NERC Logo MSTRF Logo

HOME

ANNOUNCEMENTS

CONTACTS

DATA

INSTRUMENTS

MAIN MENU

MISCELLANEOUS

PLOTS

PUBLICATIONS

SCIENCE

WIND-PROFILER
PRINCIPLES



For this page:
INTERNAL LINKS

EXTERNAL LINKS
THE NERC MST RADAR FACILITY AT ABERYSTWYTH
FILE FORMAT FOR VERSION-2 MST RADAR RADIAL DATA
WARNING: Version-2 MST radar data products are DEPRECATED
Users are encouraged to make use of version-3 radial files.
Click here to find out about earlier versions of the signal processing.

File contents
These files contain radial (i.e. along-beam) profiles of the radar return parameters (noise power, signal power, Doppler shift and spectral width) for different beam pointing directions. Data are recorded at 150 m intervals in range and cover the approximate range 2 - 20 (for the ST mode) or 58 - 96 km (for the M mode). For most purposes the Cartesian data will suffice and it should only be necessary to examine the radial data for specialist studies.
Click here to find out about the contents of other files.

File naming convention:
radar-mst_capel-dewi_YYYYMMDD_AARRR_radial_v2.na

YYYY is a 4-digit year [1990 - ]
MM is a 2-digit month [01 - 12]
DD is a 2-digit day [01 - 31]
AA is the altitude mode ['st': approximately 2 - 20 km | 'm': approximately 58 - 96 km]
RRR is the range resolution (m) [150 | 300 | 600 | 1200 | 2400 | 4800]

i.e. radar-mst_capel-dewi_20050101_st300_radial_v2.na contains 300 m resolution radial data over the ST altitude range for 1st January 2005.
Click here for the background to the file naming convention.

File location: /badc/mst/data/mst-products-v2/radial/
Click here for the location of other files.

Archiving convention: YYYY/MM
Click here for a further explanation.

Data availability: refer to Instrument performance weblog

File format
(ASCII) NASA-Ames files, with a File Format Index of 2110, are used, i.e. the same as for Version-1 data products. However, there are some small differences in the file contents. Users will therefore need to make a few modifications to any software designed for dealing with Version-1 products.

Only those aspects of the file format which are essential for reading the data will be described. For a full description of the NASA-Ames formats, consult the Gaines and Hipskind [1998] document.

The radial file for 1st January 2005 will be used as an example. Text in green represents actual file contents. Text in red is for explanatory purposes only.
Header Lines
Line 1: 88 2110
Integer 1 corresponds to the total number of header lines, nr_header_lines
Integer 2 corresponds to the File Format Index

Line 7: 2005 01 01 2005 01 10
Integers 1 - 3 correspond to the year, month and day on which the observations were made.
Integers 4 - 6 correspond to the year, month and day on which the file was created.

Line 11: 6
The number of primary variables, nr_primary_vars.

Line 13: 999.99 999.99 999.999 99.999 999 99999
Missing data values, in order, for the nr_primary_vars primary variables listed below. Greater use is made of missing data values than in the Version-1 files and so users should make sure that they make the appropriate substitutions when they read in the data.

Lines 14 - 19:
The names of the primary variables (listed below).

Line 20: 16
The number of auxilliary variables, nr_aux_vars).

Lines 23 - 38:
The names of the nr_aux_vars auxilliary variables (listed below).

Line 45: 130 3660 1
Integer 1 corresponds to the number of range gates per dwell, nr_range_gates,
Integer 2 corresponds to the total number of dwells, total_nr_dwells, and
Integer 3 corresponds to the number of different cycle formats, which is typically (but not always) equal to 1.

Lines 58 - 86:
Global data attributes following the netCDF CF 1.0 convention. This includes a processing history. It is not necessary to know this information in order to read the file.

Data reading loop
After reading the above mentioned lines, wind forward to line (nr_header_lines + 1) where the data begin. Associated with each dwell of observation there is a single line of auxiliary variables followed by nr_gates lines of primary variables. The data can therefore be read with a simple loop structure of the form (shown here in Fortran syntax):
do total_dwell_nr = 1,total_nr_dwells
  read_auxiliary_variables
  do gate_nr = 1,nr_gates
    read_primary_variables
  end do
end do
Reading auxiliary variables
Each auxiliary variables line contains the second independent variable (dwell time) followed by the 16 auxilliary variables (a mixture of floating point, F, and integer, I values), shown here for the first dwell in the file: 116 130 1 1 1 11 27.7 6.0 8 2 2 320 18 147 512 128 1
Value 1: Cycle time (s) I
Technically speaking this is the second independent variable rather than an auxiliary variable. The time is given in seconds since 00:00:00 UTC for the day in question.

