Reference
Usage
This section is intended as a quick look-up on the use of metadata items within IMPROVER; they are listed alphabetically. It also indicates whether the metadata item is part of the basic NetCDF metadata, the Climate and Forecast or CF Metadata Conventions or specific to IMPROVER and whether it must, should or could be present.
Metadata items
altitude
This CF coordinate variable holds the height above sea level
for site data
and has the standard_name attribute set to altitude.
This should be present for the site data to allow the data to be fully exploited.
axis
This CF attribute uses a single capitalised character to indicate
how a coordinate variable should be intepreted.
It can take the values X, Y, Z and T
representing the three spatial directions and time.
Axis values should be set for non-scalar coordinate variables,
which for IMPROVER gridded data are X and Y.
IMPROVER also has threshold and percentile
coordinate variables, but there are no standards for labelling these.
blend_time
This IMPROVER-specific variable
has been added to indicate when the data was processed (blended)
to generate this forecast, and can be used to indicate how ‘fresh’
the data is.
This has the long_name attribute set to blend_time,
but otherwise takes the same form as the time variable.
Ideally, this should be present.
bounds
This CF attribute provides a label pointing to a separate variable defining the bounds of a point on an axis, most commonly the start and end of a time period.
This must be present if an associated ‘bounds variable’ exists.
calendar
For IMPROVER, this CF attribute is set to gregorian to indicate
that a Gregorian (standard) calendar is used.
cell_methods
This CF attribute can be used to describe the application of simple
statistical processing to a variable
(e.g. a maximum of the temperature over a period of time).
It is a string comprised of a list of blank-separated words of the form
name: method.
The name can be a dimension of the variable, a scalar coordinate variable,
a valid standard name, or the word area.
The method should be selected from a standard list
described in the CF Metadata Conventions
It is also possible to included additional information
in the form of a comment
If any method other than point is specified for a given axis,
then bounds must also be provided for that axis.
For example, time: maximum would indicate the maximum
over the period of time described by the time bounds.
Cell methods are covered in more detail in the Statistical Processing.
Conventions
This netCDF attribute specifies a space-separated (or comma-separated if conventions have spaces in their titles) list of metadata conventions that the file conforms to. Up until CF version 1.6, strictly only the CF Metadata Conventions were allowed to be declared here, but a change at 1.7 allowed multiple conventions.
This must be set to include the appropriate version of the CF Convention which should include any other conventions that are used (although, at present, there is no entry set automatically to indicate the extensions used to support enhancements used by IMPROVER).
coordinates
This CF attribute lists any coordinates that do not appear as dimensioned coordinate variables, i.e. those that do not appear as dimensions of the main variable. This covers both scalar coordinate variables (single-valued coordinates, with no dimension) and auxillary coordinate variables (variables that contain coordinate data but are not coordinate variables, usually because they depend on more than one dimension).
This should be included where coordinates are present that
do not appear as dimensioned coordinate variables.
For IMPROVER gridded data this would typically be the
scalar coordinate variables:
blend_time, height and time
and for spot data the scalar coordinate variables:
altitude, blend_time, latitude, longitude,
met_office_site_id and time wmo_id.
forecast_reference_time
This CF variable represents the nominal data time or start time of a
model forecast run,
and has the standard_name attribute set to forecast_reference_time.
Ideally, this should no longer be used for IMPROVER data.
Warning
Use of forecast_reference_time in IMPROVER is deprecated
as it is at best unhelpful and at worst it is confusing,
as IMPROVER generates a blend from multiple sources
with different start times so there is no unique data time.
forecast_period
This CF variable represents the interval between
the forecast_reference_time and the validity time (time)
and has the standard_name attribute set to forecast_period.
Ideally, this should no longer be used for IMPROVER data.
Warning
Use of forecast_period in IMPROVER is deprecated
as it is at best unhelpful and at worst it is confusing,
as IMPROVER generates a blend from multiple sources
with different start times so there is no unique data time.
grid_mapping
This CF attribute provides a label pointing to a separate grid mapping variable, which more fully describes the map projection.
