lair.air.hrrr#

HRRR winds at a point.

Documentation: https://mesowest.utah.edu/html/hrrr/zarr_documentation/html/python_data_loading.html

Module Attributes

PROJECTION

HRRR Projection

Functions

`create_s3_chunk_url`(zarr_id, chunk_id[, prefix])	Create the s3 chunk url for a zarr id.
`create_s3_group_url`(zarr_id[, prefix])	Create the s3 group url for a zarr id.
`create_s3_subgroup_url`(zarr_id[, prefix])	Create the s3 subgroup url for a zarr id.
`decompress_chunk`(zarr_id, compressed_data)	Decompress s3 chunk data.
`generate_zarr_ids`(times, level_type, ...)	Generate ZarrId instance for multiple times and variables
`get_nearest_point`(longitude, latitude, ...)	Get the nearest point to a given latitude and longitude.
`get_value`(s3, zarr_id, chunk_id, nearest_point)	Get an array of values from am s3 zarr chunk.
`retrieve_object`(s3, s3_url)	Retrieve object from s3.

Classes

`Winds`(lat, lon, times)	Class to download HRRR winds at a point.
`ZarrId`(run_hour, level_type, var_level, ...)	Class to store information about a HRRR zarr file.

lair.air.hrrr.PROJECTION: CRS = <Projected CRS: +proj=lcc +ellps=WGS84 +a=6371229 +b=6371229 +lon_ ...> Name: unknown Axis Info [cartesian]: - E[east]: Easting (metre) - N[north]: Northing (metre) Area of Use: - undefined Coordinate Operation: - name: unknown - method: Lambert Conic Conformal (2SP) Datum: unknown - Ellipsoid: unknown - Prime Meridian: Greenwich #: HRRR Projection

class lair.air.hrrr.ZarrId(run_hour: datetime, level_type: Literal['sfc', 'prs'], var_level: str, var_name: str, model_type: Literal['anl', 'fcst'])[source]#

Class to store information about a HRRR zarr file.

Attributes

run_hour	(dt.datetime) run hour of the model
level_type	(‘sfc’ \| ‘prs’) vertical level
var_level	(str) variable level
var_name	(str) variable name
model_type	(‘anl’ \| ‘fcst’) analysis or forecast

Methods

format_chunk_id(chunk_id)

Format chunk_id for zarr url.

format_chunk_id(chunk_id)[source]#

Format chunk_id for zarr url.

Parameters:

chunk_idint: chunk_id

Returns:

str: formatted chunk_id

__init__(run_hour: datetime, level_type: Literal['sfc', 'prs'], var_level: str, var_name: str, model_type: Literal['anl', 'fcst']) → None#

class lair.air.hrrr.Winds(lat: float, lon: float, times: list[datetime])[source]#

Class to download HRRR winds at a point.

Attributes

lat	(float) latitude of point
lon	(float) longitude of point
times	(list[dt.datetime]) list of times to download
nearest_point	(xr.Dataset) nearest point to lat, lon
chunk_id	(float) chunk id of nearest point
zarr_ids	(dict[str, list[ZarrId]]) dictionary of zarr ids with var as key
data	(pd.DataFrame) wind data

__init__(lat: float, lon: float, times: list[datetime]) → None#

lair.air.hrrr.create_s3_group_url(zarr_id: ZarrId, prefix: bool = True) → str[source]#

Create the s3 group url for a zarr id.

Start of the zarr array data format

Includes metadata such as the grid

Parameters:

zarr_idZarrId: ZarrId instance
prefixbool: include prefix

Returns:

str: s3 group url

lair.air.hrrr.create_s3_subgroup_url(zarr_id: ZarrId, prefix: bool = True) → str[source]#

Create the s3 subgroup url for a zarr id.

Where the actual data variable is stored

While subgroups are part of the zarr spec, there’s not necessarily a good reason for the data in this case to have this extra level of nesting

Parameters:

zarr_idZarrId: ZarrId instance
prefixbool: include prefix

Returns:

str: s3 subgroup url

lair.air.hrrr.create_s3_chunk_url(zarr_id: ZarrId, chunk_id, prefix: bool = False) → str[source]#

Create the s3 chunk url for a zarr id.

Contains just the data for the chunk (no metadata)

Current APIs (zarr, xarray) don’t support reading zarr data by chunks, so we have to write relatively low-level code to load data on this level

Parameters:

zarr_idZarrId: ZarrId instance
chunk_idint # TODO check this: chunk id
prefixbool: include prefix

Returns:

str: s3 chunk url

lair.air.hrrr.get_nearest_point(longitude: float, latitude: float, chunk_index: Dataset) → Dataset[source]#

Get the nearest point to a given latitude and longitude.

Parameters:

longitudefloat: longitude of point
latitudefloat: latitude of point
chunk_indexxr.Dataset: chunk index dataset

Returns:

xr.Dataset: nearest point

lair.air.hrrr.retrieve_object(s3, s3_url: str)[source]#

Retrieve object from s3.

Parameters:

s3boto3.resource: s3 resource
s3_urlstr: s3 url

Returns:

bytes: compressed object data

lair.air.hrrr.generate_zarr_ids(times: list[datetime], level_type: str, variables: list[Tuple[str, str]], model_type: Literal['anl', 'fcst'])[source]#

Generate ZarrId instance for multiple times and variables

Parameters:

times[dt.datetime]: list of datetime objects with hour specified [UTC]
level_typestr: veritical level
variables[(str, str)]: list of tuples: (var_name, var_level)
model_typestr: ‘anl’ | ‘fcst’

Returns:

ids[ZarrId]: dictionary of zarr ids with var as key

lair.air.hrrr.decompress_chunk(zarr_id: ZarrId, compressed_data: bytes)[source]#

Decompress s3 chunk data.

Parameters:

zarr_idZarrId: ZarrId instance
compressed_databytes: compressed data

Returns:

np.array: decompressed chunk

lair.air.hrrr.get_value(s3, zarr_id: ZarrId, chunk_id, nearest_point: Dataset)[source]#

Get an array of values from am s3 zarr chunk.

Parameters:

s3boto3.resource: s3 resource
zarr_idZarrId: ZarrId instance
chunk_idfloat: chunk id
nearest_pointxr.Dataset: nearest point

Returns:

np.array: array of values