Summary of calibration parameters for the sonic anemometers This document is intended to emphasize the stability of various calibration parameters needed to process the sonic anemometer data. The UW sonics will be described. Whenever possible, the accuracy of the measurement will also be determined. Pathlength measurement: (cm) UW sonics: (should be 20) A B C (top) bc ac ab sonic 1: 8/85 20.0711 19.9796 19.8272 8.1661 8.0569 8.0264 8.0061 8.0645 7.9985 11/85 19.8992 19.7871 19.6107 8.1348 8.0320 8.0000 7.8909 7.9377 7.9730 9/86 19.9492 19.7917 19.7510 8.1178 8.0442 8.0264 8.0213 7.9832 8.0035 10/86 19.9695 19.8679 19.7485 8.1001 8.0061 8.0594 8.0315 8.0061 8.0416 3/90 19.8734 19.8534 19.6698 8.00357 7.97224 8.0423 8.00362 7.97198 8.04261 sonic 2: 8/85 19.8171 19.6977 19.7993 8.1305 8.3160 7.9248 7.9400 8.1813 8.0264 11/85 19.7844 19.5703 19.6276 8.0910 8.3437 7.9801 7.9420 8.2144 8.0116 9/86 19.8501 19.6418 19.7510 8.0645 8.3261 7.8816 7.9248 8.1991 8.0035 10/86 19.7866 19.6342 19.7053 8.0950 8.2804 7.9248 7.9375 8.2347 8.0670 3/90 19.7620 19.5937 19.6537 7.95100 8.18713 7.98573 7.94971 8.18799 7.98573 6/92 19.9873 19.8577 19.6317 7.9870 7.9413 8.0416 7.8923 8.2720 8.1175 The measurements were: 8/85: hand calipers measurement 11/85: two-dimensional sonic digitizer measurement (closes the worst) 9/86, 10/86: hand calipers measurement 3/90: measured using NCAR mill 6/92: hand calipers using "points" The worst-case change in measured pathlength is 0.2 cm, though this may have been due to a deliberate realignment of the array early on. The worst-case change in the last two measurements (separated by 3.5 years) is 0.08 cm. The bc,ac,ab values are distances between the top transducers. First line is measured distance, second line is predicted from other measurements. Note that worst-case closure in bc,ac,ab is 0.001 cm, from latest measurement, thus accurate to 0.001 cm. Previous measurements were much worse (typically 0.1 cm). Angles: (deg) UW sonics: (should be 60) Ah Av Bh Bv Ch Cv sonic 1: 8/85 60.43 62.53 63.59 63.49 62.32 11/85 62.3231 62.0940 60.2938 62.3737 61.7643 9/86 62.5810 60.0944 62.1173 63.0041 62.2810 10/86 62.6769 59.4125 60.8331 62.6503 62.0730 ?/87 62.18 59.94 60.92 62.72 61.87 3/90 0.264104 61.3324 61.8593 62.4181 61.6612 62.7507 sonic 2: 8/85 61.16 61.22 60.58 61.89 62.77 11/85 61.2248 60.5106 59.6408 61.9138 61.6429 9/86 61.6045 60.7287 58.4361 61.8930 61.7839 10/86 62.0517 59.4029 60.1865 61.7071 62.2408 ?/87 61.46 60.16 59.02 61.99 61.70 3/90 3.30618 60.6848 60.3923 62.6359 62.1622 61.8236 6/92 60.1124 65.8667 62.3198 57.7271 62.7474 The measurement on ?/87 was done using a coordinate measuring machine at RISOE. Note that I started using Ah (the horizontal orientation of path A with respect to a line in the center of the bottom triangle) recently. This was not used earlier since azimuthal orientation of the array (to this accuracy) was not considered important during my UW days. Differences of over 1.5 degrees are seen, even between the RISOE and NCAR measurements which closed very well. It must be possible to change the alignment of the array (in handling?) by this amount. Level output: (from tests using mill) FLAT: ATI #2: (unit 203) pitch = -0.906 sensor + 1.191 roll = -1.037 sensor - 0.386 ATI #3: (unit 204) pitch = -0.799 sensor + 1.371 roll = -0.888 sensor + 0.010 ARM/SJVAQS: UW #1: (unit 203) pitch = 0.733 sensor - 1.798 roll = 0.826 sensor + 0.960 UW #2: (unit 204) pitch = 0.778 sensor - 2.578 roll = 0.866 sensor + 1.176 Note gain in unit 204 seems RELATIVELY stable (to 3%), but that gain of 203 changed. The sign change must have been due to a 180 degree rotation of the sensor between the two mountings. Zeros: UW sonics: (counts, approximately = 0.017 cm/s) A B C S sonic 1: ?/85 +94 -283 +69 +1265 9/85 +199 +136 -39 9/85 +199 +160 -39 -96 9/85 +161 +115 -53 +700 9/86 +327 +72 -90 +1016 10/86 +249 +159 -18 +1600 10/86 +328 +126 -48 +1664 10/86 +130 +110 -10 +820 11/86 +140 +119 -47 +916 11/86 +121 +105 -50 +1228 7/90 -111 +12 +34 +956 9/12/90 -127 +8 +29 11/5/90 -110 -17 +13 2/13/91 -138 +11 +28 7/9/91 -103 -36 0 7/10/91 -121 7/12/91 -115 -89 -3 +197 7/23/91 -443 -84 +16 +1040 8/7/91 -462 -84 +8 6/8/92 -638 -19 -41 +260 7/1/92 -19 -46 sonic 2: ?/85 -7 -76 -69 +833 9/85 -108 +168 -80 +632 1/86 -10 +195 -49 +684 9/86 +20 +222 -45 +672 10/86 +22 +250 -48 +640 10/86 +45 +257 -61 +640 10/86 -40 +180 -10 +660 11/86 -28 +209 -60 +728 7/90 +180 -130 +177 +1241 9/12/90 +180 -130 +176 11/5/90 +207 -151 +160 2/13/91 +196 -164 +170 7/6/91 +180 -124 +155 8/7/91 +179 -135 +167 6/6/92 +175 -157 +166 +478 6/8/92 +32 -170 +179 +413 6/30/92 +27 -167 +176 Note that changes of +/- 20 counts are common (larger changes are probably due to threshold (70 counts/wave) or transducer changes).