This must be present for gridded data, as must the associated grid mapping variable.
height
This CF vertical coordinate variable is included in some cases to fully describe the quantity of interest, for single-level variables appearing as a scalar coordinate variable.
This should be included if there is any ambiguity in the interpretation
of quantity of interest if it is omitted.
(e.g. an inclusion of height with a value of 1.5 m
for the representatiion of screen level.)
history
Ideally, this netCDF attribute should provide a list of the applications
that have modified the original data (i.e. an audit trail),
with recommended practice being to add a date/time stamp
(in the form YYYY-MM-DDThh:mm:ssZ) and identify the software package.
However, in practice, this is far from straightforward for IMPROVER
as it processes a range of model runs,
so there is no single, sequential processing chain
from which to generate such an audit trail,
making it impossible to accurately maintain previous history information.
This is not currently set in IMPROVER.
institution
This CF attribute specifies where the original data was produced.
This must be present and should take the name of the institution from where the data originated if only data from a single model has been processed. However, it should be set to the institution running the post-processing for multi-model blended data.
latitude
This coordinate variable represents one half of the positional
information for gridded data held on a
Latitude-Longitude (strictly, equirectangular) projection.
This is also used for site positions, which are only provided
in latitude and longitude.
It has the standard_name attribute set to latitude
and units set to degrees.
Unless explicitly stated in the metadata,
the latitude and longitude can be considered as relative the WGS84
or the World Geodetic System 1984 datum.
All data must contain either this or projection_y_coordinate variable.
For gridded data, if any statistical processing over the coordinate
has been applied,
there must also be an associated latitude_bnds variable
providing the bounds over which cell_methods are applied,
although this is often included anyway to define the cell boundaries.
The latitude_bnds variable has no attributes as it is tied to the
main coordinate variable.
least_significant_digit
This is a variable attribute used by netCDF-writing software to
specify the precision that is maintained when ‘bit-shaving’
is applied to provide improved file compression.
The example value of 3LL indicates that a precision of 3 decimal places
is preserved, i.e. values precise to the nearest 0.001.
As ‘bit-shaving’ is zeroing bits
(that are providing an unrequired level precision),
this would actually be implemented as the power of 2 nearest 0.001.
This is usually included automatically where the precision is limited.
The driver for the use of ‘bit-shaving’ is that although it requires no extension to the software to read the data (the number formats in the file are not changed), it facilitates more effective reduction in file size, when lossless compression is applied.
long_name
This netCDF-specific variable attribute provides
a descriptive name that is not governed by CF.
If a CF Standard Name exists for the quantity,
this should be used and the long_name is usually omitted.s
A standard_name or long_name must be present.
longitude
This coordinate variable represents one half of the positional
information for gridded data held on a
Latitude-Longitude (strictly, equirectangular) projection.
This is also used for site positions, which are only provided
in latitude and longitude.
It has the standard_name attribute set to longitude
and units set to degrees.
Unless explicitly stated in the metadata,
the latitude and longitude can be considered as relative the WGS84
or the World Geodetic System 1984 datum.
All data must contain either this or projection_x_coordinate variable.
For gridded data, if any statistical processing over the coordinate
has been applied,
there must also be an associated longitude_bnds variable
providing the bounds over which cell_methods are applied,
although this is ofsten included anyway to define the cell boundaries.
The longitude_bnds variable has no attributes as it is tied to the
main coordinate variable.
met_office_site_id
This IMPROVER-specific coordinate variable is an 8-character string, zero-padded ID number used by the Met Office to label all sites. Within the IMPROVER code, the name is user configurable, such that it can be changed for different institutions / indices.
