Institute for Mathematics Applied to Geosciences (IMAGe)

An Example
The Functions
PWV(Precipitable Water Vapor)
a few more flights
Related sites
ATD's Rawinsonde Project

RadioSonde: A discussion of the data format

Most of this data was collected by a radiosonde which was carried aloft by a weather balloon. This data was collected as part of a special field program, but the instrument and data formats are very common in the atmospheric community.

Fundamentally; temperature, pressure, and humidity sensors are bundled with a radio transmitter and are either sent aloft on a balloon (rawindsonde) or tied to a small parachute and dropped from a plane (dropsonde). Either way, the location of the sonde is observed (either visually or with GPS), the data from the radio transmitter recorded, and the result is an ascii file that contains a header (of varying length) with descriptive information followed by a table of observational data and derived quantities.

For rawindsondes, the first entry for the table is actually the surface data from the weather station, but the remaining entries are from the sonde. Also, the wind is inferred from the position of the balloon. Since this is real data, there are missing values – which arise all the time. If you get a profile without missing values, you can be sure it is synthetic or interpolated. The datastreme ends when the operator calls it quits. This happens when a) the balloon breaks, or b) the radio reception has so many dropouts that is is not worth it anymore. Different fields have different missing flags. Where possible, we replace the "standard" missing flags with NaN


The nominal format and is a mixture of ancillary data (contained in the first ???? lines of the file), and the instrument measures and some derived fields that follow in a fixed-format table with an unknown number of lines. Part of the formatting of the table is that each column has a whitespace-separated column name. This line is crucial to the proper function of the RadioSonde package. If the datafile does not contain information as to which column is what variable ... you will have to modify getsonde.

The getsonde function actually does not care what the column ordering (i.e. the order of the variables) nor does it care if all of the ones listed below are present. Certainly, at a minimum, a profile will have pressure, temperature, and rh. Most of the necessary quantities can be derived from these.

Field varname width format description units missing value
[ 1] time 6 F6.1 Time Seconds 9999.0
[ 2] press 6 F6.1 Pressure Millibars 9999.0
[ 3] temp 5 F5.1 Dry-bulb Temperature Degrees C 999.0
[ 4] dewpt 5 F5.1 Dew Point Temperature Degrees C 999.0
[ 5] rhum 5 F5.1 Relative Humidity Percent 999.0
[ 6] uwind 6 F6.1 U Wind Component Meters/Second 9999.0
[ 7] vwind 6 F6.1 V Wind Component Meters/Second 9999.0
[ 8] wspd 5 F5.1 Wind Speed Meters/Second 999.0
[ 9] wdir 5 F5.1 Wind Direction Degrees 999.0
[10] dz 5 F5.1 Ascension Rate Meters/Second 999.0
[11] lon 8 F8.3 Longitude Degrees 9999.0
[12] lat 7 F7.3 Latitude Degrees 999.0
[13] range 5 F5.1 Variable (see below) 999.0
[14] angle 5 F5.1 Variable (see below) 999.0
[15] alt 7 F7.1 Altitude Meters 99999.0
[16] qp 4 F4.1 QC flag for Pressure Code (see below) 99.0
[17] qt 4 F4.1 QC flag for Temperature Code (see below) 99.0
[18] qh 4 F4.1 QC flag for Humidity Code (see below) 99.0
[19] qu 4 F4.1 QC flag for U Component Code (see below) 99.0
[20] qv 4 F4.1 QC flag for V Component Code (see below) 99.0
[21] quv 4 F4.1 QC flag for Ascension Rate Code (see below) 99.0

The missing value flags are in brackets []. Fields 13 and 14 are `variable' because depending on the sounding system the variables used in these positions can vary.

Fields 16 through 21 contain the Quality Control information (flags) generated at JOSS. These flags are based on the automated or visual checks made.

The JOSS QC flag codes are as follows:

99.0 Unchecked (QC information is `missing.') (`UNCHECKED')
1.0 Checked, datum seems physically reasonable. (`GOOD')
2.0 Checked, datum seems questionable on physical basis. (`MAYBE')
3.0 Checked, datum seems to be in error. (`BAD')
4.0 Checked, datum is interpolated. (`ESTIMATED')
9.0 Checked, datum was missing in original file. (`MISSING')