def process(
orography: cli.inputcube,
land_sea_mask: cli.inputcube = None,
*,
bands_config: cli.inputjson = None,
):
"""Runs topographic bands mask generation.
Reads orography and land_sea_mask fields of a cube. Creates a series of
masks, where each mask excludes data below or equal to the lower threshold
and excludes data above the upper threshold.
Args:
orography (iris.cube.Cube):
The orography on a standard grid.
land_sea_mask (iris.cube.Cube):
The land mask on standard grid, with land points set to one and
sea points set to zero. If provided sea points will be set
to zero in every band. If no land mask is provided, sea points will
be included in the appropriate topographic band.
bands_config (dict):
Definition of orography bands required.
The expected format of the dictionary is e.g
{'bounds':[[0, 50], [50, 200]], 'units': 'm'}
The default dictionary has the following form:
{'bounds': [[-500., 50.], [50., 100.],
[100., 150.],[150., 200.], [200., 250.],
[250., 300.], [300., 400.], [400., 500.],
[500., 650.],[650., 800.], [800., 950.],
[950., 6000.]], 'units': 'm'}
Returns:
iris.cube.Cube:
list of orographic band mask cube.
"""
from improver.generate_ancillaries.generate_ancillary import (
GenerateOrographyBandAncils,
THRESHOLDS_DICT,
)
if bands_config is None:
bands_config = THRESHOLDS_DICT
if land_sea_mask:
land_sea_mask = next(
land_sea_mask.slices(
[land_sea_mask.coord(axis="y"),
land_sea_mask.coord(axis="x")]))
orography = next(
orography.slices(
[orography.coord(axis="y"),
orography.coord(axis="x")]))
result = GenerateOrographyBandAncils()(orography,
bands_config,
landmask=land_sea_mask)
result = result.concatenate_cube()
return result
def process(cube: cli.inputcube):
"""Sets night values to zero for UV index.
Args:
cube (iris.cube.Cube):
Cube that will have night values set to zero. This should contain
either diagnostic values or probabilities of UV index above threshold.
Returns:
iris.cube.Cube:
Input cube with all night values set to zero.
Raises:
ValueError: If input cube is suspicious, within reason. Note that this is
a general check: the CLI expects a cube of UV index or probability of
UV index above thresold, and will raise an error if given a probability
below threshold, but will not recognise a completely inappropriate cube
(eg temperature in Kelvin). Therefore this CLI should be used with care.
"""
import numpy as np
from improver.metadata.probabilistic import is_probability
from improver.utilities.solar import DayNightMask
if is_probability(cube):
if "above_threshold" not in cube.name():
raise ValueError(f"{cube.name()} unsuitable for night masking")
mask = DayNightMask()(cube).data
# Broadcast mask to shape of input cube to account for additional dimensions.
mask = np.broadcast_to(mask, cube.shape)
# setting night values to zero.
cube.data = np.where(mask == DayNightMask().night, 0, cube.data)
return cube
def process(cube: cli.inputcube,
raw_cube: cli.inputcube = None,
*,
realizations_count: int = None,
random_seed: int = None,
ignore_ecc_bounds=False):
"""Converts an incoming cube into one containing realizations.
Args:
cube (iris.cube.Cube):
A cube to be processed.
raw_cube (iris.cube.Cube):
Cube of raw (not post processed) weather data.
If this argument is given ensemble realizations will be created
from percentiles by reshuffling them in correspondence to the rank
order of the raw ensemble. Otherwise, the percentiles are rebadged
as realizations.
realizations_count (int):
The number of ensemble realizations in the output.
random_seed (int):
Option to specify a value for the random seed for testing
purposes, otherwise the default random seed behaviours is
utilised. The random seed is used in the generation of the
random numbers used for splitting tied values within the raw
ensemble, so that the values from the input percentiles can
be ordered to match the raw ensemble.
ignore_ecc_bounds (bool):
If True, where percentiles exceed the ECC bounds range, raises a
warning rather than an exception.
Returns:
iris.cube.Cube:
The processed cube.
"""
from improver.cli import (percentiles_to_realizations,
probabilities_to_realizations)
if cube.coords('percentile'):
output_cube = percentiles_to_realizations.process(
cube,
raw_cube=raw_cube,
realizations_count=realizations_count,
random_seed=random_seed,
ignore_ecc_bounds=ignore_ecc_bounds)
elif cube.coords(var_name='threshold'):
output_cube = probabilities_to_realizations.process(
cube,
raw_cube=raw_cube,
realizations_count=realizations_count,
random_seed=random_seed,
ignore_ecc_bounds=ignore_ecc_bounds)
elif cube.coords(var_name='realization'):
output_cube = cube
else:
raise ValueError("Unable to convert to realizations:\n" + str(cube))
return output_cube
def process(orography: cli.inputcube,
land_sea_mask: cli.inputcube = None,
*,
bands_config: cli.inputjson = None):
"""Runs topographic weights generation.
Reads the orography and land_sea_mask fields of a cube. Creates a series of
topographic zone weights to indicate where an orography point sits within
the defined topographic bands. If the orography point is in the centre of
a topographic band, then a single band will have a weight 1.0.
If the orography point is at the edge of a topographic band, then the
upper band will have a 0.5 weight whilst the lower band will also have a
0.5 weight. Otherwise the weight will vary linearly between the centre of
a topographic band and the edge.
Args:
orography (iris.cube.Cube):
The orography on a standard grid.
land_sea_mask (iris.cube.Cube):
Land mask on a standard grid. If provided, sea points will be
masked and set to the default fill value. If no land mask is
provided, weights will be generated for sea points as well as land
in the appropriate topographic band.
bands_config (dict):
Definition of orography bands required.
The expected format of the dictionary is e.g
{'bounds':[[0, 50], [50, 200]], 'units': 'm'}
The default dictionary has the following form:
{'bounds': [[-500., 50.], [50., 100.],
[100., 150.],[150., 200.], [200., 250.],
[250., 300.], [300., 400.], [400., 500.],
[500., 650.],[650., 800.], [800., 950.],
[950., 6000.]], 'units': 'm'}
Returns:
iris.cube.Cube:
Cube containing the weights depending upon where the orography
point is within the topographical zones.
"""
from improver.generate_ancillaries.generate_topographic_zone_weights \
import GenerateTopographicZoneWeights
from improver.generate_ancillaries.generate_ancillary import (
THRESHOLDS_DICT)
if bands_config is None:
bands_config = THRESHOLDS_DICT
if land_sea_mask:
land_sea_mask = next(land_sea_mask.slices(
[land_sea_mask.coord(axis='y'), land_sea_mask.coord(axis='x')]))
orography = next(orography.slices(
[orography.coord(axis='y'), orography.coord(axis='x')]))
result = GenerateTopographicZoneWeights().process(
orography, bands_config, landmask=land_sea_mask)
return result
def process(cube: cli.inputcube,
*,
realizations_count: int = None,
ignore_ecc_bounds=False):
"""Converts an incoming cube into one containing realizations.
Args:
cube (iris.cube.Cube):
A cube to be processed.
realizations_count (int):
The number of ensemble realizations in the output.
ignore_ecc_bounds (bool):
If True, where percentiles exceed the ECC bounds range, raises a
warning rather than an exception.
Returns:
iris.cube.Cube:
The processed cube.
"""
from improver.cli import (percentiles_to_realizations,
probabilities_to_realizations)
if cube.coords('percentile'):
output_cube = percentiles_to_realizations.process(
cube,
realizations_count=realizations_count,
ignore_ecc_bounds=ignore_ecc_bounds)
elif cube.coords(var_name='threshold'):
output_cube = probabilities_to_realizations.process(
cube,
realizations_count=realizations_count,
ignore_ecc_bounds=ignore_ecc_bounds)
elif cube.coords(var_name='realization'):
output_cube = cube
else:
raise ValueError("Unable to convert to realizations:\n" + str(cube))
return output_cube
def process(cube: cli.inputcube):
"""Sets night values to zero for UV index.
