Quickstart
Loading and accessing data
To work with HSI data you must create a Hyper object and
it uses Satpy to get access to the data.
You need to input what files to read and what Satpy Reader should read them:
>>> from hypergas import Hyper
>>> from glob import glob
>>> filenames = glob('EMIT_L1B_*_001_20230814T100819*nc')
>>> hyp = Hyper(filenames, reader='emit_l1b') # EnMAP: "hsi_l1b"; EMIT: "emit_l1b"; PRISMA: "hyc_l1"
To load data from the files use the Hyper.load
method. Printing the Scene object will list each of the
xarray.DataArray objects currently loaded:
>>> hyp.load()
>>> print(hyp.scene)
<xarray.DataArray 'radiance' (bands: 253, y: 1280, x: 1242)>
dask.array<where, shape=(253, 1280, 1242), dtype=float32, chunksize=(253, 1280, 1242), chunktype=numpy.ndarray>
Coordinates:
fwhm (bands) float32 dask.array<chunksize=(253,), meta=np.ndarray>
* bands (bands) float32 381.0 388.4 395.8 ... 2.478e+03 2.486e+03 2.493e+03
Dimensions without coordinates: y, x
Attributes: (12/20)
start_time: 2023-08-14 10:08:19
end_time: 2023-08-14 10:08:31
long_name: Radiance Data
units: uW cm-2 sr-1 nm-1
name: radiance
file_type: ['emit_l1b_rad']
... ...
area: Shape: (1280, 1242)\nLons: [[52.19963936194076 52.1...
reader: emit_l1b
_satpy_id: DataID(name='radiance', modifiers=())
ancillary_variables: []
sza: 36.839778900146484
vza: 8.523066520690918
<xarray.DataArray 'rgb' (bands: 3, y: 1280, x: 1242)>
Coordinates:
* bands (bands) int64 650 560 470
Dimensions without coordinates: y, x
Attributes: (12/20)
start_time: 2023-08-14 10:08:19
end_time: 2023-08-14 10:08:31
long_name: true color (RGB)
units: 1
name: rgb
file_type: ['emit_l1b_rad']
... ...
area: Shape: (1280, 1242)\nLons: [[52.19963936194076 52.1...
reader: emit_l1b
_satpy_id: DataID(name='rgb', modifiers=())
ancillary_variables: []
sza: 36.839778900146484
vza: 8.523066520690918
By default, HyperGas loads data while disregarding water vapor bands.
If you want keep them, you can set drop_waterbands to False:
>>> hyp.load(drop_waterbands=False)
Visualizing data
To visualize loaded data in jupyter notebook:
>>> import os
>>> import folium
>>> from hypergas.folium_map import Map
>>> # convert to Dataset
>>> ds = hyp.scene.to_xarray(datasets='rgb')
>>> ds.attrs['filename'] = os.path.abspath(filename[0])
>>> # initialize folium map
>>> m = Map(ds, ['rgb'])
>>> m.initialize()
>>> m.plot(show_layers=[True], opacities=[0.7])
>>> # show the map in notebook
>>> layer_control = folium.LayerControl(collapsed=False, position='topleft', draggable=True)
>>> m.map.add_child(layer_control)
>>> # export to html file
>>> m.export()
Retrieval
The Hyper.retrieve function calculates the trace gas enhancement using matched filter.
You can input species like “ch4” or “co2” to run the retrieval directly.
>>> hyp.retrieve(species='ch4')
>>> ch4 = hyp.scene['ch4']
>>> print(ch4)
<xarray.DataArray 'ch4' (bands: 1, y: 1280, x: 1242)>
Dimensions without coordinates: bands, y, x
Attributes: (12/21)
start_time: 2023-08-14 10:08:19
end_time: 2023-08-14 10:08:31
long_name: methane_enhancement
units: ppb
name: ch4
file_type: ['emit_l1b_rad']
... ...
reader: emit_l1b
_satpy_id: DataID(name='ch4', modifiers=())
ancillary_variables: []
sza: 36.839778900146484
vza: 8.523066520690918
description: methane enhancement derived by the 2100~2450 nm window
If you prefer your own wavelength range, you can modify the wvl_intervals parameter,
which can be a list or multi-list.
>>> hyp.retrieve(wvl_intervals=[1300, 2500], species='ch4')
>>> hyp.retrieve(wvl_intervals=[[1600, 1750], [2110, 2450]], species='ch4')
Saving to disk
To save loaded datasets to disk as NetCDF file:
>>> vnames = ['u10', 'v10', 'sp', 'rgb', 'radiance_2100', 'ch4']
>>> loaded_names = [x['name'] for x in hyp.scene.keys()]
>>> # drop not loaded vnames
>>> vnames = [vname for vname in vnames if vname in loaded_names]
>>> # set global attrs
>>> header_attrs = {'author': 'AUTHOR'}
>>> hyp.scene.save_datasets(datasets=vnames, filename='test.nc',
header_attrs=header_attrs, writer='cf'
)