This document describes the UK Met. Office Unified Model data for the HYREX project. There are separate dataset descriptions for the common infrastructure dataset, the vertically pointing radar and associated datasets supplied by the University of Salford, and for the Chilbolton radar data.
This dataset comprises a set of Mesoscale Model forecast data outputs for some of the HYREX Intensive Observing Periods (IOP's), of which there have been 11 to date. The data have been generated by the Joint Centre for Mesoscale Meteorology (JCMM) at Reading University.
The UK Meteorological Office Operational Numerical Weather Prediction (NWP) can be configured to run the Global Model, the Limited Area Model (LAM) or the Mesoscale Model (MM). The MM is designed to supply finer scale weather information for the UK up to one day forecast time. As the main aim of the MM is to predict weather elements rather than the large scale flow patterns, conditions near the surface and the moisture distribution are important. Surface observations of temperature, humidity and wind over land are at present assimilated only in the MM. In addition, Meteosat infrared imagery, radar rainfall imagery and surface cloud reports are combined with the latest 3 hour model forecast in MOPS, the Moisture Observation Pre-processing System. The output from MOPS (cloud fraction) is converted into a set of relative humidity soundings at each model grid-point, which are then assimilated in the MM.
The UK operational Mesoscale model uses a rotated latitude and longitude coordinate system in which the computational north pole is shifted to an actual position of 37.5 deg. N, 177.5 E. This is done to obtain fairly uniform horizontal resolution over the area of interest. The corners of the computational area are approximately in actual latitude/longitude: (60.1N, 16.6W; 60.2N, 10.7E; 46.6N, 12.7W; 46.7N, 7.1E). The grid-length is 0.15 deg. in each direction which is approximately 16.8 Km, giving 92x92 grid points running from the north - west corner. In the vertical the MM uses hybrid sigma/pressure coordinates and has 31 levels with the first 28 of these being moist levels. The lower levels are based upon sigma levels, defined as pressure/surface pressure ratio implying that they closely follow the terrain and the upper levels are defined on pressure levels.
The MM is hydrostatic (the basic equations assume that vertical accelerations are small enough to be ignored compared to gravity) and uses a B-grid staggering, i.e the variables are held on the Arakawa `B' grid. In the Arakawa B grid scheme, potential temperature (theta), surface pressure (p*) and specific humidity (q) etc. are calculated on grid points but the two horizontal wind components (u,v) are calculated in the centre of the grid boxes. The variables u, v, theta, q and Z (Geopotential Height) are held at levels etak, where eta is the vertical height coordinate and k is the vertical grid index, while the time derivative eta dot is held at the intermediate levels eta(k+1/2).
The research effort at the Reading JCMM is to make detailed analyses of MM data obtained during Intensive Observation Periods (IOP's). The main tool used in this research is the UK Meteorological Office Mesoscale Model (MM), which is a particular configuration of the Unified Model (UM). For each IOP, the model is run with initial data identical to that used in the operational forecast model runs, obtained from the Met Office. The model runs consist of specific case studies of forecast durations between 6 and 24 hours and correspond to the IOP periods. The model is run with and without Observational Data Assimilation to see how the inclusion of extra upper air sounding data affect the moisture budget calculations.
One of the aims of the research at Reading with these data is to use the wind and humidity fields from the MM in the moisture budget equation. During the IOP's, estimates of the volume-averaged atmospheric moisture flux convergence affecting southern England are obtained. The area for which the moisture budget is estimated will be varied systematically in order to define the lower limits in space and time for which the moisture budget equation provides a realistic prediction of observed precipitation. In each case the relative importance of the different terms in the equation is being examined.
In order to verify the moisture budget, areal averages of rainfall are obtained from analysis of rainguage data made by IOH and then compared with the results from the model.
The output data from each model run include a variety of field types, which are written to 3 output files labelled .pp0, .pp1 and .pp2 (referred to in the "Usage Profile" column in the table below). There are single level and multi-level fields: multi-level fields are written on both model and pressure levels. The following table shows the distribution of data fields between the three files. This table is also available as a plain-text file called FieldCodes.txt in the documentation directory /badc/hyrex/doc/um.