Args:
cube (iris.cube.Cube):
Cube that will have night values set to zero.
Returns:
iris.cube.Cube:
Input cube with all night values set to zero.
"""
import numpy as np
from improver.utilities.solar import DayNightMask
mask = DayNightMask()(cube).data
# Broadcast mask to shape of input cube to account for additional dimensions.
mask = np.broadcast_to(mask, cube.shape)
# setting night values to zero.
cube.data = np.where(mask == DayNightMask().night, 0, cube.data)
return cube
def process(
neighbour_cube: cli.inputcube,
cube: cli.inputcube,
lapse_rate: cli.inputcube = None,
*,
apply_lapse_rate_correction=False,
land_constraint=False,
similar_altitude=False,
extract_percentiles: cli.comma_separated_list = None,
ignore_ecc_bounds=False,
new_title: str = None,
suppress_warnings=False,
):
"""Module to run spot data extraction.
Extract diagnostic data from gridded fields for spot data sites. It is
possible to apply a temperature lapse rate adjustment to temperature data
that helps to account for differences between the spot site's real altitude
and that of the grid point from which the temperature data is extracted.
Args:
neighbour_cube (iris.cube.Cube):
Cube of spot-data neighbours and the spot site information.
cube (iris.cube.Cube):
Cube containing the diagnostic data to be extracted.
lapse_rate (iris.cube.Cube):
Optional cube containing temperature lapse rates. If this cube is
provided and a screen temperature cube is being processed, the
lapse rates will be used to adjust the temperature to better
represent each spot's site-altitude.
apply_lapse_rate_correction (bool):
Use to apply a lapse-rate correction to screen temperature data so
that the data are a better match the altitude of the spot site for
which they have been extracted.
land_constraint (bool):
Use to select the nearest-with-land-constraint neighbour-selection
method from the neighbour_cube. This means that the grid points
should be land points except for sites where none were found within
the search radius when the neighbour cube was created. May be used
with similar_altitude.
similar_altitude (bool):
Use to select the nearest-with-height-constraint
neighbour-selection method from the neighbour_cube. These are grid
points that were found to be the closest in altitude to the spot
site within the search radius defined when the neighbour cube was
created. May be used with land_constraint.
extract_percentiles (list or int):
If set to a percentile value or a list of percentile values,
data corresponding to those percentiles will be returned. For
example "25, 50, 75" will result in the 25th, 50th and 75th
percentiles being returned from a cube of probabilities,
percentiles or realizations. Deterministic input data will raise
a warning message.
Note that for percentiles inputs, the desired percentile(s) must
exist in the input cube.
ignore_ecc_bounds (bool):
Demotes exceptions where calculated percentiles are outside the ECC
bounds range to warnings.
new_title (str):
New title for the spot-extracted data. If None, this attribute is
removed from the output cube since it has no prescribed standard
and may therefore contain grid information that is no longer
correct after spot-extraction.
suppress_warnings (bool):
Suppress warning output. This option should only be used if it
is known that warnings will be generated but they are not required.
Returns:
iris.cube.Cube:
Cube of spot data.
Raises:
ValueError:
If the percentile diagnostic cube does not contain the requested
percentile value.
ValueError:
If the lapse rate cube was provided but the diagnostic being
processed is not air temperature.
ValueError:
If the lapse rate cube provided does not have the name
"air_temperature_lapse_rate"
ValueError:
If the lapse rate cube does not contain a single valued height
coordinate.
Warns:
warning:
If diagnostic cube is not a known probabilistic type.
warning:
If a lapse rate cube was provided, but the height of the
temperature does not match that of the data used.
warning:
If a lapse rate cube was not provided, but the option to apply
the lapse rate correction was enabled.
"""
import warnings
import iris
import numpy as np
from iris.exceptions import CoordinateNotFoundError
from improver.ensemble_copula_coupling.ensemble_copula_coupling import (
ConvertProbabilitiesToPercentiles, )
from improver.metadata.probabilistic import find_percentile_coordinate
from improver.percentile import PercentileConverter
from improver.spotdata.apply_lapse_rate import SpotLapseRateAdjust
from improver.spotdata.neighbour_finding import NeighbourSelection
from improver.spotdata.spot_extraction import SpotExtraction
from improver.utilities.cube_extraction import extract_subcube
neighbour_selection_method = NeighbourSelection(
land_constraint=land_constraint,
minimum_dz=similar_altitude).neighbour_finding_method_name()
result = SpotExtraction(
neighbour_selection_method=neighbour_selection_method)(
neighbour_cube, cube, new_title=new_title)
# If a probability or percentile diagnostic cube is provided, extract
# the given percentile if available. This is done after the spot-extraction
# to minimise processing time; usually there are far fewer spot sites than
# grid points.
if extract_percentiles:
extract_percentiles = [np.float32(x) for x in extract_percentiles]
try:
perc_coordinate = find_percentile_coordinate(result)
except CoordinateNotFoundError:
if "probability_of_" in result.name():
result = ConvertProbabilitiesToPercentiles(
ecc_bounds_warning=ignore_ecc_bounds)(
result, percentiles=extract_percentiles)
result = iris.util.squeeze(result)
elif result.coords("realization", dim_coords=True):
fast_percentile_method = not np.ma.isMaskedArray(result.data)
result = PercentileConverter(
"realization",
percentiles=extract_percentiles,
fast_percentile_method=fast_percentile_method,
)(result)
else:
msg = ("Diagnostic cube is not a known probabilistic type. "
"The {} percentile could not be extracted. Extracting "
"data from the cube including any leading "
"dimensions.".format(extract_percentiles))
if not suppress_warnings:
warnings.warn(msg)
else:
constraint = [
"{}={}".format(perc_coordinate.name(), extract_percentiles)
]
perc_result = extract_subcube(result, constraint)
if perc_result is not None:
result = perc_result
else:
msg = ("The percentile diagnostic cube does not contain the "
"requested percentile value. Requested {}, available "
"{}".format(extract_percentiles,
perc_coordinate.points))
raise ValueError(msg)
# Check whether a lapse rate cube has been provided and we are dealing with
# temperature data and the lapse-rate option is enabled.
if apply_lapse_rate_correction and lapse_rate:
if not result.name() == "air_temperature":
msg = ("A lapse rate cube was provided, but the diagnostic being "
"processed is not air temperature and cannot be adjusted.")
raise ValueError(msg)
if not lapse_rate.name() == "air_temperature_lapse_rate":
msg = ("A cube has been provided as a lapse rate cube but does "
"not have the expected name air_temperature_lapse_rate: "
"{}".format(lapse_rate.name()))
raise ValueError(msg)
try:
lapse_rate_height_coord = lapse_rate.coord("height")
except (ValueError, CoordinateNotFoundError):
msg = ("Lapse rate cube does not contain a single valued height "
"coordinate. This is required to ensure it is applied to "
"equivalent temperature data.")
raise ValueError(msg)
# Check the height of the temperature data matches that used to
# calculate the lapse rates. If so, adjust temperatures using the lapse
# rate values.
if cube.coord("height") == lapse_rate_height_coord:
plugin = SpotLapseRateAdjust(
neighbour_selection_method=neighbour_selection_method)
result = plugin(result, neighbour_cube, lapse_rate)
elif not suppress_warnings:
warnings.warn(
"A lapse rate cube was provided, but the height of the "
"temperature data does not match that of the data used "
"to calculate the lapse rates. As such the temperatures "
"were not adjusted with the lapse rates.")
elif apply_lapse_rate_correction and not lapse_rate:
if not suppress_warnings:
warnings.warn(
"A lapse rate cube was not provided, but the option to "
"apply the lapse rate correction was enabled. No lapse rate "
"correction could be applied.")
