Checking rain gauge readings

I wonder how many members of this forum actually check their rain gauge readings with some sort of manual gauge (or even a second automatic gauge). I have found by experience that this is essential as things do occasionally go wrong with automatic gauges. Data from my station goes out to local farmers as well as a UK group called the Climatological Observer’s Link. If my readings are in error, some local farmers who also take readings have been known to point out if things are wildly wrong.

The problem with doing this is that I have WD set to change the day at midnight, which I think is the logical way to operate a station whose output is viewed by many members of the public. But if a check gauge is used, it is logical that it should be read at midnight which is not a task that I am prepared to take on! I normally take the reading at 9am GMT throughout the year, regardless of BST or British Summer Time. The trouble with this, is that the two readings are not directly comparable. So any corrections required may well be complicated. I generally assume that my check gauge is the more accurate as it is a British Met Office standard 5 inch diameter copper gauge completer with a special measuring cylinder.

A specific problem with this way of measuring rainfall is that the rainfall of the 1st of the month does not start until 9am on that day, in other words with this arrangement the previous month’s rain lasts until 9am on the first of the next month. If there is no rain between midnight and 9am on the first of the month then there is no problem. If there is rainfall during this period then in comparing amounts, during the month this will have to be allowed for.

If you look at the main WD page at 9am (a civilised time to take readings!), you can see that the current month’s total rainfall is shown and this includes the amount on any one day from midnight to 9am. So if you allow for the above problem at the start of the month, then the total of readings from the check gauge (which I have on a monthly check sheet) should be the same. For example if there had been 2.5mms of rain during the midnight to 9am period on the first of the month, then the automatic gauge should read 2.5mms greater than the check gauge total. If not then something is likely to be wrong with the automatic gauge readings and if the same check was performed yesterday at 9am gave correct readings, then you can assume that either the check gauge is at fault, or, more likely the auto gauge.

On my station I use an Instromet gauge as my auto gauge and it does from time to time get blocked by birds (there are spikes to deter them) or more likely it has frozen up (I do have a heater but sometimes forget to put it on). The main check gauge occasionally fills up with snow and I take it into the house and thaw it but if there is a lot of snow, then it’s very difficult to have accurate readings with any sort of rain gauge.

I would be interested to know if others have similar problems and any other simple way of solving it.

John

I have three weather stations of different makes and each gives quite different amounts of rainfall. For example, yesterday, they gave respectively 2.0, 4.1 and 6.1 mm. The first two are the seesaw type, the first one being quite old and the buckets are corroded (new ones on order). The second bucket type is brand-new. The third rain gauge is acoustic and is also new. As numbers two and three are relatively new acquisitions, I’ve been thinking about checking the calibration of all three, without the means to do so super accurately. The way I was going to do it was by placing a 15 cm funnel in a tall glass, half buried in the ground. The collected water would then be weighed in a scale to

The daily rain totals for my Davis track very well +/- with the official airport station a few miles away, that’s good enough for me :slight_smile: Everything I’ve seen on this forum over the years has persuaded me that trying to match the totals for co-located gauges of different sizes and technology leads to madness :lol:

I tend to agree with you, Niko. My motivation to do a check, however crude, is to find out which of the three rain gauges approximates best. If the Davis were new, I would not have any qualms about whether it was accurate enough for my purposes but the buckets are all corroded and it is more than 10 years old. The Oregon is new, with nice shiny buckets (for the moment) but I am less confident in the Oregon weather station. The WeatherFlow acoustic rain gauge does not inspire confidence, for the moment, as it has not yet been tweaked by the manufacturer. I’m also worried that it may give very different results according to the raindrop size and wind speed but this remains to be determined.

My problem is that the two nearest official weather stations are both about 20 km away. One is east of here on the coast with a range of mountains between them and us. The other is plonk in the middle of the island, north of here, with another range of mountains between them and us. In both cases, the vegetation around the official sites is very different from the mountainous pine forests around us, 300 m up. It is impossible to correlate the rainfall from either official site with our own, or for that matter, temperature, humidity and other measurements.

It is for this reason that I have 2 1/2 weather stations, so that I have a reasonable chance that the weather I measure here is reasonably accurate (farmers round here use my forecasts in preference to the official ones). Rainfall is the only parameter where I have some doubt as to the accuracy, and that is why I want to get a better idea of it, particularly when irrigation is required or otherwise.

I use a CoCoRaHS-type gauge as a cross-check on my modified tipping-bucket gauge, and I usually manage to read it at midnight. :slight_smile:

But any attempt to correct the tipping-bucket readings would be difficult, beyond the obvious observations that CoCo always records more rain than the tipping-bucket gauge and often records a trace when the WMR200 (now 0.25 mm per tip) shows none. The gauges are about 20 feet apart in a suburban garden, exposed as well as houses, trees and shrubs will allow, but any swirling wind must cause problems.

Every month I calculate a “correction factor”, from the ratio of one gauge’s readings to the other, and so far this year it has varied from 2% to 32%, averaging 11%. . .

Sorry, couldn’t resist…

https://youtu.be/y8hcfTFVJ9k

Ha, ha :lol: Full source https://www.youtube.com/watch?v=mEOjo0imqKE (Note: Includes one F word)

Last night, we had the mother and father of a storm and the Davis registered 16.8 mm and the Oregon 23.8 mm. Visually, I could see that when the rain was very heavy, the step on the graph was much lower on the Davis. On the other hand, when the rain was moderate, both systems tracked correctly.

As I was writing this, we had a short cloudburst with 3.4 mm on the Davis and 4.9 mm on the Oregon with respectively maximum rain rates of 1.8 mm/min and 6.0 mm/min, both at 14:51. A priori, it would seem that the Davis has a sluggish rain gauge when the rain rate is very high but probably works normally at low rain rates. Could this be due to the corrosion on the working surface of the seesaw?

I won’t mention what the WeatherFlow system has done during the stormy period other than to say that it went completely haywire for both rainfall and wind.

Actually me too my console reads actual rain is 2.0 mm but in WD it reads 9 mm, how to fix that?

I have three weather stations made up from various types of pro sensors. They all compare really well with the manual gauge, although for one, values below 0.5 mm do not show because of the bucket size. Usually have 0.1 - 0.2 mm error if any.