WMOD00583

• Contract/Permit #: 1-07-81-V0175
• Title: Some Laboratory Tests of the Accuracy of the Belfort Universal Recording Raingage - Technical Report
• Date: 3/1/1985
• Weather Modification - Doc Type: Report

Test results showed definite fluctuations in both gages, with
maximum departures from the starting point of 0.010 and 0.015
inch respectively. These are illustrated in Figures 17 and
18 in Appendix A. The departures were in opposite directions
for the two gages; one gage showed an increase from the
starting point at about the halfway point in the test. After
that time, the traces gradually converged towards the
starting point. Both units were within 0.005 inch of the
initial point at the end of the test period. The source of
these deviations is unknown but it is apparent that
precipitation totals from storm events of one-half day or
longer duration could be subject to errors of approximately
0.01 inch if these gages are typical.

i
I

(
"
I

3. Storm total accuracy.

Each of the tests in which simulated precipitation was
added to the beaker within the gage bucket allowed a precise
comparison between the "true" catch and that indicated on the
chart record. The collection beaker (with evaporation
suppressant) was weighed to the nearest 0.1 g before each
trial, and then re-weighed with the collected water after the
conclusion of each day's tests. Tests conducted before gage
trials disclosed that for the evaporation suppressant and
beakers used, evaporation losses were negligible over time
periods far exceeding those between weighings. Thus, any
discrepancies between the weighed accumulation in the beakers
and the chart readings are not due to evaporation.

In the initial tests the rise in the chart trace seemed
to lag behind that expected from the calculated input flow
rate. This suggested that perhaps hysteresis was occurring
in the gage mechanism. Consequently, with most later tests
(beginning on 18 November 1982) the chart cylinders were not
removed from the drives and the collection beakers were not
taken from the buckets and weighed until the morning after
the final test flow was ended. This allowed more time for
the hysteresis to correct itself. In addition, just before
the beaker and the chart were .removed, the gage bucket
platform was tapped with a pencil to remove any remaining
frictional drag in the mechanism.

The indicated chart levels were then determined for the
following times: (1) at the immediate end of simulated
precipitation input; (2) after several hours of zero
precipitation input; and (3) after the mechanism was tapped.
Results are presented in Table 3. It can be seen that, on
the average, the gages consistently undermeasured for all
three measurement times. The average undermeasurement at the
point simulated precipitation input was ended was 0.017
inch. The data suggest that the error increases with total
precipitation amount, but additional trials would be required
to confirm this. This level of accuracy is quite im~ressive,
as it implies that the basic gage is capable of prov1ding

10