API¶

Initialisation¶

Geedim extends the GEE API with the ee.Image.gd and ee.ImageCollection.gd accessors. To enable the accessors, Geedim must be imported:

import ee
import geedim  # noqa: F401

ee.Initialize()

If you use a linter, you may need to include the # noqa: F401 suppression comment after the geedim import to prevent it being removed.

Cloud masking¶

Cloud masking is supported on Landsat 4-9 collection 2 images and Sentinel-2 TOA and surface reflectance images. The cloudSupport property is True on images / collections with cloud mask support:

im = ee.Image('COPERNICUS/S2_SR_HARMONIZED/20211220T080341_20211220T082827_T35HKC')
im.gd.cloudSupport

True

Image and collection accessors share the same interface for cloud masking. Masks and related bands can be added with addMaskBands(), and cloud masks applied with maskClouds():

# add mask bands created with a threshold of 0.7 on the 'cs_cdf' Cloud Score+ band
im = im.gd.addMaskBands(score=0.7, cs_band='cs_cdf')
print(im.gd.bandNames)

im = im.gd.maskClouds()

['B1', 'B2', 'B3', 'B4', 'B5', 'B6', 'B7', 'B8', 'B8A', 'B9', 'B11', 'B12', 'AOT', 'WVP', 'SCL', 'TCI_R', 'TCI_G', 'TCI_B', 'MSK_CLDPRB', 'MSK_SNWPRB', 'QA10', 'QA20', 'QA60', 'MSK_CLASSI_OPAQUE', 'MSK_CLASSI_CIRRUS', 'MSK_CLASSI_SNOW_ICE', 'CLOUD_SCORE', 'FILL_MASK', 'CLOUDLESS_MASK', 'CLOUD_DIST']

The addMaskBands() reference documents the masking parameters for images / collections.

Filtering¶

Collections can be filtered on the filled or cloud-free portions of a given region, and other criteria, with filter():

region = ee.Geometry.Rectangle(24.35, -33.75, 24.45, -33.65)
coll = ee.ImageCollection('COPERNICUS/S2_SR_HARMONIZED')

# filter by date range, region bounds, and a lower limit of 60% on the cloud-free
# portion of region
filt_coll = coll.gd.filter(
    '2021-10-01', '2022-04-01', region=region, cloudless_portion=60
)

The schemaTable and propertiesTable properties allow the collection contents to be displayed. schemaPropertyNames defines a set of image properties to include in the tables:

# include the VEGETATION_PERCENTAGE property in schemaTable & propertiesTable
filt_coll.gd.schemaPropertyNames += ('VEGETATION_PERCENTAGE',)

print(filt_coll.gd.schemaTable)
print(filt_coll.gd.propertiesTable)

ABBREV     NAME                             DESCRIPTION
---------  -------------------------------  ------------------------------------------------
INDEX      system:index                     Earth Engine image index
DATE       system:time_start                Image capture date/time (UTC)
FILL       FILL_PORTION                     Portion of region pixels that are valid (%)
CLOUDLESS  CLOUDLESS_PORTION                Portion of filled pixels that are cloud-free (%)
RADQ       RADIOMETRIC_QUALITY              Radiometric quality check
GEOMQ      GEOMETRIC_QUALITY                Geometric quality check
SAA        MEAN_SOLAR_AZIMUTH_ANGLE         Solar azimuth angle (deg)
SZA        MEAN_SOLAR_ZENITH_ANGLE          Solar zenith angle (deg)
VAA        MEAN_INCIDENCE_AZIMUTH_ANGLE_B1  View (B1) azimuth angle (deg)
VZA        MEAN_INCIDENCE_ZENITH_ANGLE_B1   View (B1) zenith angle (deg)
VP         VEGETATION_PERCENTAGE            Percentage of pixels classified as vegetation