# Remove the internal model_grid_hash attribute if present.
result.attributes.pop("model_grid_hash", None)
return result
def process(
wind_speed: cli.inputcube,
sigma: cli.inputcube,
target_orography: cli.inputcube,
standard_orography: cli.inputcube,
silhouette_roughness: cli.inputcube,
vegetative_roughness: cli.inputcube = None,
*,
model_resolution: float,
output_height_level: float = None,
output_height_level_units="m",
):
"""Wind downscaling.
Run wind downscaling to apply roughness correction and height correction
to wind fields as described in Howard and Clark (2007). All inputs must
be on the same standard grid.
Args:
wind_speed (iris.cube.Cube):
Cube of wind speed on standard grid.
Any units can be supplied.
sigma (iris.cube.Cube):
Cube of standard deviation of model orography height.
Units of field: m.
target_orography (iris.cube.Cube):
Cube of orography to downscale fields to.
Units of field: m.
standard_orography (iris.cube.Cube):
Cube of orography on standard grid. (interpolated model orography).
Units of field: m.
silhouette_roughness (iris.cube.Cube):
Cube of model silhouette roughness.
Units of field: dimensionless.
vegetative_roughness (iris.cube.Cube):
Cube of vegetative roughness length.
Units of field: m.
model_resolution (float):
Original resolution of model orography (before interpolation to
standard grid)
Units of field: m.
output_height_level (float):
If only a single height level is desired as output from
wind-downscaling, this option can be used to select the height
level. If no units are provided with 'output_height_level_units',
metres are assumed.
output_height_level_units (str):
If a single height level is selected as output using
'output_height_level', this additional argument may be used to
specify the units of the value entered to select the level.
e.g hPa.
Returns:
iris.cube.Cube:
The processed Cube.
Rises:
ValueError:
If the requested height value is not found.
"""
import warnings
import iris
from iris.exceptions import CoordinateNotFoundError
from improver.utilities.cube_extraction import apply_extraction
from improver.wind_calculations import wind_downscaling
if output_height_level_units and output_height_level is None:
warnings.warn(
"output_height_level_units has been set but no "
"associated height level has been provided. These units "
"will have no effect."
)
try:
wind_speed_iterator = wind_speed.slices_over("realization")
except CoordinateNotFoundError:
wind_speed_iterator = [wind_speed]
wind_speed_list = iris.cube.CubeList()
for wind_speed_slice in wind_speed_iterator:
result = wind_downscaling.RoughnessCorrection(
silhouette_roughness,
sigma,
target_orography,
standard_orography,
model_resolution,
z0_cube=vegetative_roughness,
height_levels_cube=None,
)(wind_speed_slice)
wind_speed_list.append(result)
wind_speed = wind_speed_list.merge_cube()
non_dim_coords = [x.name() for x in wind_speed.coords(dim_coords=False)]
if "realization" in non_dim_coords:
wind_speed = iris.util.new_axis(wind_speed, "realization")
if output_height_level is not None:
constraints = {"height": output_height_level}
units = {"height": output_height_level_units}
single_level = apply_extraction(
wind_speed, iris.Constraint(**constraints), units
)
if not single_level:
raise ValueError(
"Requested height level not found, no cube "
"returned. Available height levels are:\n"
"{0:}\nin units of {1:}".format(
wind_speed.coord("height").points, wind_speed.coord("height").units
)
)
wind_speed = single_level
return wind_speed
def process(
cube: cli.inputcube,
*,
coordinates: cli.comma_separated_list = None,
percentiles: cli.comma_separated_list = None,
ignore_ecc_bounds=False,
):
r"""Collapses cube coordinates and calculate percentiled data.
Calculate percentiled data over a given coordinate by collapsing that
coordinate. Typically used to convert realization data into percentiled
data, but may calculate over any dimension coordinate. Alternatively
calling this with a dataset containing probabilities will convert those
to percentiles using the ensemble coupla coupling plugin. If no particular
percentiles are given at which to calculate values and no
'number of percentiles' to calculate are specified, the
following defaults will be used.
'[0, 5, 10, 20, 25, 30, 40, 50, 60, 70, 75, 80, 90, 95, 100]'
Args:
cube (iris.cube.Cube):
A Cube for processing.
coordinates (str or list):
Coordinate or coordinates over which to collapse data and
calculate percentiles; e.g. 'realization' or 'latitude,longitude'.
This argument must be provided when collapsing a coordinate or
coordinates to create percentiles, but is redundant when
converting probabilities to percentiles and may be omitted. This
coordinate(s) will be removed and replaced by a percentile
coordinate.
percentiles (list):
Optional definition of percentiles at which to calculate data.
ignore_ecc_bounds (bool):
If True, where calculated percentiles are outside the ECC bounds
range, raises a warning rather than an exception.
Returns:
iris.cube.Cube:
The processed Cube.
Raises:
ValueError:
If the cube name does not contain 'probability_of\_' and
coordinates isn't used.
Warns:
Warning:
If 'probability_of\_' is in the cube name and coordinates is used.
"""
import warnings
import numpy as np
from improver.ensemble_copula_coupling.ensemble_copula_coupling import (
ConvertProbabilitiesToPercentiles, )
from improver.metadata.probabilistic import is_probability
from improver.percentile import PercentileConverter
if percentiles is not None:
percentiles = [float(p) for p in percentiles]
if is_probability(cube):
result = ConvertProbabilitiesToPercentiles(
ecc_bounds_warning=ignore_ecc_bounds)(cube,
percentiles=percentiles)
if coordinates:
warnings.warn("Converting probabilities to percentiles. The "
"provided COORDINATES_TO_COLLAPSE variable will "
"not be used.")
else:
if not coordinates:
raise ValueError("To collapse a coordinate to calculate "
"percentiles, a coordinate or list of "
"coordinates must be provided.")
# Switch back to use the slow scipy method if the cube contains masked
# data which the numpy method cannot handle.
fast_percentile_method = True
if np.ma.is_masked(cube.data):
# Check for masked points:
fast_percentile_method = False
elif np.ma.isMaskedArray(cube.data):
# Check if we have a masked array with an empty mask. If so,
# replace it with a non-masked array:
cube.data = cube.data.data
result = PercentileConverter(
coordinates,
percentiles=percentiles,
fast_percentile_method=fast_percentile_method,
)(cube)
return result
def process(cube: cli.inputcube,
coefficients: cli.inputcube = None,
land_sea_mask: cli.inputcube = None,
*,
distribution,
realizations_count: int = None,
randomise=False,
random_seed: int = None,
ignore_ecc_bounds=False,
predictor='mean',
shape_parameters: cli.comma_separated_list = None):
"""Applying coefficients for Ensemble Model Output Statistics.
Load in arguments for applying coefficients for Ensemble Model Output
Statistics (EMOS), otherwise known as Non-homogeneous Gaussian
Regression (NGR). The coefficients are applied to the forecast
that is supplied, so as to calibrate the forecast. The calibrated
forecast is written to a cube. If no coefficients are provided the input
forecast is returned unchanged.
Args:
cube (iris.cube.Cube):
A Cube containing the forecast to be calibrated. The input format
could be either realizations, probabilities or percentiles.
coefficients (iris.cube.Cube):
A cube containing the coefficients used for calibration or None.
If none then then input is returned unchanged.
land_sea_mask (iris.cube.Cube):
A cube containing the land-sea mask on the same domain as the
forecast that is to be calibrated. Land points are "
"specified by ones and sea points are specified by zeros. "
"If not None this argument will enable land-only calibration, in "
"which sea points are returned without the application of "
"calibration."
distribution (str):
The distribution for constructing realizations, percentiles or
probabilities. This should typically match the distribution used
for minimising the Continuous Ranked Probability Score when
estimating the EMOS coefficients. The distributions available are
those supported by :data:`scipy.stats`.
realizations_count (int):
Option to specify the number of ensemble realizations that will be
created from probabilities or percentiles for input into EMOS.
randomise (bool):
Option to reorder the post-processed forecasts randomly. If not
set, the ordering of the raw ensemble is used. This option is
only valid when the input format is realizations.
random_seed (int):
Option to specify a value for the random seed for testing
purposes, otherwise the default random seen behaviour is utilised.
