UKMO Unified Model data (mesoscale grid)

The UK Meteorological Office has provided the 6-hourly analyses and the intervening hourly forecasts from the mesoscale version of the Unified Model for the flight days of CLARE (if you are interested in other days during the CLARE campaign then you will have to use ECMWF model data, which includes every day of the campaign but only over Chilbolton). Two types of file are provided: the basic model model data at each model level along the flight tracks and model data plus derived 35 and 94 GHz radar attenuation parameters for Chilbolton only on the flight days. The latter provides information necessary to correct for attenuation and use differential attenuation to estimate liquid water content. See also the brief notes at the end of this document.

Model data on flight days

The flights were fortunately all along the same 260 degree azimuth to Chilbolton, and for each flight day I have interpolated the model fields on to six points along the flight track of the aircraft. The points are 10 km apart. The model resolution is around 12.25 km in the horizontal. Interpolation has been performed using an inverse distance method in the horizontal, but no interpolation has been performed in the vertical (the data is on the original 35 levels). The longitude and latitude of the interpolation points are close to the following:

Dist. from Chilbolton     Longitude (deg. East)    Latitude (deg. North)
   0 km                       -1.4370                  51.1445
  10 km                       -1.579                   51.129
  20 km                       -1.719                   51.113
  30 km                       -1.861                   51.098
  40 km                       -2.002                   51.082
  50 km                       -2.143                   51.066

The format used is similar to the ECMWF model data for the CLARE campaign, which was provided by ECMWF already interpolated over Chilbolton.

Each flight day consists of 6 compressed files corresponding to the 6 points along the flight track. Only Chilbolton data has been extracted for non-flight days. These files are named:

UKMO_yymmdd_kk_var.Z

where kk is the distance from Chilbolton in km. The compressed size of each file is around 36 kBytes, uncompressed is 135 kBytes.

Right-click on the following links to download a day of data:

UKMO_981007_00_var.Z (7 October)
UKMO_981007_10_var.Z (7 October)
UKMO_981007_20_var.Z (7 October)
UKMO_981007_30_var.Z (7 October)
UKMO_981007_40_var.Z (7 October)
UKMO_981007_50_var.Z (7 October)

UKMO_981013_00_var.Z (13 October)
UKMO_981013_10_var.Z (13 October)
UKMO_981013_20_var.Z (13 October)
UKMO_981013_30_var.Z (13 October)
UKMO_981013_40_var.Z (13 October)
UKMO_981013_50_var.Z (13 October)

UKMO_981014_00_var.Z (14 October)
UKMO_981014_10_var.Z (14 October)
UKMO_981014_20_var.Z (14 October)
UKMO_981014_30_var.Z (14 October)
UKMO_981014_40_var.Z (14 October)
UKMO_981014_50_var.Z (14 October)

UKMO_981016_00_var.Z (16 October)
UKMO_981016_10_var.Z (16 October)
UKMO_981016_20_var.Z (16 October)
UKMO_981016_30_var.Z (16 October)
UKMO_981016_40_var.Z (16 October)
UKMO_981016_50_var.Z (16 October)

UKMO_981020_00_var.Z (20 October)
UKMO_981020_10_var.Z (20 October)
UKMO_981020_20_var.Z (20 October)
UKMO_981020_30_var.Z (20 October)
UKMO_981020_40_var.Z (20 October)
UKMO_981020_50_var.Z (20 October)

UKMO_981021_00_var.Z (21 October)
UKMO_981021_10_var.Z (21 October)
UKMO_981021_20_var.Z (21 October)
UKMO_981021_30_var.Z (21 October)
UKMO_981021_40_var.Z (21 October)
UKMO_981021_50_var.Z (21 October)

UKMO_981022_00_var.Z (22 October)
UKMO_981022_10_var.Z (22 October)
UKMO_981022_20_var.Z (22 October)
UKMO_981022_30_var.Z (22 October)
UKMO_981022_40_var.Z (22 October)
UKMO_981022_50_var.Z (22 October)

Each file starts with a 3 line header. The first line explains the second. The four numbers on the second line are latitude, longitude, the y-location of the interpolation point on the mesoscale grid and the x-location of the point on the mesoscale grid (where the first grid-point in each direction is `1'). The third line describes the data which follows. The first two columns of data are 'date' and 'time', the date and hour (UTC) of the analysis or forecast.

The data is held on the 35 model levels. Level 1 is the lowest level and level 35 is the highest. Note that this is the opposite convention to that used by ECMWF. The order of the remaining variables is as follows:

Radar attenuation parameters

From the model data I have calculated a number of radar parameters that are of interest to anyone who wishes to correct radar data at 35 or 94 GHz for attenuation, or use attenuation to measure liquid water content. These are attenuation coefficents for atmospheric gases and liquid water, and the dielectric parameter |K|² which is a function of temperature at millimetre wavelengths. The files are:

UKMO_981007_diffatten_var.Z (7 October)
UKMO_981013_diffatten_var.Z (13 October)
UKMO_981014_diffatten_var.Z (14 October)
UKMO_981016_diffatten_var.Z (16 October)
UKMO_981020_diffatten_var.Z (20 October)
UKMO_981021_diffatten_var.Z (21 October)
UKMO_981022_diffatten_var.Z (22 October)

The format is the same as for the basic model data except that after the date and time, the columns in the file correspond to the following parameters:

lev - integer model level
zlevs - height above mean sea level (m)
p - pressure (Pa)
u - zonal wind component (m/s)
v - meridional wind component (m/s)
T - temperature (K)
q - water vapour density (kg/m³)
LWC - cloud liquid water content (kg/m³)
IWC - cloud ice water content (kg/m³)
RH - relative humidity (percent/100.)
omega - vertical velocity in pressure coordinates (Pa/s)
Ag35 - one-way gaseous attenuation at 35 GHz, using model temperature, pressure and humidity (dB/km)
Ag94 - one-way gaseous attenuation at 94 GHz, using model temperature, pressure and humidity (dB/km)
Ag35sat - one-way gaseous attenuation at 35 GHz, using model temperature and pressure but assuming saturation with respect to water or ice depending on the temperature (dB/km)
Ag94sat - one-way gaseous attenuation at 94 GHz, using model temperature and pressure but assuming saturation with respect to water or ice depending on the temperature (dB/km)
Al35 - one-way liquid water attenuation (Rayleigh approximation) at 35 GHz (dB km^-1 (g m^-3)^-1)
Al94 - one-way liquid water attenuation (Rayleigh approximation) at 94 GHz (dB km^-1 (g m^-3)^-1)
K235 - dielectric parameter |K|² of liquid water at 35 GHz (dimensionless)
K294 - dielectric parameter |K|² of liquid water at 94 GHz (dimensionless)

Notes

Relative humidity was calculated directly from q and T and is taken to be with respect to ice below 0°C and with respect to liquid water above. Note that the water contents are in kg/m³ rather than kg/kg as used by the ECMWF.

The attenuation at 35 and 94 GHz due to atmospheric gases (predominantly water-vapour and oxygen),`Ag35' and `Ag94', were calculated from the UKMO model parameters using the line-by-line model of Liebe (1985), Radio Science, 20, 1069-1089. The dielectric constant of liquid water from which the liquid-water attenuation and |K|² were calculated was taken from Liebe, Manabe and Hufford (1989), IEEE AP, 37, 1617-1623.

Many thanks to Ed Dicks and Peter Panagi for providing the raw data.

Problems with extraction should be addressed to Robin Hogan.