INDEX                                  DATE               FILL CLOUDLESS RADQ   GEOMQ    SAA   SZA    VAA  VZA    VP
-------------------------------------- ---------------- ------ --------- ------ ------ ----- ----- ------ ---- -----
20211006T075809_20211006T082043_T35HKC 2021-10-06 08:29 100.00     99.33 PASSED PASSED 44.88 37.00 183.01 3.06 22.25
20211021T075951_20211021T082750_T35HKC 2021-10-21 08:29 100.00    100.00 PASSED PASSED 49.71 32.07 183.15 3.07 14.52
20211031T080051_20211031T082748_T35HKC 2021-10-31 08:29 100.00    100.00 PASSED PASSED 53.80 29.38 182.67 3.07 23.20
20211115T080109_20211115T082145_T35HKC 2021-11-15 08:29 100.00     90.59 PASSED PASSED 60.95 26.59 181.27 3.06  7.68
20211130T080301_20211130T082811_T35HKC 2021-11-30 08:29 100.00     67.26 PASSED PASSED 67.94 25.46 180.43 3.06 14.30
20211210T080321_20211210T082814_T35HKC 2021-12-10 08:29 100.00     73.88 PASSED PASSED 71.57 25.60 182.17 3.06 16.56
20211220T080341_20211220T082827_T35HKC 2021-12-20 08:29 100.00     81.07 PASSED PASSED 73.81 26.29 183.00 3.07 21.30
20211225T080239_20211225T082126_T35HKC 2021-12-25 08:29  99.83     85.90 PASSED PASSED 74.37 26.83 182.86 3.06  7.51
20220109T080321_20220109T082806_T35HKC 2022-01-09 08:29 100.00    100.00 PASSED PASSED 73.66 28.85 182.17 3.06 28.63
20220119T080241_20220119T082733_T35HKC 2022-01-19 08:29 100.00    100.00 PASSED PASSED 71.52 30.51 180.43 3.06 18.67
20220124T080119_20220124T081814_T35HKC 2022-01-24 08:29 100.00    100.00 PASSED PASSED 70.05 31.42 179.96 3.06 13.94
20220129T080201_20220129T082905_T35HKC 2022-01-29 08:29 100.00    100.00 PASSED PASSED 68.32 32.34 180.45 3.06 20.57
20220208T080101_20220208T082541_T35HKC 2022-02-08 08:29 100.00    100.00 PASSED PASSED 64.36 34.34 180.77 3.07 22.98
20220218T080001_20220218T082230_T35HKC 2022-02-18 08:29 100.00     99.97 PASSED PASSED 59.90 36.50 181.99 3.06 22.38
20220223T075919_20220223T082859_T35HKC 2022-02-23 08:29 100.00    100.00 PASSED PASSED 57.62 37.67 181.52 3.06 25.82
20220228T075851_20220228T082659_T35HKC 2022-02-28 08:29 100.00    100.00 PASSED PASSED 55.24 38.84 182.47 3.06 28.50
20220305T075809_20220305T082125_T35HKC 2022-03-05 08:29 100.00     99.64 PASSED PASSED 52.93 40.09 181.19 3.06 26.51
20220310T075741_20220310T082729_T35HKC 2022-03-10 08:29 100.00     97.49 PASSED PASSED 50.60 41.35 182.55 3.06 13.33
20220315T075659_20220315T082125_T35HKC 2022-03-15 08:29 100.00     96.10 PASSED PASSED 48.38 42.68 182.00 3.06 15.17
20220330T075611_20220330T082727_T35HKC 2022-03-30 08:29 100.00    100.00 PASSED PASSED 42.13 46.78 181.65 3.07 21.63

Compositing¶

Collections can be composited using composite(). By default, cloud is masked in the component images before compositing. E.g. to form a cloud-free median composite:

# create and filter a collection
region = ee.Geometry.Rectangle(24.35, -33.75, 24.45, -33.65)
coll = ee.ImageCollection('COPERNICUS/S2_SR_HARMONIZED')
filt_coll = coll.gd.filter('2021-10-01', '2022-04-01', region=region)

# composite
comp_im = filt_coll.gd.composite(method='median')

CompositeMethod documents supported values for the method parameter. The mosaic, q_mosaic, and medoid methods prioritise images in their sort order i.e. when more than one image pixel qualifies for selection, they select the first one. Images can be sorted by closeness to the date parameter, or by cloud-free portion of the region parameter. If neither date or region are supplied, images are sorted by capture date.