The random seed is used in the generation of the random numbers
used for either the randomise option to order the input
percentiles randomly, rather than use the ordering from the raw
ensemble, or for splitting tied values within the raw ensemble,
so that the values from the input percentiles can be ordered to
match the raw ensemble.
ignore_ecc_bounds (bool):
If True, where the percentiles exceed the ECC bounds range,
raises a warning rather than an exception. This occurs when the
current forecasts is in the form of probabilities and is
converted to percentiles, as part of converting the input
probabilities into realizations.
predictor (str):
String to specify the form of the predictor used to calculate
the location parameter when estimating the EMOS coefficients.
Currently the ensemble mean ("mean") and the ensemble
realizations ("realizations") are supported as the predictors.
shape_parameters (float or str):
The shape parameters required for defining the distribution
specified by the distribution argument. The shape parameters
should either be a number or 'inf' or '-inf' to represent
infinity. Further details about appropriate shape parameters
are available in scipy.stats. For the truncated normal
distribution with a lower bound of zero, as available when
estimating EMOS coefficients, the appropriate shape parameters
are 0 and inf.
Returns:
iris.cube.Cube:
The calibrated forecast cube.
Raises:
ValueError:
If the current forecast is a coefficients cube.
ValueError:
If the coefficients cube does not have the right name of
"emos_coefficients".
ValueError:
If the forecast type is 'percentiles' or 'probabilities' and the
realizations_count argument is not provided.
"""
import warnings
import numpy as np
from iris.exceptions import CoordinateNotFoundError
from improver.calibration.ensemble_calibration import (
ApplyCoefficientsFromEnsembleCalibration)
from improver.ensemble_copula_coupling.ensemble_copula_coupling import (
EnsembleReordering, ConvertLocationAndScaleParametersToPercentiles,
ConvertLocationAndScaleParametersToProbabilities,
ConvertProbabilitiesToPercentiles, RebadgePercentilesAsRealizations,
ResamplePercentiles)
from improver.calibration.utilities import merge_land_and_sea
from improver.metadata.probabilistic import find_percentile_coordinate
current_forecast = cube
if current_forecast.name() in ['emos_coefficients', 'land_binary_mask']:
msg = "The current forecast cube has the name {}"
raise ValueError(msg.format(current_forecast.name()))
if coefficients is None:
msg = ("There are no coefficients provided for calibration. The "
"uncalibrated forecast will be returned.")
warnings.warn(msg)
return current_forecast
if coefficients.name() != 'emos_coefficients':
msg = ("The current coefficients cube does not have the "
"name 'emos_coefficients'")
raise ValueError(msg)
if land_sea_mask and land_sea_mask.name() != 'land_binary_mask':
msg = ("The land_sea_mask cube does not have the "
"name 'land_binary_mask'")
raise ValueError(msg)
original_current_forecast = current_forecast.copy()
try:
find_percentile_coordinate(current_forecast)
input_forecast_type = "percentiles"
except CoordinateNotFoundError:
input_forecast_type = "realizations"
if current_forecast.name().startswith("probability_of"):
input_forecast_type = "probabilities"
conversion_plugin = ConvertProbabilitiesToPercentiles(
ecc_bounds_warning=ignore_ecc_bounds)
elif input_forecast_type == "percentiles":
# Initialise plugin to resample percentiles so that the percentiles are
# evenly spaced.
conversion_plugin = ResamplePercentiles(
ecc_bounds_warning=ignore_ecc_bounds)
if input_forecast_type in ["percentiles", "probabilities"]:
if not realizations_count:
raise ValueError(
"The current forecast has been provided as {0}. "
"These {0} need to be converted to realizations "
"for ensemble calibration. The realizations_count "
"argument is used to define the number of realizations "
"to construct from the input {0}, so if the "
"current forecast is provided as {0} then "
"realizations_count must be defined.".format(
input_forecast_type))
current_forecast = conversion_plugin.process(
current_forecast, no_of_percentiles=realizations_count)
current_forecast = (
RebadgePercentilesAsRealizations().process(current_forecast))
# Apply coefficients as part of Ensemble Model Output Statistics (EMOS).
ac = ApplyCoefficientsFromEnsembleCalibration(predictor=predictor)
location_parameter, scale_parameter = ac.process(
current_forecast, coefficients, landsea_mask=land_sea_mask)
if shape_parameters:
shape_parameters = [np.float32(x) for x in shape_parameters]
# Convert the output forecast type (i.e. realizations, percentiles,
# probabilities) to match the input forecast type.
if input_forecast_type == "probabilities":
result = ConvertLocationAndScaleParametersToProbabilities(
distribution=distribution,
shape_parameters=shape_parameters).process(
location_parameter, scale_parameter, original_current_forecast)
elif input_forecast_type == "percentiles":
perc_coord = find_percentile_coordinate(original_current_forecast)
result = ConvertLocationAndScaleParametersToPercentiles(
distribution=distribution,
shape_parameters=shape_parameters).process(
location_parameter,
scale_parameter,
original_current_forecast,
percentiles=perc_coord.points)
elif input_forecast_type == "realizations":
# Ensemble Copula Coupling to generate realizations
# from the location and scale parameter.
no_of_percentiles = len(current_forecast.coord('realization').points)
percentiles = ConvertLocationAndScaleParametersToPercentiles(
distribution=distribution,
shape_parameters=shape_parameters).process(
location_parameter,
scale_parameter,
original_current_forecast,
no_of_percentiles=no_of_percentiles)
result = EnsembleReordering().process(percentiles,
current_forecast,
random_ordering=randomise,
random_seed=random_seed)
if land_sea_mask:
# Fill in masked sea points with uncalibrated data.
merge_land_and_sea(result, original_current_forecast)
return result
def process(
cube: cli.inputcube,
mask: cli.inputcube = None,
*,
neighbourhood_output,
neighbourhood_shape="square",
radii: cli.comma_separated_list,
lead_times: cli.comma_separated_list = None,
degrees_as_complex=False,
weighted_mode=False,
area_sum=False,
percentiles: cli.comma_separated_list = DEFAULT_PERCENTILES,
halo_radius: float = None,
):
"""Runs neighbourhood processing.
Apply the requested neighbourhood method via the
NeighbourhoodProcessing plugin to a Cube.
Args:
cube (iris.cube.Cube):
The Cube to be processed.
mask (iris.cube.Cube):
A cube to mask the input cube. The data should contain 1 for
usable points and 0 for discarded points.
Can't be used with "percentiles" as neighbourhood_output (Optional)
neighbourhood_output (str):
The form of the results generated using neighbourhood processing.
If "probabilities" is selected, the mean probability with a
neighbourhood is calculated. If "percentiles" is selected, then
the percentiles are calculated with a neighbourhood. Calculating
percentiles from a neighbourhood is only supported for a circular
neighbourhood.
Options: "probabilities", "percentiles".
neighbourhood_shape (str):
Name of the neighbourhood method to use. Only a "circular"
neighbourhood shape is applicable for calculating "percentiles"
output.
Options: "circular", "square".
Default: "square".
radii (list of float):
The radius or a list of radii in metres of the neighbourhood to
apply.
If it is a list, it must be the same length as lead_times, which
defines at which lead time to use which nbhood radius. The radius
will be interpolated for intermediate lead times.
lead_times (list of int):
The lead times in hours that correspond to the radii to be used.
If lead_times are set, radii must be a list the same length as
lead_times.
degrees_as_complex (bool):
Include this option to process angles as complex numbers.
Not compatible with circular kernel or percentiles.
weighted_mode (bool):
Include this option to set the weighting to decrease with radius.