FIELD FIELD TIME DOMAIN USAGE CODE NAME PROFILE PROFILE PROFILE ______________________________________________________________________________ 1 Pstar after timestep (ts) T15MR DS PP2 2 U compnt of wind (wrt Model Grid) T15MR DM PP2 3 V compnt of wind (wrt Model Grid) T15MR DM PP2 4 Theta after timestep (ts) T15MR DM PP2 10 Specific Humidity after timestep (ts) T15MR DMW PP2 24 Surface Temperature after timestep T15MR DS PP0 3225 10 metre wind U-comp (wrt Model Grid) T15MR DS PP2 3226 10 metre wind V-comp (wrt Model Grid) T15MR DS PP2 3229 Evap from soil surf -amount kg/m2/ts T15MA DS PP2 3230 Evap from canopy - amount kg/m2/ts T15MA DS PP2 3232 Evaporation from sea (GBM) kg/m2/s T15MA DS PP2 3234 Surface Latent Heat Flux w/m2 T15MR DS PP2 3236 Temperature at 1.5m T15MR DS PP0 3237 Specific Humidity at 1.5m T15MR DS PP2 4004 Temperature after large scale precip T15MR DM PP1 4201 large scale rain amount kg/m2/ts T15MA DS PP2 4202 large scale snow amount kg/m2/ts T15MA DS PP2 4203 large scale rainfall rate kg/m2/s T15MR DS PP2 4204 large scale snowfall rate kg/m2/s T15MR DS PP2 4205 cloud liquid water after ls precip T15MR DMW PP1 4206 cloud ice content after ls precip T15MR DMW PP1 5201 convective rain amount kg/m2/ts T15MA DS PP2 5202 convective snow amount kg/m2/ts T15MA DS PP2 5205 convective rainfall rate kg/m2/s T15MR DS PP2 5206 convective snowfall rate kg/m2/s T15MR DS PP2 5207 Pressure at convective cloud base T15MR DS PP0 5208 Pressure at convective cloud top T15MR DS PP0 5209 Temperature after convection T15MR DM PP1 12201 Omega (In Advection) on model levels (U-Grid) T15MR DM PP2 12202 Omega (In Advection) on pressure levels (U-Grid) T15MR DP PP0 15201 U comp (Westerly) of wind on P levs T15MR DP PP0 15202 V comp (Southerly) of wind on P levs T15MR DP PP0 15216 T on pressure levs (U grid). T15MR DP PP0 15222 Omega on press levs (U grid). T15MR DP PP0 15226 Specific Humidity, P levs; (U grid) T15MR DP PP0 16202 Geopotential height: pressure levels T15MR DP PP0 16203 Temperature on pressure levels T15MR DP PP0 16222 Pressure at mean sea levels T15MR DS PP2 16225 Geopotential ht of model levels T15MR DM PP1 *** (see warning below) ________________________________________________________________________
Note that all winds are held on the U-grid. The U and V winds on Pressure levels are true Westerly and Southerly wind components but all other winds are orientated with the model grid.
The Time Profiles indicate the times at which the variables are written out:
All of the diagnostics are output from the model at the end of a time step unless explicitly stated, e.g. there are some temp fields at intermediate stages of the time step: temp after LS Precip or temp after convection, with the LS Precip and convection referring to stages during a time-step.
The Domain Profiles indicate the spatial characteristics of the output variables:
The Usage Profiles simply indicate to which PP file the particular output variable is written.
*** WARNING! There is a problem with one of the UM diagnostics due to a bug in the then current version (V3.4) of the UM. This means that field 16225 "Geopotential height of model levels" is wrong. The orography was added to this field twice by mistake. Orography data is available from the directory /badc/hyrex/data/um to allow for correction of this problem.
The corners of the UM computational area are approximately:
60.1 deg. N, 16.6 deg. W 60.2 deg. N, 10.7 deg. E 46.6 deg. N, 12.7 deg. W 46.7 deg. N, 7.1 deg. E
The grid-length is 0.15 deg. in each direction, which is approximately 16.8 Km, giving 92x92 grid points running from the north-west corner.
In the vertical the MM uses hybrid sigma/pressure coordinates (see section 1.1) and has 31 levels with the first 28 of these being moist levels. The data fields are output both on model levels and the following pressure levels (in mBar):
1000, 950, 900, 850, 800, 750, 700, 650, 600, 550, 500, 450, 400, 350, 300, 250.