Exporting¶

Preparation¶

Images are exported with the projection and bounds given by their crs, transform and shape properties; and with data type given by their dtype property:

im = ee.Image('COPERNICUS/S2_SR_HARMONIZED/20211220T080341_20211220T082827_T35HKC')

print(im.gd.crs)
print(im.gd.transform)
print(im.gd.shape)
print(im.gd.dtype)

EPSG:32735
(10, 0, 199980, 0, -10, 6300040)
(10980, 10980)
uint32

Collections are exported with the projection, bounds and data type given by the first collection image.

Both the image and collection accessors have a prepareForExport() method with the same parameters. This can be called before exporting to change the projection, bounds and data type:

Note

This is required for:

images without a fixed projection (e.g. composites)
collections whose images don’t have a fixed projection, or don’t share the same projection, bounds and data type

region = ee.Geometry.Rectangle(24.35, -33.75, 24.45, -33.65)
prep_im = im.gd.prepareForExport(
    crs='EPSG:3857', region=region, scale=30, dtype='uint16'
)

Projection and bounds can be defined with the crs, region and scale / shape; or crs, crs_transform and shape parameters. Other parameters alter resampling, selected bands and scale / offset - see the ee.Image.gd.prepareForExport() or ee.ImageCollection.gd.prepareForExport() docs for details.

GeoTIFF¶

Image¶

ee.Image.gd.toGeoTIFF() exports an image to a GeoTIFF file:

# create and prepare an image
im = ee.Image('COPERNICUS/S2_SR_HARMONIZED/20211220T080341_20211220T082827_T35HKC')
region = ee.Geometry.Rectangle(24.35, -33.75, 24.45, -33.65)
prep_im = im.gd.prepareForExport(region=region, scale=30, dtype='uint16')

# export
prep_im.gd.toGeoTIFF('s2.tif')

Image properties are written to the GeoTIFF default namespace tags, and bandProps are written to the band tags:

import rasterio as rio

with rio.open('s2.tif') as ds:
    # default namespace tags
    print(ds.tags())
    # band 1 tags
    print(ds.tags(bidx=1))