Otherwise a constant weighting is assumed.
weighted_mode is only applicable for calculating "probability"
neighbourhood output using the circular kernel.
area_sum (bool):
Return sum rather than fraction over the neighbourhood area.
percentiles (float):
Calculates value at the specified percentiles from the
neighbourhood surrounding each grid point. This argument has no
effect if the output is probabilities.
halo_radius (float):
Set this radius in metres to define the excess halo to clip. Used
where a larger grid was defined than the standard grid and we want
to clip the grid back to the standard grid. Otherwise no clipping
is applied.
Returns:
iris.cube.Cube:
A processed Cube.
Raises:
RuntimeError:
If weighted_mode is used with the wrong neighbourhood_output.
RuntimeError:
If degree_as_complex is used with
neighbourhood_output='percentiles'.
RuntimeError:
If degree_as_complex is used with neighbourhood_shape='circular'.
"""
from improver.nbhood import radius_by_lead_time
from improver.nbhood.nbhood import (
GeneratePercentilesFromANeighbourhood,
NeighbourhoodProcessing,
)
from improver.utilities.pad_spatial import remove_cube_halo
from improver.wind_calculations.wind_direction import WindDirection
if neighbourhood_output == "percentiles":
if weighted_mode:
raise RuntimeError("weighted_mode cannot be used with"
'neighbourhood_output="percentiles"')
if degrees_as_complex:
raise RuntimeError("Cannot generate percentiles from complex "
"numbers")
if neighbourhood_shape == "circular":
if degrees_as_complex:
raise RuntimeError(
"Cannot process complex numbers with circular neighbourhoods")
if degrees_as_complex:
# convert cube data into complex numbers
cube.data = WindDirection.deg_to_complex(cube.data)
radius_or_radii, lead_times = radius_by_lead_time(radii, lead_times)
if neighbourhood_output == "probabilities":
result = NeighbourhoodProcessing(
neighbourhood_shape,
radius_or_radii,
lead_times=lead_times,
weighted_mode=weighted_mode,
sum_only=area_sum,
re_mask=True,
)(cube, mask_cube=mask)
elif neighbourhood_output == "percentiles":
result = GeneratePercentilesFromANeighbourhood(
radius_or_radii,
lead_times=lead_times,
percentiles=percentiles,
)(cube)
if degrees_as_complex:
# convert neighbourhooded cube back to degrees
result.data = WindDirection.complex_to_deg(result.data)
if halo_radius is not None:
result = remove_cube_halo(result, halo_radius)
return result
def process(
cube: cli.inputcube,
coefficients: inputcoeffs = None,
land_sea_mask: cli.inputcube = None,
*,
realizations_count: int = None,
randomise=False,
random_seed: int = None,
ignore_ecc_bounds=False,
predictor="mean",
):
"""Applying coefficients for Ensemble Model Output Statistics.
Load in arguments for applying coefficients for Ensemble Model Output
Statistics (EMOS), otherwise known as Non-homogeneous Gaussian
Regression (NGR). The coefficients are applied to the forecast
that is supplied, so as to calibrate the forecast. The calibrated
forecast is written to a cube. If no coefficients are provided the input
forecast is returned unchanged.
Args:
cube (iris.cube.Cube):
A Cube containing the forecast to be calibrated. The input format
could be either realizations, probabilities or percentiles.
coefficients (iris.cube.CubeList):
A cubelist containing the coefficients used for calibration or None.
If none then then input is returned unchanged.
land_sea_mask (iris.cube.Cube):
A cube containing the land-sea mask on the same domain as the
forecast that is to be calibrated. Land points are "
"specified by ones and sea points are specified by zeros. "
"If not None this argument will enable land-only calibration, in "
"which sea points are returned without the application of "
"calibration."
realizations_count (int):
Option to specify the number of ensemble realizations that will be
created from probabilities or percentiles when applying the EMOS
coefficients.
randomise (bool):
Option to reorder the post-processed forecasts randomly. If not
set, the ordering of the raw ensemble is used. This option is
only valid when the input format is realizations.
random_seed (int):
Option to specify a value for the random seed for testing
purposes, otherwise the default random seen behaviour is utilised.
The random seed is used in the generation of the random numbers
used for either the randomise option to order the input
percentiles randomly, rather than use the ordering from the raw
ensemble, or for splitting tied values within the raw ensemble,
so that the values from the input percentiles can be ordered to
match the raw ensemble.
ignore_ecc_bounds (bool):
If True, where the percentiles exceed the ECC bounds range,
raises a warning rather than an exception. This occurs when the
current forecasts is in the form of probabilities and is
converted to percentiles, as part of converting the input
probabilities into realizations.
predictor (str):
String to specify the form of the predictor used to calculate
the location parameter when estimating the EMOS coefficients.
Currently the ensemble mean ("mean") and the ensemble
realizations ("realizations") are supported as the predictors.
Returns:
iris.cube.Cube:
The calibrated forecast cube.
Raises:
ValueError:
If the current forecast is a coefficients cube.
ValueError:
If the coefficients cube does not have the right name of
"emos_coefficients".
ValueError:
If the forecast type is 'percentiles' or 'probabilities' and the
realizations_count argument is not provided.
"""
import warnings
from improver.calibration.ensemble_calibration import ApplyEMOS
if coefficients is None:
msg = ("There are no coefficients provided for calibration. The "
"uncalibrated forecast will be returned.")
warnings.warn(msg)
return cube
if land_sea_mask and land_sea_mask.name() != "land_binary_mask":
msg = "The land_sea_mask cube does not have the name 'land_binary_mask'"
raise ValueError(msg)
calibration_plugin = ApplyEMOS()
result = calibration_plugin(
cube,
coefficients,
land_sea_mask=land_sea_mask,
realizations_count=realizations_count,
ignore_ecc_bounds=ignore_ecc_bounds,
predictor=predictor,
randomise=randomise,
random_seed=random_seed,
)
return result
def process(cube: cli.inputcube,
raw_cube: cli.inputcube = None,
*,
realizations_count: int = None,
random_seed: int = None,
ignore_ecc_bounds=False):
"""Convert probabilities to ensemble realizations using Ensemble Copula
Coupling.
Probabilities are first converted to percentiles, which are then either
rebadged as realizations or reordered if the raw_cube argument is given.
Args:
cube (iris.cube.Cube):
Cube to be processed.
raw_cube (iris.cube.Cube):
Cube of raw (not post processed) weather data.
If this argument is given ensemble realizations will be created
from percentiles by reshuffling them in correspondence to the rank
order of the raw ensemble. Otherwise, the percentiles are rebadged
as realizations.
realizations_count (int):
Optional definition of the number of ensemble realizations to
be generated. These are generated though an intermediate
percentile representation. Theses percentiles will be
distributed regularly with the aim of dividing into blocks
of equal probability. If the raw_cube is given
and the number of realization is not given the number
of realizations is taken from the number of realizations
in the raw_cube.
random_seed (int):
Option to specify a value for the random seed for testing
purposes, otherwise the default random seed behaviours is
utilised. The random seed is used in the generation of the
random numbers used for splitting tied values within the raw
ensemble, so that the values from the input percentiles can
be ordered to match the raw ensemble.
ignore_ecc_bounds (bool):
If True, where percentiles (calculated as an intermediate output
before realization) exceed to ECC bounds range, raises a warning
rather than an exception.
Returns:
iris.cube.Cube:
Processed result Cube.
"""
from improver.ensemble_copula_coupling.ensemble_copula_coupling import (
ConvertProbabilitiesToPercentiles, RebadgePercentilesAsRealizations,
EnsembleReordering)
if realizations_count is None and raw_cube:
# If realizations_count is not given, take the number from the raw
# ensemble cube.
realizations_count = len(raw_cube.coord("realization").points)
result = ConvertProbabilitiesToPercentiles(
ecc_bounds_warning=ignore_ecc_bounds)(
cube, no_of_percentiles=realizations_count)
if raw_cube:
result = EnsembleReordering()(result,
raw_cube,
random_ordering=False,
random_seed=random_seed)
else:
result = RebadgePercentilesAsRealizations()(result)
return result
def process(
start_cube: cli.inputcube,
end_cube: cli.inputcube,
*,
interval_in_mins: int = None,
times: cli.comma_separated_list = None,
interpolation_method="linear",
):
"""Interpolate data between validity times.
