SGP/SONDE/C1 - A few bad temperature values at the beginning of the sounding

I've never seen this before.  The temperature reported by the radiosonde dropped over 30 
degC just after launch (usually a symptom of a broken temperature sensor), but then 
recovered after 14 seconds of bad readings.  The temperatures from 2 seconds through 14 seconds 
after launch are incorrect, the remainder of the sounding looks fine.  I have sent the 
data files to Vaisala for their consideration.

Only the first bad temperature reading will trigger a standard mentor QC flag.  I 

suggest interpolating the temperature values from 2 sec though 14 sec using the surface 
value and the value at 16 sec as endpoints.

• tdry
• dp

• tdry
• dp