There are 24 model runs covering various time ranges during the Intensive Observing Periods (IOP's) 1-4 and 6, plus 2 other dates. The UM was run for only a subset of each IOP, over a sufficient time period during the IOP to catch the precipitation events. The dates of the IOPs covered by UM runs are as follows:
IOP Date ______________ 1 19/10/94 2 05/12/94 3 06/12/94 4 14/02/95 6 10/07/95 ______________
There are two additional runs covering 16/07/95 and 07/06/96. The model runs consist of specific case studies of forecast durations between 6 and 24 hours, therefore on some dates there are several runs.
The model has a time step of 5 min., but the data are output every 3 time steps, i.e. 15 mins.
The full dates and times covered by each model run are shown below. The start time of the runs is constrained by the start data from the UKMO: it is only possible to start the model at 0Z, 6Z, 12Z and 18Z if not using data assimilation and at 21Z, 3Z, 9Z, 15Z if using assimilation. This table is also available in a plain-text version in the UM documentation directory /badc/hyrex/doc/um, filename Files.txt.
File START END IOP (Date) ASSIM mr5.pp0 06Z 14Z 1 (19/10/94) NO mr5.pp1 06Z 14Z 1 (19/10/94) NO mr5.pp2 06Z 14Z 1 (19/10/94) NO mr7.pp0 00Z 14Z 1 (19/10/94) NO mr7.pp1 00Z 14Z 1 (19/10/94) NO mr7.pp2 00Z 14Z 1 (19/10/94) NO mr8.pp0 00Z 12Z 3 (07/12/94) NO mr8.pp1 00Z 12Z 3 (07/12/94) NO mr8.pp2 00Z 12Z 3 (07/12/94) NO mr9.pp0 18Z 12Z 3 (06/12/94) NO mr9.pp1 18Z 12Z 3 (06/12/94) NO mr9.pp2 18Z 12Z 3 (06/12/94) NO mr10.pp0 12Z 11Z 3 (06/12/94) NO mr10.pp1 12Z 11Z 3 (06/12/94) NO mr10.pp2 12Z 11Z 3 (06/12/94) NO mr6.pp0 03Z 14Z 1 (19/10/94) YES mr6.pp1 03Z 14Z 1 (19/10/94) YES mr6.pp2 03Z 14Z 1 (19/10/94) YES mr11.pp0 21Z 12Z 3 (06/12/94) YES mr11.pp1 21Z 12Z 3 (06/12/94) YES mr11.pp2 21Z 12Z 3 (06/12/94) YES mr12.pp0 15Z 12Z 3 (06/12/94) YES mr12.pp1 15Z 12Z 3 (06/12/94) YES mr12.pp2 15Z 12Z 3 (06/12/94) YES mr13.pp0 06Z 18Z 2 (05/12/94) YES mr13.pp1 06Z 18Z 2 (05/12/94) YES mr13.pp2 06Z 18Z 2 (05/12/94) YES mr14.pp0 03Z 18Z 2 (05/12/94) YES mr14.pp1 03Z 18Z 2 (05/12/94) YES mr14.pp2 03Z 18Z 2 (05/12/94) YES mr15.pp0 12Z 18Z 2 (05/12/94) NO mr15.pp1 12Z 18Z 2 (05/12/94) NO mr15.pp2 12Z 18Z 2 (05/12/94) NO mr16.pp0 09Z 18Z 2 (05/12/94) YES mr16.pp1 09Z 18Z 2 (05/12/94) YES mr16.pp2 09Z 18Z 2 (05/12/94) YES mr17.pp0 06Z 20Z 4 (14/02/95) NO mr17.pp1 06Z 20Z 4 (14/02/95) NO mr17.pp2 06Z 20Z 4 (14/02/95) NO mr18.pp0 03Z 20Z 4 (14/02/95) YES mr18.pp1 03Z 20Z 4 (14/02/95) YES mr18.pp2 03Z 20Z 4 (14/02/95) YES mr19.pp0 12Z 20Z 4 (14/02/95) NO mr19.pp1 12Z 20Z 4 (14/02/95) NO mr19.pp2 12Z 20Z 4 (14/02/95) NO mr20.pp0 09Z 20Z 4 (14/02/95) YES mr20.pp1 09Z 20Z 4 (14/02/95) YES mr20.pp2 09Z 20Z 4 (14/02/95) YES mr21.pp0 00Z 21Z 6 (10/07/95) NO mr21.pp1 00Z 21Z 6 (10/07/95) NO mr21.pp2 00Z 21Z 6 (10/07/95) NO mr22.pp0 21Z 21Z 6 (10/07/95) YES mr22.pp1 21Z 21Z 6 (10/07/95) YES mr22.pp2 21Z 21Z 6 (10/07/95) YES mr23.pp0 06Z 21Z 6 (10/07/95) NO mr23.pp1 06Z 21Z 6 (10/07/95) NO mr23.pp2 06Z 21Z 6 (10/07/95) NO mr24.pp0 03Z 21Z 6 (10/07/95) YES mr24.pp1 03Z 21Z 6 (10/07/95) YES mr24.pp2 03Z 21Z 6 (10/07/95) YES mr25.pp0 12Z 21Z 6 (10/07/95) NO mr25.pp1 12Z 21Z 6 (10/07/95) NO mr25.pp2 12Z 21Z 6 (10/07/95) NO mr26.pp0 09Z 21Z 6 (10/07/95) YES mr26.pp1 09Z 21Z 6 (10/07/95) YES mr26.pp2 09Z 21Z 6 (10/07/95) YES mr27.pp0 09Z 24Z - (16/07/95) YES mr27.pp1 09Z 24Z - (16/07/95) YES mr27.pp2 09Z 24Z - (16/07/95) YES mr28.pp0 09Z 24Z - (07/06/96) YES mr28.pp1 09Z 24Z - (07/06/96) YES mr28.pp2 09Z 24Z - (07/06/96) YES
N.B. Where the ending time is before the starting time, this means that the run ended the following day. Model runs 1-4 are not included as they were not part of the HYREX project. Further details about the Intense Observing Periods are available, including the complete dates and times and descriptions of the weather conditions.