Interpolate data to intermediate times between the validity times of two
cubes. This can be used to fill in missing data (e.g. for radar fields)
or to ensure data is available at the required intervals when model data
is not available at these times.
Args:
start_cube (iris.cube.Cube):
Cube containing the data at the beginning.
end_cube (iris.cube.Cube):
Cube containing the data at the end.
interval_in_mins (int):
Specifies the interval in minutes at which to interpolate between
the two input cubes.
A number of minutes which does not divide up the interval equally
will raise an exception.
If intervals_in_mins is set then times can not be used.
times (str):
Specifies the times in the format {YYYYMMDD}T{HHMM}Z
at which to interpolate between the two input cubes.
Where {YYYYMMDD} is year, month, day and {HHMM} is hour and minutes
e.g 20180116T0100Z. More than one time can be provided separated
by a comma.
If times are set, interval_in_mins can not be used.
interpolation_method (str):
["linear", "solar", "daynight"]
Specifies the interpolation method;
solar interpolates using the solar elevation,
daynight uses linear interpolation but sets night time points to
0.0 linear is linear interpolation.
Returns:
iris.cube.CubeList:
A list of cubes interpolated to the desired times. The
interpolated cubes will always be in chronological order of
earliest to latest regardless of the order of the input.
"""
from improver.utilities.cube_manipulation import MergeCubes
from improver.utilities.temporal import cycletime_to_datetime, iris_time_to_datetime
from improver.utilities.temporal_interpolation import TemporalInterpolation
(time_start,) = iris_time_to_datetime(start_cube.coord("time"))
(time_end,) = iris_time_to_datetime(end_cube.coord("time"))
if time_end < time_start:
# swap cubes
start_cube, end_cube = end_cube, start_cube
if times is not None:
times = [cycletime_to_datetime(timestr) for timestr in times]
result = TemporalInterpolation(
interval_in_minutes=interval_in_mins,
times=times,
interpolation_method=interpolation_method,
)(start_cube, end_cube)
return MergeCubes()(result)
def process(
cube: cli.inputcube,
coefficients: cli.inputcube = None,
land_sea_mask: cli.inputcube = None,
*,
distribution,
realizations_count: int = None,
randomise=False,
random_seed: int = None,
ignore_ecc_bounds=False,
predictor="mean",
shape_parameters: cli.comma_separated_list = None,
):
"""Applying coefficients for Ensemble Model Output Statistics.
Load in arguments for applying coefficients for Ensemble Model Output
Statistics (EMOS), otherwise known as Non-homogeneous Gaussian
Regression (NGR). The coefficients are applied to the forecast
that is supplied, so as to calibrate the forecast. The calibrated
forecast is written to a cube. If no coefficients are provided the input
forecast is returned unchanged.
Args:
cube (iris.cube.Cube):
A Cube containing the forecast to be calibrated. The input format
could be either realizations, probabilities or percentiles.
coefficients (iris.cube.Cube):
A cube containing the coefficients used for calibration or None.
If none then then input is returned unchanged.
land_sea_mask (iris.cube.Cube):
A cube containing the land-sea mask on the same domain as the
forecast that is to be calibrated. Land points are "
"specified by ones and sea points are specified by zeros. "
"If not None this argument will enable land-only calibration, in "
"which sea points are returned without the application of "
"calibration."
distribution (str):
The distribution for constructing realizations, percentiles or
probabilities. This should typically match the distribution used
for minimising the Continuous Ranked Probability Score when
estimating the EMOS coefficients. The distributions available are
those supported by :data:`scipy.stats`.
realizations_count (int):
Option to specify the number of ensemble realizations that will be
created from probabilities or percentiles for input into EMOS.
randomise (bool):
Option to reorder the post-processed forecasts randomly. If not
set, the ordering of the raw ensemble is used. This option is
only valid when the input format is realizations.
random_seed (int):
Option to specify a value for the random seed for testing
purposes, otherwise the default random seen behaviour is utilised.
The random seed is used in the generation of the random numbers
used for either the randomise option to order the input
percentiles randomly, rather than use the ordering from the raw
ensemble, or for splitting tied values within the raw ensemble,
so that the values from the input percentiles can be ordered to
match the raw ensemble.
ignore_ecc_bounds (bool):
If True, where the percentiles exceed the ECC bounds range,
raises a warning rather than an exception. This occurs when the
current forecasts is in the form of probabilities and is
converted to percentiles, as part of converting the input
probabilities into realizations.
predictor (str):
String to specify the form of the predictor used to calculate
the location parameter when estimating the EMOS coefficients.
Currently the ensemble mean ("mean") and the ensemble
realizations ("realizations") are supported as the predictors.
shape_parameters (float or str):
The shape parameters required for defining the distribution
specified by the distribution argument. The shape parameters
should either be a number or 'inf' or '-inf' to represent
infinity. Further details about appropriate shape parameters
are available in scipy.stats. For the truncated normal
distribution with a lower bound of zero, as available when
estimating EMOS coefficients, the appropriate shape parameters
are 0 and inf.
Returns:
iris.cube.Cube:
The calibrated forecast cube.
Raises:
ValueError:
If the current forecast is a coefficients cube.
ValueError:
If the coefficients cube does not have the right name of
"emos_coefficients".
ValueError:
If the forecast type is 'percentiles' or 'probabilities' and the
realizations_count argument is not provided.
"""
import warnings
import numpy as np
from improver.calibration.ensemble_calibration import ApplyEMOS
if cube.name() in ["emos_coefficients", "land_binary_mask"]:
msg = "Invalid forecast cube provided (name '{}')"
raise ValueError(msg.format(cube.name()))
if coefficients is None:
msg = ("There are no coefficients provided for calibration. The "
"uncalibrated forecast will be returned.")
warnings.warn(msg)
return cube
if coefficients.name() != "emos_coefficients":
msg = "Invalid coefficients cube provided (name '{}')"
raise ValueError(msg.format(coefficients.name()))
if land_sea_mask and land_sea_mask.name() != "land_binary_mask":
msg = "The land_sea_mask cube does not have the " "name 'land_binary_mask'"
raise ValueError(msg)
if shape_parameters:
shape_parameters = [np.float32(x) for x in shape_parameters]
calibration_plugin = ApplyEMOS()
result = calibration_plugin(
cube,
coefficients,
land_sea_mask=land_sea_mask,
realizations_count=realizations_count,
ignore_ecc_bounds=ignore_ecc_bounds,
predictor=predictor,
distribution=distribution,
shape_parameters=shape_parameters,
randomise=randomise,
random_seed=random_seed,
)
return result
def process(cube: cli.inputcube,
mask: cli.inputcube,
weights: cli.inputcube = None,
*,
radii: cli.comma_separated_list,
lead_times: cli.comma_separated_list = None,
area_sum=False,
return_intermediate=False):
""" Module to process land and sea separately before combining them.
Neighbourhood the input dataset over two distinct regions of land and sea.
If performed as a single level neighbourhood, a land-sea mask should be
provided. If instead topographic_zone neighbourhooding is being employed,
the mask should be one of topographic zones. In the latter case a weights
array is also needed to collapse the topographic_zone coordinate. These
weights are created with the improver generate-topography-bands-weights
CLI and should be made using a land-sea mask, which will then be employed
within this code to draw the distinction between the two surface types.
Args:
cube (iris.cube.Cube):
A cube to be processed.
mask (iris.cube.Cube):
A cube containing either a mask of topographic zones over land or
a land-sea mask.
weights (iris.cube.Cube):
A cube containing the weights which are used for collapsing the
dimension gained through masking. These weights must have been
created using a land-sea mask. (Optional).
radii (list of float):
The radius or a list of radii in metres of the neighbourhood to
apply.