Although this precise variable is not appropriate for most users other than the Met Office, it is advisable to implement some form of site identification that has unique elements and is complete.
mosg__
This is intended to indicate a MOSG (Met Office standard grid) namespace. It prefixes attributes to show that they are separate from the CF Metadata Conventions attributes.
mosg__model_configuration
This is an IMPROVER-specific global attribute and provides a space-separated list of model identifiers denoting which sources have contributed to the blend. The naming is fairly arbitary, but at the Met Office we have chosen to indicate the models in a coded form:
gl= global model
uk= high-resolution UK domain model
nc= (extrapolation-based) nowcast
with a secondary component indicating whether the
source is deterministic (det) or an ensemble (ens).
For example, uk_ens indicates our UK ensemble model, MOGREPS-UK.
mosg__model_run
This is an IMPROVER-specific global attribute
which extends the information provided by
mosg__model_configuration, to detail the contribution
of specific model runs (also known as cycles) to the blend.
This is represented as a list of new line (\n) separated
composite entries of the form:
model identifier:cycle time in format yyyymmddTHHMMZ:weight
percentile
This is an IMPROVER-specific coordinate variable that holds
the set of percentile levels for which values of the variable of
interest are generated.
It has a long_name attribute set to percentile
and a units attribute set to %
This must be present for percentile variables.
positive
Indicates the direction in which values of the vertical coordinate increase,
i.e. where the vertical coordinate is pressure,
the positive attribute is down.
This should be present for vertical coordinates.
projection_x_coordinate
This coordinate variable represents one half of the positional
information for gridded data held on non-Latitude-Longitude projections.
For example, the Met Office uses a Lambert azimuthal equal area (LAEA)
projection for the IMPROVER UK domain.
It has a standard_name attribute set to projection_x_coordinate,
and in the case of the LAEA projection,
the units attribute is set to m.
This must be provided for gridded data
on a non-Latitude-Longitude projection.
For gridded data, if any statistical processing over the coordinate
has been applied,
there must also be an associated projection_x_coordinate_bnds variable
providing the bounds over which cell_methods are applied,
although this is often included anyway to define the cell boundaries.
The projection_x_coordinate_bnds variable has no attributes
as it is tied to the main coordinate variable.
Note
For Met Office data using Lambert azimuthal equal area (LAEA) projection, the coordinate can be considered as relative to ETRS89 or the European Terrestrial Reference System 1989 although this is not explicit in the metadata. The European Terrestrial Reference System 1989 is a a datum based on WGS84, but fixed on 1-Jan-1989 to be anchored to the Eurasian continental plate. This is realised through a TRF (the European Terrestrial Reference Frame or ETRF). ETRS89 is ideal for a Europe-wide consistent mapping and datasets, and is an EU INSPIRE directive standard. In practice, it is close enough to WGS84 to make no difference for most applications of post-processed meteorological data.
projection_y_coordinate
This coordinate variable represents one half of the positional
information for gridded data held on non-Latitude-Longitude projections.
For example, the Met Office uses a Lambert azimuthal equal area (LAEA) grid
for the IMPROVER UK domain.
It has a standard_name attribute set to projection_y_coordinate,
and in the case of the LAEA projection,
the units attribute is set to m.
This must be provided for gridded data
on a non-Latitude-Longitude projection.
For gridded data, if any statistical processing over the coordinate
has been applied,
there must also be an associated projection_y_coordinate_bnds variable
providing the bounds over which cell_methods are applied,
although this is often included anyway to define the cell boundaries.
The projection_y_coordinate_bnds variable has no attributes
as it is tied to the main coordinate variable.
realization
This CF coordinate variable is used for indexing ensemble members
and has the standard_name attribute set to realization.
This is not usually seen in the metadata of IMPROVER output files,
IMPROVER usually generates probabilities of exceedance or percentiles.
However, it will be seen in the input file metadata
and may be seen in the output data cell_methods
where processing has been applied over realizations
(e.g. realization: mean for mean wind direction).
By convention, realization zero is the unperturbed or control member.
source
This CF attribute specifies the method of production of the original data.
This must be present and should take the value of the original source
of the data (typically an NWP model)
when no significant post-processing has been applied.
However, where significant adjustment of the data has occurred
or a number input sources have been blended,
it should be set to IMPROVER.