Value 2: Number of range gates I
This is the same as nr_gates given in line 45 of the header and so can be ignored.

Value 3: Cycle number I

Value 4: Cycle format number I

Value 5: Dwell number (within current cycle) I
which is different from total_dwell_nr (except for the first cycle).

Value 6: Beam pointing number I
which ranges between 0 and 16. The same information is contained within the next 2 auxilliary variables. Follow
this link to see the relationship between beam number and beam pointing direction.

Value 7: Beam pointing azimuth (degrees clockwise from North) F
which can have values of 0.0° (for a vertically directed beam only), 27.5°, 72.5°, 117.5°, 162.5°, 207.5°, 252.5°, 297.5° or 342.5°.

Value 8: Beam pointing zenith angle (degrees from vertical) F
which can have values of 0.0°, 4.2°, 6.0°, 8.5° or 12.0°.

Value 9: Transmitter pulse length (μs) I
which can have values of 1, 2, 4, 8, 16 or 32 μs.

Value 10: Transmitter sub-pulse length (μs) I
which can have values of 1, 2, or 4 μs, for pulse lengths of 4 μs and longer for which pulse coding has been used. If the sub-pulse length is equal to the pulse length, then no pulse coding has been used.

Value 11: Receiver bandwidth (μs) I
this is equal to the transmitter sub-pulse length, or to the pulse length if no pulse coding has been used, and determines the range resolution (values shown in brackets). Possible values are 1 μs (150 m), 2 μs (300 m), 4 μs (600 m), 8 μs (1200 m), 16 μs (2400 m) and 32 μs (4800 m). This value is the same for all dwells within a single file.

Value 12: Inter-pulse period (μs) I
This determines the maximum unambiguous range from which radar returns can be received (values in brackets). It can have values of 80 μs (12 km), 160 μs (24 km), 320 μs (48 km) and 640 μs (96 km).

Value 13: Bottom range gate number I
The bottom and top (next value) range gate numbers are given for reference only. The total number number of range gates was already given in value 2 and the range of each gate from the radar is shown in the first column of data for each dwell.

Value 14: Top range gate number I
See value 13.

Value 15: Number of coherent integrations I
This is the number of raw radar samples averaged.

Value 16: Discrete Fourier transform length I
This is the number of (averaged) samples from which a Doppler power spectrum is derived.

Value 17: Number of incoherent integrations I
This is the number of Doppler power spectra averaged together before the principal spectral parameters are evaluated.

Reading primary variables
These lines contain 7 values: the first independent variable (range from the radar) followed by the 6 primary variables, as shown below for the first line of the first cycle. These are a mixture of floating point, F, and integer, I values.
1645.0 41.98 59.23 0.176 0.510 32 32799
Value 1: Range from the radar (m) F
Technically speaking this is the first independent variable rather than a primary variable. The same range gate grid is used for all dwells within the file and so only needs to be saved once. The range is given in metres away from the radar, which is at 50 m above mean sea level.

Value 2: Spectral noise power (dB) F
The noise power is included for reference and is not typically used as a parameter in its own right.

Value 3: Radar return signal power (dB) F
P_dB = 10 × log10[P_linear]

Value 4: Radial air velocity (m/s) F
Positive values imply motion AWAY from the radar.

Value 5: Radar return spectral width (m/s) F
This corresponds to an e-1/2 half-width. Follow
this link for further information.

Value 6: Peak power spectral density (PSD) relative to the mean noise PSD (dB) I
This is used for reliability flagging. Signals with a value of greater than 10 dB are typically clearly distinguishable above the background noise. Signals with values of less than 10 dB might still be reliable but should be treated with caution.

Value 7: Reliability flag I
Note that this uses a different convention to that used for the reliability flags in the Version-1 files. A larger range of information is now stored bit-wise in a 16-bit integer, with the 15th bit being used as an overall indicator of data reliability. Therefore data may be regarded as reliable if:

flag value >= 32768

The convention has been adopted so that more pieces of information can be added to the flag value at a later date, without changing the acceptance condition. At present, only 6 of the available 16 bits are used:

Bit number Information
0 peak signal PSD to noise PSD >= threshold value
1 time continuity threshold exceded
2 complementary beams are available
3 complementary beam factor >= threshold value
4 complementary beam factor is significant
15 overall reliability flag
The threshold values used for a particular file are quoted in header lines 52 - 56.

Internal Links:
Return to the top of the page
Gaining access to the data
File naming convention
Data archiving convention
Data locations
The differences between signal processing versions
The contents of other data files
External Links:
Full description of the NASA-Ames formats: Gaines and Hipskind [1998]
Page maintained by David Hooper
Last updated 9th November 2006