This set of web pages was triggered by my participation in a field test for the Climalytic TROPO rain gauge. As I had done a detailed comparison of the classic "Stratus" gauge from Productive Alternatives and a new gauge from Outback Blue I was interested in joining. Also, it appeared to me that the folks behind Tropo weren't as harsh on themselves as they could be, so I expected to find some interesting issues to explore.
Some of my pages here were inspired by the testing outline from Climalytics but go beyond the request. Some are from interesting issues raised when comparing my collection of gauges. Some are "I can do better than that," generally in terms of having a more precise weigh scale or doing extra tests.
The entries in the navigation bar at the top include:
Home
This page.
Outback Blue
My comparison between the CoCoRaHS approved Stratus gauge and the "upstart" Outback Blue
gauge. The latter design has been changed quite a bit.
Tenths
This tests the weight of water in the inner tube at each 0.1" of rainfall.
Inches
This tests how well several inches of rainfall can be measured by using the inner tube as an intermediate measure.
Drying
This examines issues about how much water is left on a tube after drying it well.
Funnels
The gauge expects the funnel inner diameter to be 4.00". They don't seem to be!
Expansion
This looks at the "Coefficient of Linear Thermal Expansion" of polycarbonate and what it
means for rain gauges.
Air density
My rain gauges weigh more when they're cold!
The Tropo gauge, and most others on the consumer market, coolect rain in a 4" diameter funnel that fils a narrower tube. Typically, 1" of rain will show up as 10" in the narrow tube. Tropo has a 12" scale and its polycarbonate plastic has a smaller meniscus than Stratus's polybutyrate. Together, it's significantly easier to read amounts on the Tropo scale.
There's a better way. 1" of rain through the funnel will weigh close to 206 grams, so 0.01" of rain weighs about 2g, and there are inexpensive weigh scales that read to 0.1g or even 0.01g. They should be used inside as leveling and still air are both important to get a good reading. So visual scales are still worthwhile, and weighing is very good for checking their accuracy.
Late in 2018 I bought a cheap, err, inexpensive 0.01 gram resolution scale with a max loading 500g to experiment with weighing snow. I went looking for a 0.1g scale, but at $11.99, how could I refuse? I soon spent $16.85 on calibration weights to see if it really was accurate. It is, but it has a few quirks that I can work around.
I later bought a 0.1g kitchen scale with a 5 kg max load and a 500g calibration weight ($27, $10). I think scale is manufactured by the same OEM and has its own quirks.
Above are the scales and weights. Weighing the same thing twice at the same time is a good check that the scales are accurately zeroed. The 100g weight normally stays in the film can as one of the 0.01g scale's quirks is that weights ending in .98, .99, .01, and .02 are rounded to the whole gram. The film can both protects the weight and adds an offset to avoid that quirk.
The 0.01g scale has a max load of 500g, though I can push that a little beyond. The Tropo outer tube weighs about 503g, so I have to use the 0.1g scale when weighing rain in it.
Even with the Stratus gauge, I've seen people claim they can read the scale more accurately than 0.01". The tropo gauge will tempt many people to do so. Those of us with kitchen scales with 0.1g resolution can do so more readily and more accurately. However there are several reasons to stick with 0.01" (or 0.2mm) resolution.
As should be clear as you read the rest of the pages here, there a lot of sources of error:
Rain that stays in the funnel.
When rain begins, raindrops first wet the funnel's surface. These can be either discrete droplets or a smooth surface wetmness. When rain ends, what's still on the funnel evaporates readily if dry air moves in. Or, as nighttime falls, that can bring dew which we do not want to measure. The water left on the funnel can be about 0.01" of rain.
Inch plus rainfalls
To read big rain events, we use the inner tube to measure the total rain fall in pieces up to one inch at a time. Doing so misses some of the water that sticks to the outer tube, and see higher amounts from the inner tube from it not completely draining.
4 inches?
All the rain gauges described here are four inch gauges. My measurements, rounded to the neareast inch, matches. However, looking at a one tenth inch resolution, my oldest Stratus gauge has a diameter of 3.9". All the rest are 4.0" gauges. Looking at a one hundreth inch resolution, the Tropo gauge is the closest, but only 3.98". (At this point I have some concerns about my measuring skills!) A 1% error in diameter becomes a 2% error in area.
Scale and meniscus
Various problems can lead to the scale on the inner tube being not quite right. (On Tropo it appears to be nearly perfect.) I suspect that many people don't read the meniscus well, especially on the Stratus gauge. Even if you do read it well, that meniscus is an error source, and a 10X greater source when trying to read to 0.001".
Thermal expansion
Evaporation
If we set aside the errors, what do we gain with 0.001" resolution? A better question is what do we learn from measuring precipitation? Those include information about plant growth, droughts, and floods. More important that 0.001" resolution are the effects of sun, clouds, temperature, humidity, wind, and ground cover. There are significant challenges with measuring each of them. So while, we can measure rainfall and get a number in the thousandths, it simply doesn't provide a useful amount of extra informatiom.
Written 2023 Jan 28, text last updated 2023 Mar 6.