Often, careful consideration of when it is appropriate to set this
to reference IMPROVER is required to avoid the metadata being misleading.
It is probably not worth including a version of the IMPROVER software,
unless this can be reliably supplied.
spot_index
This IMPROVER-specific dimension is used as an increasing integer value index for sites.
ssp__
This is intended to indicate a SPP (statistical post-processing) namespace. It prefixes atributes to show that they are separate from the CF Metadata Conventions attributes.
ssp__relative_to_threshold
This is an IMPROVER-specific varaible attribute indicating the nature of the threshold inequality for a probability and takes one of the following four values:
greater_thangreater_than_or_equal_toless_thanless_than_or_equal_to
standard_name
This CF attribute provides a descriptive name
from the governed CF Standard Name list.
If no standard_name exists for the quantity,
a long_name must be used.
A standard_name or long_name must be present.
string5 / string8
These IMPROVER-specific arbitary constants are used to dimension the character length of the string variable holding zero padded WMO identifiers and Met Office identifiers, respectively.
threshold
This is an IMPROVER-specific coordinate variable that holds
the set of values of the variable of interest for which the
probability values are generated.
The IMPROVER code uses var_name="threshold" to detect a threshold variable
as a different standard_name or long_name attributes will be set for
different quantities to represent the variable of interest.
The appropriate units for this will also be set.
This must be present for probability variables.
time
This CF Variable provides the time at which the parameter value is valid,
and has a standard_name attribute set to time.
This is an 64-bit integer in units of seconds since 1970-01-01 00:00:00
This must be present.
If any statistical processing over time has been applied
(e.g. accumulation, maxiumum, etc),
there must also be time_bnds variable
providing the time bounds over which cell_methods are applied.
time_bnds has no attributes as it is tied to the main time variable.
title
This netCDF global attribute provides a succinct description of what is in the file and should be something that could be used on a plot to help describe the data.
This must be present, but there is no generally prescribed form that is must take.
units
This netCDF variable attribute provides the units of measurement for the quantity in a string form recognised by the Unidata’s UDUNITS package
This must be present and for IMPROVER this must be SI units,
with the exception that degrees (rather than radians)
are used for wind direction.
Non-dimensional quantities, such as IMPROVER probabilities,
have units set to 1.
weather_code
This IMPROVER variable provides a weather code in the form of an integer value.
It has a long_name attribute set to weather_code
and a units attribute set to 1.
It also has weather_code and weather_code_meaning attributes
which can used to map code values to a short description;
the values use for the Met Office IMPROVER implementation are
shown in the table below.
Code |
Description |
|---|---|
0 |
Clear_Night |
1 |
Sunny_Day |
2 |
Partly_Cloudy_Night |
3 |
Partly_Cloudy_Day |
4 |
Dust |
5 |
Mist |
6 |
Fog |
7 |
Cloudy |
8 |
Overcast |
9 |
Light_Shower_Night |
10 |
Light_Shower_Day |
11 |
Drizzle |
12 |
Light_Rain |
13 |
Heavy_Shower_Night |
14 |
Heavy_Shower_Day |
15 |
Heavy_Rain |
16 |
Sleet_Shower_Night |
17 |
Sleet_Shower_Day |
18 |
Sleet |
19 |
Hail_Shower_Night |
20 |
Hail_Shower_Day |
21 |
Hail |
22 |
Light_Snow_Shower_Night |
23 |
Light_Snow_Shower_Day |
24 |
Light_Snow |
25 |
Heavy_Snow_Shower_Night |
26 |
Heavy_Snow_Shower_Day |
27 |
Heavy_Snow |
28 |
Thunder_Shower_Night |
29 |
Thunder_Shower_Day |
30 |
Thunder |
wmo_id
This IMPROVER-specific coordinate variable
is a 5-character string, zero-padded ID number for WMO sites.
For non-WMO sites it is set to the string None.
It has a long_name attribute set to wmo_id.
This is optional and only relevant for WMO sites.