ea Miahse EAN ARF ks hee ndings kon ak Bia a eleca: } factor for our b apLE 5. L ; Pperienced at Eniwetom These Notes apply also to Table 20 <U. Date and Time refer to the 180th meridian. = = & (assuming normal ™ a lstently on the Level is correct within 5 mbe, except for the more accurate surface value, which is taken Height values are correct within 20 m. for levels between 850 and 600 mb. (inclusive); zz = within 30 m. between 500 and 300; within 50 m at 200; 2asis of real or - yhat can equally wel] | F level (see above). TT is correct within 1* C. up to 300 mb. and within 2° above 300 mb., assuming no gross Low cloud . i ' ted middle cloud | 2 given in hundreds 7 vement is to four instrument failure and no major error on the part of the observer . 5 i RH is correct within 10% and most values are correct within 5%, except when values are in parentheses, when RH may be in error by as much as 20%. DD is given to the nearest 10° and about 95% of the values shown give the true value to the nearest 10* interval. The remaining 5% are in error by a full 10° step. minute intervals. iE re correct within * error by 1° or moree ' afternoons with few i a: normal. while during the = least at greatest heights. 1 the values shown m4 +t to occur momen- | ; with the effect of : e). tions. : F 2The values by 10* intervals are based on more accurate readings half of which may be in The 5% figure is based on the assumption that the error distribution is FF values are correct within 10-15%, the accuracy being greatest at lowest heights and NOTE: The above estimates of the reliability of the various observations are based on considering both instrumental and observer errors, not including any consistent bias. Thus such factors were considered as accuracy of elevation and azimuth angles (instrumental) and the fact that in plotting there were inaccuracies introduced by the thickness of pencil lines. In any TABIE6. bility decidedly lerefor. and within 100 m. at 150 and 100 mb. These inaccuracies are in addition to those associated solely with estimating the pressure a high frequency of a stima it is almost certainly correct within 5 minutes. from the station barometer (see Notes for Table Ae ans in the tropical above. Other- , Served at one of the = seldom reported at Where the time given is precisely 0000 or 1200 much as 15 minutes. it represents the scheduled release time and may be in error by as ch it i largely depends YlF vise, . FRED: RAWINSONDE OBSERVATIONS. Pply also P exact tine of THREE~HOURLY OBSERVATIONS. These Notes a These recording gaze BRUCE: to Tables 8, 21, and 2h. Date and Time refer to 180th meridian, and times given are correct within 5 minutes, = ME j t TT, TT, lye TT, TxTxs and TT T,Tx, nin were all checked, one against the others, and minor adjustments were made in some instances in a ccorda d nce with the following rules. Direct reading dry-bu ~ lb and wot-bulb temperatures were taken as being correct except in two instances (for all Tables Al fae OnEms ™ ose em SS nee ceeee mg a a COMncecarne ee cgeeae mente i same A ee tegen Hen 1 om es