Laserfiche WebLink
<br />precipitation totals for 3- to II-hour storms accurate to <br />about the nearest 0.02 inch for an 8-inch orifice. Larger <br />orifice gages, as often used to measure snowfall, would yield <br />even better accuracy wi th the same gage mechanism. <br /> <br />\ <br /> <br />The storm total accuracy could be improved, in any <br />event, if the final chart reading wel:e made as far after <br />the end of the storm as possible. ~lble 3 shows that <br />the average undermeasurement was approximately halved <br />when the final measurement was made 9 to 16 hours <br />following the end of the precipitation. It would not <br />always be possible to do this in an actual field situation <br />where frequent storm events occurred. An additional problem <br />would be that the lag would also tend to obscure the <br />actual ending time of precipitation. <br /> <br />The resul ts are even better whEm the gage is subj ected <br />to an outside mechanical disturbance before the storm totals <br />are read. The average error was reduced to almost zero when <br />the chart readings were made after the platform was tapped <br />with a pencil. The small tap with the pencil did not <br />significantly displace the gage mechanism and may replicate <br />rather well a bit of wind disturbancE~ on the gage. 'l'h1.S is <br />not to suggest that gages be preferentially operated in <br />areas subject to wind action, as the gage inaccuracies due to <br />wind effects are well documented and far greater than those <br />discussed here. <br /> <br />4. Short term accuracy. <br /> <br />It may be desirable to analyze the results of cloud <br />seeding efforts on a time scale less than that of the total <br />storm duration. Hourly precipitation data, for example, <br />could be beneficial if sufficiently accurate. The gage <br />accuracies for precipi ta tion events of between 1- and 3-hour <br />dura tion were estimated from the cont~rol flow rate da ta. <br />As described previously, the experimE~ntal design <br />incorporated a second set of beakers which could be swung <br />into place to intercept the test flow to the gage. These <br />control flows (usually of one-half-hour duration) were used <br />as a check to insure that the simulated precipitation rate <br />being added to the gage remained appr'oximately constant with <br />time. Each day's test was started wi.th an intercepted <br />n control flow II meas ur ement. At the conclusion of this <br />period, the control beaker was removE!d, allowing the flow to <br />enter the test beaker wi thin the gage- bucket. The test 1:10w <br />was maintained for a per iod ranging f rom approximately 1- to <br />3-hours, after which another half-hour control flow sample <br />was normally taken. The test would continue in this <br />alternating manner for the remainder of the day with all test <br />flow precipitation accumulating in the gage. <br /> <br />11 <br />