If it is a list, it must be the same length as lead_times, which
defines at which lead time to use which nbhood radius. The radius
will be interpolated for intermediate lead times.
lead_times (list of int):
The lead times in hours that correspond to the radii to be used.
If lead_times are set, radii must be a list the same length as
lead_times. Lead times must be given as integer values.
area_sum (bool):
Return sum rather than fraction over the neighbourhood area.
return_intermediate (bool):
Include this option to return a cube with results following
topographic masked neighbourhood processing of land points and
prior to collapsing the topographic_zone coordinate. If no
topographic masked neighbourhooding occurs, there will be no
intermediate cube and a warning.
Returns:
(tuple): tuple containing:
**result** (iris.cube.Cube):
A cube of the processed data.
**intermediate_cube** (iris.cube.Cube):
A cube of the intermediate data, before collapsing.
Raises:
ValueError:
If the topographic zone mask has the attribute
topographic_zones_include_seapoints.
IOError:
if a weights cube isn't given and a topographic_zone mask is given.
ValueError:
If the weights cube has the attribute
topographic_zones_include_seapoints.
RuntimeError:
If lead times are not None and has a different length to radii.
TypeError:
A weights cube has been provided but no topographic zone.
"""
import warnings
import numpy as np
from improver.nbhood.nbhood import NeighbourhoodProcessing
from improver.nbhood.use_nbhood import (
ApplyNeighbourhoodProcessingWithAMask,
CollapseMaskedNeighbourhoodCoordinate)
sum_or_fraction = 'sum' if area_sum else 'fraction'
masking_coordinate = intermediate_cube = None
if any([
'topographic_zone' in coord.name()
for coord in mask.coords(dim_coords=True)
]):
if mask.attributes['topographic_zones_include_seapoints'] == 'True':
raise ValueError('The topographic zones mask cube must have been '
'masked to exclude sea points, but '
'topographic_zones_include_seapoints = True')
if not weights:
raise TypeError('A weights cube must be provided if using a mask '
'of topographic zones to collapse the resulting '
'vertical dimension.')
if weights.attributes['topographic_zones_include_seapoints'] == 'True':
raise ValueError('The weights cube must be masked to exclude sea '
'points, but topographic_zones_include_seapoints '
'= True')
masking_coordinate = 'topographic_zone'
land_sea_mask = weights[0].copy(data=weights[0].data.mask)
land_sea_mask.rename('land_binary_mask')
land_sea_mask.remove_coord(masking_coordinate)
# Create land and sea masks in IMPROVER format (inverse of
# numpy standard) 1 - include this region, 0 - exclude this region.
land_only = land_sea_mask.copy(
data=np.logical_not(land_sea_mask.data).astype(int))
sea_only = land_sea_mask.copy(data=land_sea_mask.data.astype(int))
else:
if weights is not None:
raise TypeError('A weights cube has been provided but will not be '
'used')
land_sea_mask = mask
# In this case the land is set to 1 and the sea is set to 0 in the
# input mask.
sea_only = land_sea_mask.copy(
data=np.logical_not(land_sea_mask.data).astype(int))
land_only = land_sea_mask.copy(data=land_sea_mask.data.astype(int))
if lead_times is None:
radius_or_radii = float(radii[0])
else:
if len(radii) != len(lead_times):
raise RuntimeError("If leadtimes are supplied, it must be a list"
" of equal length to a list of radii.")
radius_or_radii = [float(x) for x in radii]
lead_times = [int(x) for x in lead_times]
if return_intermediate is not None and masking_coordinate is None:
warnings.warn('No topographic_zone coordinate found, so no '
'intermediate file will be saved.')
# Section for neighbourhood processing land points.
if land_only.data.max() > 0.0:
if masking_coordinate is None:
result_land = NeighbourhoodProcessing(
'square',
radius_or_radii,
lead_times=lead_times,
sum_or_fraction=sum_or_fraction,
re_mask=True).process(cube, land_only)
else:
result_land = ApplyNeighbourhoodProcessingWithAMask(
masking_coordinate,
radius_or_radii,
lead_times=lead_times,
sum_or_fraction=sum_or_fraction,
re_mask=False).process(cube, mask)
if return_intermediate:
intermediate_cube = result_land.copy()
# Collapse the masking coordinate.
result_land = CollapseMaskedNeighbourhoodCoordinate(
masking_coordinate, weights=weights).process(result_land)
result = result_land
# Section for neighbourhood processing sea points.
if sea_only.data.max() > 0.0:
result_sea = NeighbourhoodProcessing('square',
radius_or_radii,
lead_times=lead_times,
sum_or_fraction=sum_or_fraction,
re_mask=True).process(
cube, sea_only)
result = result_sea
# Section for combining land and sea points following land and sea points
# being neighbourhood processed individually.
if sea_only.data.max() > 0.0 and land_only.data.max() > 0.0:
# Recombine cubes to be a single output.
combined_data = result_land.data.filled(0) + result_sea.data.filled(0)
result = result_land.copy(data=combined_data)
return result, intermediate_cube
def process(temperature: cli.inputcube,
orography: cli.inputcube = None,
land_sea_mask: cli.inputcube = None,
*,
max_height_diff: float = 35,
nbhood_radius: int = 7,
max_lapse_rate: float = -3 * DALR,
min_lapse_rate: float = DALR,
dry_adiabatic=False):
"""Calculate temperature lapse rates in units of K m-1 over orography grid.
Args:
temperature (iris.cube.Cube):
Air temperature data. This is required even when returning DALR,
as this defines the grid on which lapse rates are required.
orography (iris.cube.Cube):
Orography data.
land_sea_mask (iris.cube.Cube):
Binary land-sea mask data. True for land-points, False for sea.
max_height_diff (float):
Maximum allowable height difference between the central point and
points in the neighbourhood over which the lapse rate will be
calculated.
nbhood_radius (int):
Radius of neighbourhood in grid points around each point. The
neighbourhood is a square array with side length
2*nbhood_radius + 1. The default value of 7 is from the reference
paper (see plugin documentation).
max_lapse_rate (float):
Maximum lapse rate allowed, in K m-1.
min_lapse_rate (float):
Minimum lapse rate allowed, in K m-1.
dry_adiabatic (bool):
If True, returns a cube containing the dry adiabatic lapse rate
rather than calculating the true lapse rate.
Returns:
iris.cube.Cube:
Lapse rate (K m-1)
Raises:
ValueError: If minimum lapse rate is greater than maximum.
ValueError: If Maximum height difference is less than zero.
ValueError: If neighbourhood radius is less than zero.
RuntimeError: If calculating the true lapse rate and orography or
land mask arguments are not given.
"""
import numpy as np
from improver.lapse_rate import LapseRate
from improver.metadata.amend import amend_attributes
attributes_dict = {
"title": "remove",
"source": "remove",
"history": "remove",
"um_version": "remove"
}
if dry_adiabatic:
result = temperature.copy(data=np.full_like(temperature.data, DALR))
result.rename('air_temperature_lapse_rate')
result.units = U_DALR
amend_attributes(result, attributes_dict)
return result
if min_lapse_rate > max_lapse_rate:
msg = 'Minimum lapse rate specified is greater than the maximum.'
raise ValueError(msg)
if max_height_diff < 0:
msg = 'Maximum height difference specified is less than zero.'
raise ValueError(msg)
if nbhood_radius < 0:
msg = 'Neighbourhood radius specified is less than zero.'
raise ValueError(msg)
if orography is None or land_sea_mask is None:
msg = 'Missing orography and/or land mask arguments.'
raise RuntimeError(msg)
result = LapseRate(max_height_diff=max_height_diff,
nbhood_radius=nbhood_radius,
max_lapse_rate=max_lapse_rate,
min_lapse_rate=min_lapse_rate).process(
temperature, orography, land_sea_mask)
amend_attributes(result, attributes_dict)
return result
def process(
cube: cli.inputcube,
mask: cli.inputcube,
weights: cli.inputcube = None,
*,
neighbourhood_shape="square",
radii: cli.comma_separated_list,
lead_times: cli.comma_separated_list = None,
area_sum=False,
):
""" Module to process land and sea separately before combining them.
