proudly sponsored by
|
archiving of SOWACS proudly sponsored by |
|
NOTE: To get off this list, send email to majordomo@aqua.ccwr.ac.za with the body of the message containing the line: unsubscribe sowacs Greg, The details you provide provoke more thought. I suspect that the displayed counts during the standard count are scaled to a 16 s reading time even though the counting time during the standard count is 8 s. Gauges with those firmware revisions probably have displayed counts scaled to 16 s. What do the manuals for those gauges say? Are they taking standard counts as 32 counts of 8 s each? The Chi ratios reported by your gauges average to 0.98 and 1.07 for gauges 706 and 3521, respectively. The respective calculated Chi ratios average 1.39 and 1.51. The correlation between the CPN Chi ratios and the calculated ones in your first message indicate a scaling problem because the calculated ones are uniformly larger. Also, a linear regression gives a slope of 0.82 and 0.70 for calculated Chi ratio vs. CPN Chi ratio (r^2 values of 0.99) for your gauges 706 and 3521, respectively. When I multiply my 16 s counts by two and recalculate the Chi ratio, I get a slope of 0.71 for Chi ratio for data multiplied by two vs. Chi ratio from original data. The slope is 0.5 if I multiply the data by four. I suggest that you divide the counts you recorded by two and recalculate the means and Chi ratios. I think you will end up with values very close to what the gauges are reporting. We are about to do our yearly cross-calibration here and will check it out with our five CPN and one Troxler gauges. I'll post the details here. Regards, Steve At 09:03 AM 4/24/2000 +0000, you wrote: > >Steve, > >Thanks for your comments. I have checked the pertinent information you >indicated and found the firmware revisions were 503A-230 and 503B-3.0. The >number of counts used to calculate the Chi ratio were the 32 counts >displayed as >the instrument standard count was taken. The number of seconds for each count >was determined by the firmware (16 sec as you indicated). The counting time >was >32 sec for unit 706 and 16 sec for unit 3521. The standard count determined by >the instrument was the same as the manually calculated average of the 32 >readings recorded in all instances. A recent standard count taken with unit >706 >set to 16 sec count time produced an instrument Chi ratio of 1.12 and a >calculated Chi ratio of 1.59 which is consistent with previous results. > >My concern is that all the calculated Chi ratios are outside the recommended >operating range for the instrument (0.75 - 1.25) which I suspect affects the >repeatability of the readings. > >Gregg > >__________________________________________ >> Gregg Dill, P.Eng., Irrigation Specialist >> Alberta Agriculture, Food and Rural Development >> Agriculture Center >> 100, 5401 - 1st Avenue South >> Lethbridge, Alberta, Canada T1J 4V6 >> Telephone +1 403-381-5136 FAX +1 403-382-4526 >>______________________________________________ ___________________ Steve Evett, Soil Scientist USDA-ARS, 2300 Experiment Staton Rd., Bushland, TX 79012 USA 806-356-5775, FAX: 806-356-5750 srevett@ag.gov, http://www.cprl.ars.usda.gov/programs/