked out of level; they must be exactly level to be accurate. During high rain rates, water is lost during the transition from the full tipper to the empty tipper. Long term climatologies show tipping buckets collect about 70% of what a co-located 8″ standard rain gage collects. Tipping buckets, whether heated or not, don’t work worth diddly in snow, which is a significant amount of yearly precipitation for the northern portions of the country, and which often causes significant and prolonged snow melt flooding. Knowing the amount of water accumulated from snow is critical to accurate flood forecasts. For that matter, unattended automated weighing gages at the Local Climate Data airport sites can under-report in extreme events such as a foot or more of wet heavy snow, when the funnel heater can’t keep up and the snow bridges the funnel and additional snow blows over the top. On the other hand, there are about 20 Fisher-Porter gages in the Middle and Upper Mississippi Valley which have telemetry, and have observers available to remove the funnel for snow events and install the funnel for rain events, and these Fisher-Porter gages have measured the heavy snow events quite nicely and are quite comparable to the 8″ standard rain gage reports. The Fisher-Porter network is in the process of having all the mechanical linkages, punch tape and springs replaced with a load cell and datalogger. Units that have been upgraded have shown a drastic decrease in maintenance and lost data.
One major deficiency in the COOP network precipitation gages is the lack of wind screens. Research has shown wind above 10 mph has a significant effect on the catch of unshielded gages, particularly during snow. Check out http://ams.confex.com/ams/90annual/techprogram/session_23952.htm Very few gages outside the ASOS network have wind screens.
If one could obtain quality all-weather precipitation data for $200/gage, I’m sure the NWS would jump up and down with joy, as budgets are always an issue. But, the unattended all-weather precipitation observation problem is more complicated than that.

the location of the gauge is or paramount importance.
not just to miss the rain shadow of a building… but the larger view.
i worked at a certain place in the los angeles area for about 30 years. during this time i drove a certain freeway about 15 miles each way 5 days a week.
over that period of time i observed that at one intersection between freeways it very rarely rained. at another about three miles away, if we would get any rain in the LA basin it would be at least foggy and probably be raining there.
there was also the appearance that in the los angeles/long beach harbor there would be rain in the small hours of the morning that would not be noticed by the rest of the town but we would see roads washed clean, parking lots with puddles and when asked about it the graveyard shift would say “yes it rained heavily for about a half hour about 0400 this morning”.
has this kind of thing been addressed?
C

Walter Dnes:
sir, i grew up in montana and every kid in the state knows how to “check the snow depth”. :-)))))
c