Author Topic: Rain gauge comparison and calibration controller  (Read 3511 times)

0 Members and 1 Guest are viewing this topic.

Offline SLOweather

  • Global Moderator
  • Forecaster
  • *****
  • Posts: 3456
    • Weatherelement Moline IL
Rain gauge comparison and calibration controller
« on: January 19, 2016, 01:42:27 PM »
I have several Davis rain gauges, a Rainwise, what appears to be a Texas Electronics, and a couple of Hydreon RG-11s, laying about.

So, I decided to do a project that's been on my mind for a while. I have resurrected an old Rugid Computers RUG9 controller from the Pile Of Cool Stuff, and started programming it to be a rain gauge comparator (for starters). So far it counts up on each of 5 inputs, and resets the counts to zero if I press the 1 key while on the rain gauge screen. Right now, I'm figuring out the programming to calculate rain rate.




Offline SLOweather

  • Global Moderator
  • Forecaster
  • *****
  • Posts: 3456
    • Weatherelement Moline IL
Re: Rain gauge comparison and calibration controller
« Reply #1 on: January 19, 2016, 10:02:13 PM »
OK. I got some good work done earlier, while it was raining. (1.77" today! Woo hoo! The creek in the canyon below our property is running for the first time in years!)

It took a programming trick, but I now have a basic rain rate (inches per hour) calculation running. I had to add a software delay timer to the digital input to debounce the tipper switches so I could latch the number of seconds since the last tip.)

Offline SLOweather

  • Global Moderator
  • Forecaster
  • *****
  • Posts: 3456
    • Weatherelement Moline IL
Re: Rain gauge comparison and calibration controller
« Reply #2 on: January 22, 2016, 01:38:00 AM »
Today I did a couple of test runs with my Novalynx calibrator, one on the old Rainwise RGP, one on an older Davis. I wasn't really looking for any accuracy results. Just a test of the system.

I did notice one thing interesting. The calculated rain rate in inches per hour varied more than I expected during a run. Some was apparently due to the decreasing head in the calibrator bottle. Some I'm not sure of.  I need to program up a logger so I can look at the run data in a spreadsheet.

Also, I'm considering calibrating a peristaltic chemical feed pump to feed a known flow rate to the gauges.

Offline miraculon

  • Sunrise Side Weather
  • Forecaster
  • *****
  • Posts: 4107
  • KE8DAF
    • Sunrise Side Weather in Rogers City MI USA
Re: Rain gauge comparison and calibration controller
« Reply #3 on: January 22, 2016, 08:49:26 AM »
I am curious about which orifice you used in the NovaLynx calibrator. My first attempt with my homebrew calibrator used a 1/16" hole drilled into a brass plug. This was way too big resulting in rates in the 10-12"/hr range. I found a #72 gas orifice that I used and got much better results. I have to watch that this small orifice doesn't start clogging with minerals from the water though. I might use distilled water next time.

I also noticed a slower tip rate near the end of my calibration run when the funnel was almost empty.

This was my first attempt with the 1/16" orifice: http://www.wxforum.net/index.php?topic=27402.msg264111#msg264111

Here is the-redo with the smaller orifice: http://www.wxforum.net/index.php?topic=27839.msg268957#msg268957

On the 2nd attempt, I used the Escali scale that I bought for CoCoRaHS instead of eyeballing the beaker markings. That way I was able to get the exact amount for the calibration runs.

Greg H.


Blitzortung Stations #706 and #1682
CoCoRaHS: MI-PI-1
CWOP: CW4114 and KE8DAF-13
WU: KMIROGER7
Amateur Radio Callsign: KE8DAF

Offline SLOweather

  • Global Moderator
  • Forecaster
  • *****
  • Posts: 3456
    • Weatherelement Moline IL
Re: Rain gauge comparison and calibration controller
« Reply #4 on: January 22, 2016, 10:49:33 AM »
I've been using the 1/32" orifice exclusively. It appears to be a plastic hose barb. The Novalynx calibrator has a sort of "constant head" design, rather like a hummingbird feeder or chicken waterer.