Neighbourhood the input dataset over two distinct regions of land and sea.
If performed as a single level neighbourhood, a land-sea mask should be
provided. If instead topographic_zone neighbourhooding is being employed,
the mask should be one of topographic zones. In the latter case a weights
array is also needed to collapse the topographic_zone coordinate. These
weights are created with the improver generate-topography-bands-weights
CLI and should be made using a land-sea mask, which will then be employed
within this code to draw the distinction between the two surface types.
Args:
cube (iris.cube.Cube):
A cube to be processed.
mask (iris.cube.Cube):
A cube containing either a mask of topographic zones over land or
a land-sea mask. If this is a land-sea mask, land points should be
set to one and sea points set to zero.
weights (iris.cube.Cube):
A cube containing the weights which are used for collapsing the
dimension gained through masking. These weights must have been
created using a land-sea mask. (Optional).
neighbourhood_shape (str):
Name of the neighbourhood method to use.
Options: "circular", "square".
Default: "square".
radii (list of float):
The radius or a list of radii in metres of the neighbourhood to
apply.
If it is a list, it must be the same length as lead_times, which
defines at which lead time to use which nbhood radius. The radius
will be interpolated for intermediate lead times.
lead_times (list of int):
The lead times in hours that correspond to the radii to be used.
If lead_times are set, radii must be a list the same length as
lead_times. Lead times must be given as integer values.
area_sum (bool):
Return sum rather than fraction over the neighbourhood area.
Returns:
(tuple): tuple containing:
**result** (iris.cube.Cube):
A cube of the processed data.
Raises:
ValueError:
If the topographic zone mask has the attribute
topographic_zones_include_seapoints.
IOError:
if a weights cube isn't given and a topographic_zone mask is given.
ValueError:
If the weights cube has the attribute
topographic_zones_include_seapoints.
RuntimeError:
If lead times are not None and has a different length to radii.
TypeError:
A weights cube has been provided but no topographic zone.
"""
import numpy as np
from improver.nbhood.nbhood import NeighbourhoodProcessing
from improver.nbhood.use_nbhood import ApplyNeighbourhoodProcessingWithAMask
masking_coordinate = None
if any(
"topographic_zone" in coord.name() for coord in mask.coords(dim_coords=True)
):
if mask.attributes["topographic_zones_include_seapoints"] == "True":
raise ValueError(
"The topographic zones mask cube must have been "
"masked to exclude sea points, but "
"topographic_zones_include_seapoints = True"
)
if not weights:
raise TypeError(
"A weights cube must be provided if using a mask "
"of topographic zones to collapse the resulting "
"vertical dimension."
)
if weights.attributes["topographic_zones_include_seapoints"] == "True":
raise ValueError(
"The weights cube must be masked to exclude sea "
"points, but topographic_zones_include_seapoints "
"= True"
)
masking_coordinate = "topographic_zone"
land_sea_mask = weights[0].copy(data=weights[0].data.mask)
land_sea_mask.rename("land_binary_mask")
land_sea_mask.remove_coord(masking_coordinate)
# Create land and sea masks in IMPROVER format (inverse of
# numpy standard) 1 - include this region, 0 - exclude this region.
land_only = land_sea_mask.copy(
data=np.logical_not(land_sea_mask.data).astype(int)
)
sea_only = land_sea_mask.copy(data=land_sea_mask.data.astype(int))
else:
if weights is not None:
raise TypeError("A weights cube has been provided but will not be " "used")
land_sea_mask = mask
# In this case the land is set to 1 and the sea is set to 0 in the
# input mask.
sea_only = land_sea_mask.copy(
data=np.logical_not(land_sea_mask.data).astype(int)
)
land_only = land_sea_mask.copy(data=land_sea_mask.data.astype(int))
if lead_times is None:
radius_or_radii = float(radii[0])
else:
if len(radii) != len(lead_times):
raise RuntimeError(
"If leadtimes are supplied, it must be a list"
" of equal length to a list of radii."
)
radius_or_radii = [float(x) for x in radii]
lead_times = [int(x) for x in lead_times]
# Section for neighbourhood processing land points.
if land_only.data.max() > 0.0:
if masking_coordinate is None:
result_land = NeighbourhoodProcessing(
neighbourhood_shape,
radius_or_radii,
lead_times=lead_times,
sum_only=area_sum,
re_mask=True,
)(cube, land_only)
else:
result_land = ApplyNeighbourhoodProcessingWithAMask(
masking_coordinate,
neighbourhood_shape,
radius_or_radii,
lead_times=lead_times,
collapse_weights=weights,
sum_only=area_sum,
)(cube, mask)
result = result_land
# Section for neighbourhood processing sea points.
if sea_only.data.max() > 0.0:
result_sea = NeighbourhoodProcessing(
neighbourhood_shape,
radius_or_radii,
lead_times=lead_times,
sum_only=area_sum,
re_mask=True,
)(cube, sea_only)
result = result_sea
# Section for combining land and sea points following land and sea points
# being neighbourhood processed individually.
if sea_only.data.max() > 0.0 and land_only.data.max() > 0.0:
# Recombine cubes to be a single output.
combined_data = result_land.data.filled(0) + result_sea.data.filled(0)
result = result_land.copy(data=combined_data)
return result
def process(
cube: cli.inputcube,
raw_cube: cli.inputcube = None,
*,
realizations_count: int = None,
random_seed: int = None,
ignore_ecc_bounds=False,
):
"""Converts an incoming cube into one containing realizations.
Args:
cube (iris.cube.Cube):
A cube to be processed.
raw_cube (iris.cube.Cube):
Cube of raw (not post processed) weather data.
If this argument is given ensemble realizations will be created
from percentiles by reshuffling them in correspondence to the rank
order of the raw ensemble. Otherwise, the percentiles are rebadged
as realizations.
realizations_count (int):
The number of ensemble realizations in the output.
random_seed (int):
Option to specify a value for the random seed when reordering percentiles.
This value is for testing purposes only, to ensure reproduceable outputs.
It should not be used in real time operations as it may introduce a bias
into the reordered forecasts.
ignore_ecc_bounds (bool):
If True where percentiles (calculated as an intermediate output
before realization) exceed the ECC bounds range, raises a
warning rather than an exception.
Returns:
iris.cube.Cube:
The processed cube.
"""
from improver.ensemble_copula_coupling.ensemble_copula_coupling import (
ConvertProbabilitiesToPercentiles,
EnsembleReordering,
RebadgePercentilesAsRealizations,
ResamplePercentiles,
)
from improver.metadata.probabilistic import is_probability
if cube.coords("realization"):
return cube
if not cube.coords("percentile") and not is_probability(cube):
raise ValueError("Unable to convert to realizations:\n" + str(cube))
if realizations_count is None:
try:
realizations_count = len(raw_cube.coord("realization").points)
except AttributeError:
# raised if raw_cube is None, hence has no attribute "coord"
msg = "Either realizations_count or raw_cube must be provided"
raise ValueError(msg)
if cube.coords("percentile"):
percentiles = ResamplePercentiles(
ecc_bounds_warning=ignore_ecc_bounds)(
cube, no_of_percentiles=realizations_count)
else:
percentiles = ConvertProbabilitiesToPercentiles(
ecc_bounds_warning=ignore_ecc_bounds)(
cube, no_of_percentiles=realizations_count)
if raw_cube:
result = EnsembleReordering()(percentiles,
raw_cube,
random_seed=random_seed)
else:
result = RebadgePercentilesAsRealizations()(percentiles)
return result