PROPHET Data Protocols


I. Accessing the archive for data submission/retrieval

1. Use ftp or a program that uses ftp to connect to
or ''

2. Log in
Username: 'aoss\prophet'

3. Change directory to the appropriate directory 4. To place data on archive use:
put <filename>
To retrieve data from archive use:
get <filename>
If you have any notes or comments you would like to put with data you may
place additional files with your data (i.e. a .readme file). There is no
limit on the number of characters in filenames, so please use names that
can fairly clearly identify the data in the files. Please send Mary Anne an
e-mail when you place data in the archive.


II. Data Formats

Data submitted to the archive should be in columns of ASCII text,
tab delimited or comma separated (text file).
The first column should be the time (GMT) associated with each
reported data point in the form: YYMMDDHHMMSS
(e.g. 000712141530 for a data point at 14:15:30 07/12/00)
Data time could also be reported in other useful formats as
additional columns (e.g. Year, Julian day, Second of day)
The time reported for averaged data should be the time
corresponding to the center of the averaging period. Where
possible, data should be reported as 1 minute/2 minute/5 minute
or longer averages.
Data file headers should include information on sample averaging
time. Headers for data from accumulated samples should include
information about the sample collection period (data files could
also include sample start and stop times).
Data file headers (or associated .readme files) should provide
information regarding measurement uncertainty.
Missing data should be replaced with a value -999. Data below
the detection limit should be replaced with a value -777.
We should keep in mind that correlation analyses require
coincident data--data file headers should specify whether or not
interpolation between data points is valid for conducting
correlation analyses