In a Davis gauge, it gives rates of about 2" per hour. which is higher than I would like. My calibrator routine in the Rugid counts seconds between tips and calculates the rate each tip.

Offline SLOweather

  • Global Moderator
  • Forecaster
  • *****
  • Posts: 3456
    • Weatherelement Moline IL
Re: Rain gauge comparison and calibration controller
« Reply #5 on: January 22, 2016, 04:18:00 PM »
I clamped a Stenner Classic Series adjustable peristaltic pump to a table and installed a Koflo 100 ml calibration column on the intake port. The Stenner has a #1 size pump tube in it

At the highest setting, it's supposed to pump 9.1 liters per day.

The initial 100 ml test run took 10.50 minutes to pump 100 ml. I calculate that to be a rain rate of 1.05" per hour into a Davis gauge. The drops come off the end of the discharge tubing at about 2 per second.

This might work...

Offline SLOweather

  • Global Moderator
  • Forecaster
  • *****
  • Posts: 3456
    • Weatherelement Moline IL
Re: Rain gauge comparison and calibration controller
« Reply #6 on: January 22, 2016, 04:25:26 PM »
One other result from today...

After programming up a data logger module, and doing a test run on the Rainwise 8" gauge with the Novalynx calibrator, and then looking at the data in Excel...

Most tips took 22 seconds, with a few outliers of 21 and 23 seconds. Then at the end, the last few, 22, 22, 23, 26, 29, 33 seconds.

Offline ValentineWeather

  • Forecaster
  • *****
  • Posts: 6364
    • Valentine Nebraska's Real-Time Weather
Re: Rain gauge comparison and calibration controller
« Reply #7 on: January 22, 2016, 06:50:55 PM »
The 1/32" nozzle will give a rain rate of 2" per hour on an 8" gauge. The spoon tippers on a Davis are smaller being its a 6.5" diameter gauge so the rain/rate is going to be higher also.

Novalynx factory calibrates their gauges at 6" per hour which is above the rated 1% error of 2 inches per hour, for overall good performance I guess. So they use the 1/16 nozzle on the 8" diameter gauge at factory.
Randy

Offline CW2274

  • Forecaster
  • *****
  • Posts: 6731
    • Conditions @ CW2274 West Tucson-Painted Hills Ranch
Re: Rain gauge comparison and calibration controller
« Reply #8 on: January 22, 2016, 07:12:10 PM »
I got my Rainwise and e-mailed to ask what exactly equates to an inch of rain and the reply was 800ml so that's what I've calibrated to at about 2 to 3 inches per hour.

Offline ValentineWeather

  • Forecaster
  • *****
  • Posts: 6364
    • Valentine Nebraska's Real-Time Weather
Re: Rain gauge comparison and calibration controller
« Reply #9 on: January 22, 2016, 07:21:25 PM »
I got my Rainwise and e-mailed to ask what exactly equates to an inch of rain and the reply was 800ml so that's what I've calibrated to at about 2 to 3 inches per hour.

I think you will like it. Only reason I went with a Texas Electronics was the snow melt ability and less evaporation hopfully. Plus I got sick of scraping snow off the sides where it wouldn't melt. The TE is all aluminium so heat conduction will prevent this snow buildup on the edges they guaranteed me.   :-P
Randy

Offline CW2274

  • Forecaster
  • *****
  • Posts: 6731
    • Conditions @ CW2274 West Tucson-Painted Hills Ranch
Re: Rain gauge comparison and calibration controller
« Reply #10 on: January 22, 2016, 07:26:50 PM »
I got my Rainwise and e-mailed to ask what exactly equates to an inch of rain and the reply was 800ml so that's what I've calibrated to at about 2 to 3 inches per hour.