The UM data is output into 3 files per model run, which range in size between 80 and 480 Mbytes. (The filenames are tabulated under 'Temporal Coverage' above). These are IEEE binary files in Met. Office "PP" format, consisting of unpacked headers (mostly integer data) followed by floating point field data.
Data are compressed using gzip to save disk space.
Software to read and plot the um data is available at: /badc/hyrex/software/um/.
As mentioned above, the UM data are stored in very large binary files, so software is required to read the files and extract the fields of interest. JCMM have provided a UNIX shell script exfield, which runs a FORTRAN program to read the data and extract required fields. Both are available in the directory /badc/hyrex/software/um. There are versions of the software for Cray (sort_cray.f) and IEEE (Sun etc - sort_ieee.f) platforms. The software has also been run on the BADC's DEC UNIX machine login.badc.rl.ac.uk, by compiling with the -convert big_endian qualifier to the f77 compiler as follows:
$ f77 -o sort.x -convert big_endian sort_ieee.f
(Comparisons of sample data values on login.badc.rl.ac.uk and a Sun machine showed the same data values were output on both systems, so the floating point values are successfully "byte-swapped" on the DEC system.) You can use the exfield software to extract subsets of data on login.badc.rl.ac.uk, ready for FTP to your local system. Note that you may need to ask BADC Support for an account on login.badc.rl.ac.uk to process the data files before transferring the data to your local machine.
Using exfield you can select fields by their field code, select a sub-area from the model grid by specifying the latitude and longitude limits (as grid array indices), select specific pressure or model levels or ranges of levels, and specify the data output times or frequency (as dump time indices). The program produces an output file containing header information such as pressure levels and field codes contained in the model run data, and writes the selected data fields to a PP format file.
If you want to use exfield to extract subsets of data on login.badc.rl.ac.uk, you need to copy the exfield shell script to your private filestore and edit it to select the data you require. However, there is no need to copy the sort executable - you can set up an alias to use the copy in the software directories.
JCMM have provided a program called JPLOT to manipulate and plot the output data from the exfield read and subsetting software. JPLOT can be found in the software/um directory. JPLOT is a fully-interactive X-windows widget based IDL graphical display program, designed to run on IEEE UNIX systems. It is intended to handle data from a variety of sources including UM data extractions.
JPLOT can display data in a number of ways:
The user can then display the 2D sections in a number of ways, for example using images, contours or surface plots. Also various tool are available such as zoom utilities and other data field manipulation widgets. UM data may be displayed either on a rectangle corresponding to the equatorial grid (which the limited area models are run on) or with a standard cylindrical map projection. An overplotting facility also exists for plotting one field over another. The user can interactively alter various display parameters such as contouring parameters (values, linestyles, labelling, etc.) and axis values and titles etc. PostScript files can be created for hardcopies of graphical output.
More information about JPLOT is available from the JCMM website.