RELEASE AND RECORDER RECORD B4-17 
7.3. Personal Bias. The observer will use no personal bias while eval- 
uating the recorder record, super adiabatic lapse rates, ''supers'’, will 
not be hidden by "re-evaluating" the recorder record. Assuming the out- 
rigger was fully extended, the proper lengthoftrain was placed between 
the balloon and the radiosonde, and no obvious malfunctioning of the 
radiosonde and/or ground equipment is evident, the recorder record will 
be evaluated as accurate. 
7.4 Labeling ''Supers". All segments on the recorder record which are 
"super" will be labeled rechecked, with arrows pointing to the bounding 
levels which were rechecked. 
7.5 Doubtful and Missing Data. When the radiosonde or ground equip- 
ment functions abnormally, the accuracy of the raob data will be class- 
ified in accordance with the following paragraphs. A precise deter- 
mination of the accuracy of the data willoften be difficult or impossible. 
Classification, therefore, will be based upon the possible error char- 
acteristic of common situations. 
7.5.1 Temperature. 
a. When the possible error is 1°C or less, the data will be re- 
garded as accurate. 
b. When the possible error is more than 1°C but not more than 
3°C, the data will be regarded as doubtful. 
c. When the possible error is more than 3°C, the data will be re- 
garded as missing. 
7.5.2 Relative Humidity. 
a. When the possible error is 10 percentor less, the data will be 
regarded as accurate. (Note that the relative humidity may 
continue accurate within the +10 percent range even though 
the temperature in the stratum is classified as doubtful.) 
b. When the possible error is more than 10 percent, the data will 
be regarded as missing. Relative humidity data will not be 
classified as doubtful. 
c. Relative humidity data will be classified as missing whenever 
the radiosonde fails to transmit humidity or whenever the hu- 
midity traces are unreadable and the temperature is -40°C or 
warmer. If, however, the missing stratum is shallow (50 mb 
I> IL=(g}e) FMH#3 