{'AOT_RETRIEVAL_ACCURACY': '0', 'CLOUDY_PIXEL_PERCENTAGE': '7.464998', 'CLOUD_COVERAGE_ASSESSMENT': '7.464998', 'CLOUD_SHADOW_PERCENTAGE': '0.476946', 'DARK_FEATURES_PERCENTAGE': '2.535131', 'DATASTRIP_ID': 'S2A_OPER_MSI_L2A_DS_VGS2_20211220T100229_S20211220T082827_N03.01', 'DATATAKE_IDENTIFIER': 'GS2A_20211220T080341_033923_N03.01', 'DATATAKE_TYPE': 'INS-NOBS', 'DEGRADED_MSI_DATA_PERCENTAGE': '0', 'FORMAT_CORRECTNESS': 'PASSED', 'GENERAL_QUALITY': 'PASSED', 'GENERATION_TIME': '1639994549000', 'GEOMETRIC_QUALITY': 'PASSED', 'GRANULE_ID': 'L2A_T35HKC_A033923_20211220T082827', 'HIGH_PROBA_CLOUDS_PERCENTAGE': '4.461143', 'LICENSE': 'https://developers.google.com/earth-engine/datasets/catalog/COPERNICUS_S2_SR_HARMONIZED#terms-of-use', 'MEAN_INCIDENCE_AZIMUTH_ANGLE_B1': '183.002238919314', 'MEAN_INCIDENCE_AZIMUTH_ANGLE_B10': '183.209778422443', 'MEAN_INCIDENCE_AZIMUTH_ANGLE_B11': '182.210025828693', 'MEAN_INCIDENCE_AZIMUTH_ANGLE_B12': '181.730903604473', 'MEAN_INCIDENCE_AZIMUTH_ANGLE_B2': '187.678466476237', 'MEAN_INCIDENCE_AZIMUTH_ANGLE_B3': '185.366615351959', 'MEAN_INCIDENCE_AZIMUTH_ANGLE_B4': '184.132241490031', 'MEAN_INCIDENCE_AZIMUTH_ANGLE_B5': '183.682922936926', 'MEAN_INCIDENCE_AZIMUTH_ANGLE_B6': '183.3563508309', 'MEAN_INCIDENCE_AZIMUTH_ANGLE_B7': '183.109691838117', 'MEAN_INCIDENCE_AZIMUTH_ANGLE_B8': '186.357005411029', 'MEAN_INCIDENCE_AZIMUTH_ANGLE_B8A': '182.941314717219', 'MEAN_INCIDENCE_AZIMUTH_ANGLE_B9': '182.928253148734', 'MEAN_INCIDENCE_ZENITH_ANGLE_B1': '3.06615278603653', 'MEAN_INCIDENCE_ZENITH_ANGLE_B10': '2.61190347479859', 'MEAN_INCIDENCE_ZENITH_ANGLE_B11': '2.79184501565815', 'MEAN_INCIDENCE_ZENITH_ANGLE_B12': '2.99404018653664', 'MEAN_INCIDENCE_ZENITH_ANGLE_B2': '2.4284683919248', 'MEAN_INCIDENCE_ZENITH_ANGLE_B3': '2.53509223726786', 'MEAN_INCIDENCE_ZENITH_ANGLE_B4': '2.6563004578057', 'MEAN_INCIDENCE_ZENITH_ANGLE_B5': '2.72988386474952', 'MEAN_INCIDENCE_ZENITH_ANGLE_B6': '2.80815177009197', 'MEAN_INCIDENCE_ZENITH_ANGLE_B7': '2.89096735679217', 'MEAN_INCIDENCE_ZENITH_ANGLE_B8': '2.47809796571547', 'MEAN_INCIDENCE_ZENITH_ANGLE_B8A': '2.97743523242124', 'MEAN_INCIDENCE_ZENITH_ANGLE_B9': '3.16193398356126', 'MEAN_SOLAR_AZIMUTH_ANGLE': '73.8073718283396', 'MEAN_SOLAR_ZENITH_ANGLE': '26.2914700823507', 'MEDIUM_PROBA_CLOUDS_PERCENTAGE': '2.397248', 'MGRS_TILE': '35HKC', 'NODATA_PIXEL_PERCENTAGE': '0', 'NOT_VEGETATED_PERCENTAGE': '39.013728', 'PROCESSING_BASELINE': '03.01', 'PRODUCT_ID': 'S2A_MSIL2A_20211220T080341_N0301_R035_T35HKC_20211220T100229', 'RADIATIVE_TRANSFER_ACCURACY': '0', 'RADIOMETRIC_QUALITY': 'PASSED', 'REFLECTANCE_CONVERSION_CORRECTION': '1.0326712354468', 'SATURATED_DEFECTIVE_PIXEL_PERCENTAGE': '0', 'SENSING_ORBIT_DIRECTION': 'DESCENDING', 'SENSING_ORBIT_NUMBER': '35', 'SENSOR_QUALITY': 'PASSED', 'SNOW_ICE_PERCENTAGE': '0.00505', 'SOLAR_IRRADIANCE_B1': '1884.69', 'SOLAR_IRRADIANCE_B10': '367.15', 'SOLAR_IRRADIANCE_B11': '245.59', 'SOLAR_IRRADIANCE_B12': '85.25', 'SOLAR_IRRADIANCE_B2': '1959.66', 'SOLAR_IRRADIANCE_B3': '1823.24', 'SOLAR_IRRADIANCE_B4': '1512.06', 'SOLAR_IRRADIANCE_B5': '1424.64', 'SOLAR_IRRADIANCE_B6': '1287.61', 'SOLAR_IRRADIANCE_B7': '1162.08', 'SOLAR_IRRADIANCE_B8': '1041.63', 'SOLAR_IRRADIANCE_B8A': '955.32', 'SOLAR_IRRADIANCE_B9': '812.92', 'SPACECRAFT_NAME': 'Sentinel-2A', 'system-asset_size': '1735937022', 'system-footprint': "{'geodesic': False, 'crs': {'type': 'name', 'properties': {'name': 'EPSG:32735'}}, 'type': 'Polygon', 'coordinates': [[[254220, 6262390], [263820, 6262390], [263820, 6273760], [254220, 6273760], [254220, 6262390]]]}", 'system-index': '20211220T080341_20211220T082827_T35HKC', 'system-time_end': '1639988982907', 'system-time_start': '1639988982907', 'THIN_CIRRUS_PERCENTAGE': '0.606607', 'UNCLASSIFIED_PERCENTAGE': '1.75378', 'VEGETATION_PERCENTAGE': '21.298875', 'WATER_PERCENTAGE': '27.451491', 'WATER_VAPOUR_RETRIEVAL_ACCURACY': '0', 'AREA_OR_POINT': 'Area'}
{'center_wavelength': '0.4439', 'description': 'Aerosols', 'gee-scale': '0.0001', 'gee-wavelength': '443.9nm (S2A) / 442.3nm (S2B)', 'gsd': '60', 'name': 'B1'}