I think you will like it. Only reason I went with a Texas Electronics was the snow melt ability and less evaporation hopfully. Plus I got sick of scraping snow off the sides where it wouldn't melt. The TE is all aluminium so heat conduction will prevent this snow buildup on the edges they guaranteed me.   :-P
Me too. I've had the damn thing for three weeks but can't drum up any rain (now :roll:) in this El Nino to compare with my VP2!

Offline ValentineWeather

  • Forecaster
  • *****
  • Posts: 6364
    • Valentine Nebraska's Real-Time Weather
Re: Rain gauge comparison and calibration controller
« Reply #11 on: January 23, 2016, 10:55:59 AM »
Something for those doing their own gauge calibrations, if you want a more precise measurement of water maximum accuracy is obtained by using a laboratory scale, weight of water (1 ml = 1 gm).
Randy

Offline SLOweather

  • Global Moderator
  • Forecaster
  • *****
  • Posts: 3456
    • Weatherelement Moline IL
Re: Rain gauge comparison and calibration controller
« Reply #12 on: January 25, 2016, 04:51:42 PM »
I think my variable rate peristaltic pump will be a good calibrator. I can even program the controller to run it for a set length of time to match the volume desired. Of course, the problem is that this is a spare unit from our water company and runs on 240 VAC. The only 240 V plug I have is the garage.

Offline SLOweather

  • Global Moderator
  • Forecaster
  • *****
  • Posts: 3456
    • Weatherelement Moline IL
Re: Rain gauge comparison and calibration controller
« Reply #13 on: January 25, 2016, 04:56:56 PM »
In setting up this calibration mechanism, one thing I did was dowload whatever docs I could for the gauges I have.

Texas Electronics has about the best explanation of field calibrating and rates:


Quote
Absolutely accurate calibration can be obtained only with laboratory equipment, but an approximate field check can easily be made.  The tipping bucket mechanism is a simple and highly reliable device.  The transmitter must be located in a clear area, away from trees, buildings, etc. To obtain accurate readings, the unit must be mounted level and be free of foreign material, dust, or other debris. The transmitter must be calibrated with the rate of flow of water through the tipping bucket mechanism under control. 

At least 36 seconds should be allowed to fill one side of the tipping bucket.  This represents a maximum flow rate of one inch of rain per hour.  If the flow exceeds that rate, then the instrument will read low even if properly calibrated. Decreasing the rate of flow, on the other hand, will not materially affect calibration.  The reason for this is obvious if the tipping bucket assembly is observed when the weight of the water begins to tip the bucket.

Some time is required for the bucket to tip (a few milliseconds).  During a portion of this time, water flows into the empty bucket. This creates a small margin of error and the faster the flow rate, the greater the error. At flow rates of one inch per hour or less, water drips slowly into the bucket instead of flows, allowing the bucket to tip between drips, and eliminate the cause of the error.

Offline CW2274

  • Forecaster
  • *****
  • Posts: 6731
    • Conditions @ CW2274 West Tucson-Painted Hills Ranch
Re: Rain gauge comparison and calibration controller
« Reply #14 on: January 25, 2016, 04:57:54 PM »
I think my variable rate peristaltic pump will be a good calibrator. I can even program the controller to run it for a set length of time to match the volume desired. Of course, the problem is that this is a spare unit from our water company and runs on 240 VAC. The only 240 V plug I have is the garage.
Transformer?

Offline SLOweather

  • Global Moderator
  • Forecaster
  • *****
  • Posts: 3456
    • Weatherelement Moline IL
Re: Rain gauge comparison and calibration controller
« Reply #15 on: January 25, 2016, 05:12:58 PM »
Caveat: Please don't use any of these results as gospel or for purchasing decisions. All of these are OLD. The Rainwise is at least 20 years old, the Davis is a free-standing one from a WMII with a black, (not silvered), tipper, and the TE one I got used in new condition in a goody box I bought from WxTEch. I had to make feet for it.

This is currently to see how different rain gauges act, and how well this calibration lash up functions.

After one run on the 3 current gauges...

