Giant Icebergs of the Ross Sea, in situ drift and weather measurements

The data are provided in two formats:
The first, and recommended, format is a Matlab *.mat file that contains a data structure. This structure is self-explanatory, i.e., the names of the icebergs and the field variables appear in the structured naming convention. For example: the structure “iceberg” consists of the following sub structures:

iceberg =

B15K: [1x1 struct]
B15A: [1x1 struct]
B15J: [1x1 struct]
NIB: [1x1 struct]
DIB: [1x1 struct]
C16: [1x1 struct]
coastline: [1x1 struct]

The substructures are named by the iceberg on which the station that collected the data was located as of July 1 2008, e.g., C16 contains the data for iceberg C16. One exception: the two stations on B15K were originally located on B15A, but broke off as part of B15K when B15A split on 8 October, 2003. NIB stands for Nascent Iceberg on the Ross Ice Shelf, DIB stands for the Drygalski Ice Tongue/iceberg C25 (as the station duration spanned the spawning of a new iceberg when C16 struck the Drygalski Ice Tongue). The structure “coastline” represents a nice digitization of the Antarctic coast provided by Kelly Brunt, of the Raytheon Polar Services Co.

The substructures below the iceberg name contain the data in various raw and processed forms:

>> iceberg.B15A

ans =

desc: 'Iceberg data: constructed on 30-Jul-2008 13:33:24'
raw: [1x1 struct]
time: [1x1 struct]
pos: [1x1 struct]
T: [1x1 struct]
RH: [1x1 struct]
pressure: [1x1 struct]
windspeed: [1x1 struct]
solar: [1x1 struct]
winddirection: [1x1 struct]
orientation: [1x1 struct]
snowdistance: [1x1 struct]
errors: [1x1 struct]
station: [1x1 struct]
square_of_wind_speed: [1x1 struct]

The “raw” substructure contains the raw data processed from the ARGOS data transmission (including the hexidecimal ARGOS code):

>> iceberg.B15A.raw

ans =

desc: 'Raw unprocessed data from one hex file, split into ARGOS data and station data'
hex: [1x1 struct]
cr10x: [1x1 struct]
station_month: {1x50 cell}
station_year: {1x50 cell}
file: [1x1 struct]
argos: [1x1 struct]

The “raw.cr10x” data contains the data that has been decoded from the ARGOS (including engineering data transmitted on a less frequent basis), but no data processing and quality control have been provided:

>> iceberg.B15A.raw.cr10x

ans =

desc: 'Raw Decimal data converted from HEX'
engineering: [1x1 struct]
output: [212024x16 double]
data: [2689x16 double]

The processed data is located in the following structures: time, pos (position), T (temperatures), RH (relative humidity), pressure, windspeed, solar (if available), winddirection, orientation, snowdistance, square of wind speed. The “error” structure contains a flag set where original hex data is suspected of having an error.

>> iceberg.B15A

ans =

desc: 'Iceberg data: constructed on 30-Jul-2008 13:33:24'
raw: [1x1 struct]
time: [1x1 struct]
pos: [1x1 struct]
T: [1x1 struct]
RH: [1x1 struct]
pressure: [1x1 struct]
windspeed: [1x1 struct]
solar: [1x1 struct]
winddirection: [1x1 struct]
orientation: [1x1 struct]
snowdistance: [1x1 struct]
errors: [1x1 struct]
station: [1x1 struct]
square_of_wind_speed: [1x1 struct]

Units for data are as follows:

Time: either Matlab time (in days) or standard y m d h m s format
Temperature: degrees centigrade
Lat/lon: degrees, degrees minutes decimal minutes
RH: non-dimensional (percent)
Pressure: millibars
Windspeed: m/s
Winddirection: radians (unwrapped)
Orientation: radians (unwrapped)
Snow distance: meters
Solar: watts per meter squared per whole sky (downward flux) integrated over all band.

Iceberg C16 has two epochs. Epoch 1 represents data that was collected by storage module (no ARGOS transmission involved), Epoch 2 represents data that was only collected by ARGOS transmission)

Iceberg B15J (originally B15A prior to 8 October 2003) contains two stations: M1 and M2.

Nascent Iceberg is located on the Ross Ice Shelf, and is not an iceberg.

Data precision is as listed for the instruments used (Garmin GPS36, Campbell Scientific weather instruments (see publication)). Wind direction data is very poorly constrained due to difficulty working with magnetic compasses (which were inaccurate and unstable) and the difficulty of correcting iceberg orientation. For scientists interested in using the wind direction data, we recommend that the data be re-processed from the original ARGOS hex stream.

Physical description of the stations are available at the AMRC of the University of Wisconsin. A generic description is provided here: The station is set on a 10 meter tower which is dug into the firn surface about 2 meters. At the top of the station are GPS, windspeed and wind direction (RM Young instrument), pyranometer and downward looking snow distance sensor. Two thermometers (one with RH sensor) are mounted on the tower at different heights typically separated by 2-3 meters. Barometric pressure is sensed in an environmental enclosure that (in the case of C16) was buried at times.

All data are interpolated to a common time line for the convenience of comparison (with sample interval of 20 minutes). Look for the “common” substructures under each data heading including time.

The second format for data is ascii text files with headers that explain the variables. These files are one per station and are self explanatory.