8
$\begingroup$

My question is simple - how accurate is a tipping-bucket rain gauge?

I have read about what causes the errors (wind, bucket mechanics for extreme events, ...) and I know the accuracy is strictly related to the rain rate, in the sense that, e.g., TB gauges tend to highly underestimate in extreme conditions.

However, I'd like to have a number, say, errors around +-25%. I know the number itself won't be accurate, it's just to have the order of magnitude. References appreciated, since I myself looked around with no success :-)

Improve this question
$\endgroup$
2
  • 1
    $\begingroup$ You ask about accuracy but error statistics won't always include all sources of inaccuracy. You need to have the gauge located properly and we have learned the hard way that maintenance is a critical issue for remote locations because bird droppings and spider webs can foul the mechanism. $\endgroup$
    – haresfur
    Commented Sep 15, 2015 at 22:54
  • $\begingroup$ You are right. I have found good informations in this survey, section 3.3, where different sources are reviewed separately. Hard to give a single value, then. $\endgroup$
    – natario
    Commented Sep 16, 2015 at 9:31

2 Answers 2

Reset to default
7
$\begingroup$

The following simulation results shows that the Tipping Bucket estimates suffer from significant errors if based on time scales less than ten to fifteen minutes.

Comparison of Tipping-Bucket Error Statistics:

Comparison of Tipping-Bucket Error Statistics:

enter image description here

enter image description here

(Source of above diagrams)

In this study we attempted to characterize the errors associated with tipping-bucket rain gauges when used to provide rainfall intensity estimates at small temporal scales. The main source of the TB gauge sampling error is its sampling mechanism and its inability to capture the small temporal features of the rainfall time series. We found significant error levels in the 1-min estimates especially at low rain rates, but we also found that as the time scale of the computed rain rates increased, the error decreased substantially. With time scales longer than 15 min, the error becomes negligible. The gauge’s performance and its associated errors are sensitive to the applied sampling interval and the bucket volume. Therefore, we recommend sampling intervals on the order of 5–10 s, along with a bucket size no larger than 0.254 mm (0.01). Our analysis wasbased on rainfall observations that were mostly dominated by convective storms with few stratiform events. Similar analysis may be needed to investigate the scope of applicability of the developed formulas under different climatological regimes. However, the formulas serve as a first-order approximation that can be used in practical applications. (Source)

Improve this answer
$\endgroup$
3
  • 1
    $\begingroup$ So, the errors you describe are basically caused by uncertainty of the tipping point of the bucket. Sometimes they tip early, sometimes late. When you average over a lot of tips, though, the accuracy is good. $\endgroup$
    – Daniel Griscom
    Commented Sep 14, 2015 at 1:49
  • $\begingroup$ Thank you. What is the difference between time scale and sampling resolution ? I understand the former but not the other one (going to read the paper now) $\endgroup$
    – natario
    Commented Sep 14, 2015 at 8:43
  • $\begingroup$ Time Scale is the averaging period for reporting data. Sampling Resolution determines the number of samples used per reported data point. $\endgroup$
    – f.thorpe ♦
    Commented Sep 26, 2015 at 2:24
4
$\begingroup$

As the previous answer made clear, tipping bucket raingauges incur errors related to tip increments, rainfall amount, intensity and duration - all for obvious reasons. Many raingauges, TB and otherwise, have funnel filters to stop clogging from leaves, dead lizards, etc. In the case of small rainfall events the wetting and clogging of the filter can cause huge errors. It's important to note that these errors are separate from, and subordinate to, the errors of wetting, splash, and micro-aeodynamic turbulence. These further errors typically amount to an undercatch of about 7%, or very much greater errors in the case of strong horizontal components of wind velocity. Also, the collection area of the TB raingauge is significant. TB rim diameters range from 10 cm to 50 cm with the former being highly inaccurate. Whatever the design of raingauge / pluviometer, it is best to calibrate it with a ground-level rim set in an anti-splash grid.

Improve this answer
$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.