. Collected reprints / Atlantic Oceanographic and Meteorological Laboratories [and] Pacific Oceanographic Laboratories. Oceanography MEASUREMENT DEPARTURE 4 ACCURACY REQUIREMENT - ISMG. \ ' \ â â¢â â 'â r\V\.\ Fig. 7. Departure of pyranoiuetci seusoi clue to random errors in measurement. Time indicates tin* period \ci which the data arc integrated. lie 8. Departure of pyranomctei sen so l due to random errors of (1) spatial sampling and (2) data integration. Time indicates the period over which the data arc averaged. 15) using the data from the Krenkel on which three pyranometers
. Collected reprints / Atlantic Oceanographic and Meteorological Laboratories [and] Pacific Oceanographic Laboratories. Oceanography MEASUREMENT DEPARTURE 4 ACCURACY REQUIREMENT - ISMG. \ ' \ â â¢â â 'â r\V\.\ Fig. 7. Departure of pyranoiuetci seusoi clue to random errors in measurement. Time indicates tin* period \ci which the data arc integrated. lie 8. Departure of pyranomctei sen so l due to random errors of (1) spatial sampling and (2) data integration. Time indicates the period over which the data arc averaged. 15) using the data from the Krenkel on which three pyranometers were located, we have evaluated the crroi due to visual integration alone. In this way it was possi- ble to separate the total random erroi (fig- 7) into the two components as shown ill Fig. 8. and to examine how they varied as a function ol integration time. The information in Fig. 8 is useful in illustrating the relationship between the accuracy required for GATE measurements (5' <,) and the random errors ol spatial sampling and data integration; it also shows how this relationship depends on the period of integration. For example, if the length of the intercomparison were only one hour, it is evident from Fig. 8 that the de- parture ol a single pyranometer from the average of all pyranometers is likely to be near 6% due to the random error sources. This is larger than the accuracy require- ment specified by ISMG and, of course, would not pro- vide an adecpiate basis lor standardizing pyranometers. Clearly, it is most desirable to use a long integration period to minimize the random pan ol the measurement differences. The present ISMG plan suggests that three-day inter- comparisons will be conducted at sea during the main held phases with approximately the same ship spacing as in GIST. 1 he estimates in Fig. 8 suggest that if the pyranometer data arc integrated for a three-day period, the uncertainty in individual sensor measurement due to random sources will he abo
Size: 2232px × 1119px
Photo credit: © The Book Worm / Alamy / Afripics
License: Licensed
Model Released: No
Keywords: ., bookcentury1900, bookcollectionamericana, booksponsorlyrasismemb