The 8" Rainwise is pretty solid at 21-22 seconds between tips, increasing the last few tips as the water ran out.

The 6.5" Davis was ok, with tips in the middle if the run ranging from 15 to 19 seconds, with most being 16-17 seconds. As usual, the tips increased 21, 25, 28 at the ends of the run. I got a little thrown because beyond that, there are some 4661, 13162, 9772, 893, 620, 269, 120, etc tips in the log. Then I remembered that it rained after the calibration run and before I downloaded the log.

The 6" Texas Electronics gauge is a little more erratic.  Mostly in the 15-17 second range, some 14s and then a few weird outlliers in the 29. 30, 32, 33 range. I think I'll check the tipper and then rerun it.

Another thing that bears inspection is to see if there is a pattern to the low and high times, like one side is not calibrated the same as the other.

Offline SLOweather

  • Global Moderator
  • Forecaster
  • *****
  • Posts: 3456
    • Weatherelement Moline IL
Re: Rain gauge comparison and calibration controller
« Reply #16 on: January 26, 2016, 02:29:33 PM »
One other thing Texas Electronics mentions in their manual, that I haven't seen anywhere else:

Quote
After transmitter installation, remove the gold funnel and observe the black tipping bucket. It should not be held
in a dead center position by the magnetic attraction of the bucket magnet and the magnetic switch. Press either
end of the bucket down against the stop to be sure it is not centered.

I have seen that happen in Dave gauges, especially ones with worn bearings.

Offline kcidwx

  • Forecaster
  • *****
  • Posts: 330
Re: Rain gauge comparison and calibration controller
« Reply #17 on: February 05, 2016, 11:39:00 AM »
In setting up this calibration mechanism, one thing I did was dowload whatever docs I could for the gauges I have.

Texas Electronics has about the best explanation of field calibrating and rates:


Quote
Absolutely accurate calibration can be obtained only with laboratory equipment, but an approximate field check can easily be made.  The tipping bucket mechanism is a simple and highly reliable device.  The transmitter must be located in a clear area, away from trees, buildings, etc. To obtain accurate readings, the unit must be mounted level and be free of foreign material, dust, or other debris. The transmitter must be calibrated with the rate of flow of water through the tipping bucket mechanism under control. 

At least 36 seconds should be allowed to fill one side of the tipping bucket.  This represents a maximum flow rate of one inch of rain per hour.  If the flow exceeds that rate, then the instrument will read low even if properly calibrated. Decreasing the rate of flow, on the other hand, will not materially affect calibration.  The reason for this is obvious if the tipping bucket assembly is observed when the weight of the water begins to tip the bucket.

Some time is required for the bucket to tip (a few milliseconds).  During a portion of this time, water flows into the empty bucket. This creates a small margin of error and the faster the flow rate, the greater the error. At flow rates of one inch per hour or less, water drips slowly into the bucket instead of flows, allowing the bucket to tip between drips, and eliminate the cause of the error.

I think what he meant to say is the full bucket. But yes that's what causes the under reporting.
« Last Edit: February 05, 2016, 05:20:01 PM by kcidwx »
Meteorological Technician
NWS Certified Aviation Weather Observer
B.S. Meteorology

Offline ValentineWeather

  • Forecaster
  • *****
  • Posts: 6364
    • Valentine Nebraska's Real-Time Weather
Re: Rain gauge comparison and calibration controller
« Reply #18 on: March 05, 2016, 12:22:05 AM »
Somehow I didn't see these last post or get notifications possibly. 
I ended up calibrating the TE at the 2" per/hr rate. I looked into the siphon RG type and found the non siphon are actually more accurate below 2" per/hr so glad I went with the non siphon gauge. At 6 inch per/hr rate I'm losing about .05/.04 or 5% for 1" rain.  Rates that high and above are rare in my area for extended time and only occur for brief periods during T/S.

I found this link showing rates and likelihood of occurrence.   http://hdsc.nws.noaa.gov/hdsc/pfds/
Randy