Collection¶

ee.ImageCollection.gd.toGeoTIFF() exports a collection to GeoTIFF files. The split parameter controls whether exported files correspond to collection bands or images:

from pathlib import Path

# create and prepare a collection (with two images and three bands)
coll = ee.ImageCollection('COPERNICUS/S2_SR_HARMONIZED')
region = ee.Geometry.Rectangle(24.35, -33.75, 24.45, -33.65)
coll = coll.filterBounds(region).limit(2)
prep_coll = coll.gd.prepareForExport(
    region=region, scale=30, dtype='uint16', bands=['B4', 'B3', 'B2']
)

# create export directory
dirname = Path('s2')
dirname.mkdir()

# export (one file for each collection band)
prep_coll.gd.toGeoTIFF(dirname, split='bands')

# display exported files
[fp.name for fp in dirname.glob('*.tif')]

['B3.tif', 'B2.tif', 'B4.tif']

When split is images, image properties are written to the GeoTIFF default namespace tags, and bandProps are written to the band tags (see the image example).

Nodata¶

By default, GeoTIFF file nodata tags are set to the nodata value of their corresponding images. Both ee.Image.gd.toGeoTIFF() and ee.ImageCollection.gd.toGeoTIFF() have a nodata parameter that allows this to be changed. E.g.:

# set masked pixels to a new nodata value
nodata = 65535
prep_im = prep_im.unmask(nodata)

# export, setting nodata to the new value
prep_im.gd.toGeoTIFF('s2_nodata.tif', nodata=nodata)

# display GeoTIFF nodata
with rio.open('s2_nodata.tif') as ds:
    print(ds.nodata)

65535.0

Paths and URIs¶

The file argument in ee.Image.gd.toGeoTIFF() and dirname argument in ee.ImageCollection.gd.toGeoTIFF() can be local paths or remote URIs. See the related note in the command line section for more information.

NumPy¶

Image¶

ee.Image.gd.toNumPy() exports an image to a NumPy ndarray:

# create and prepare an image (with 3 bands)
im = ee.Image('COPERNICUS/S2_SR_HARMONIZED/20211220T080341_20211220T082827_T35HKC')
region = ee.Geometry.Rectangle(24.35, -33.75, 24.45, -33.65)
prep_im = im.gd.prepareForExport(
    region=region, scale=30, dtype='uint16', bands=['B4', 'B3', 'B2']
)

# export (3D array with bands along the third dimension)
array = prep_im.gd.toNumPy()

# display array format
print(type(array))
print(array.shape)
print(array.dtype)

<class 'numpy.ndarray'>
(379, 320, 3)
uint16

Collection¶

ee.ImageCollection.gd.toNumPy() exports a collection to a NumPy ndarray. The split parameter controls the layout of collection bands and images in the exported array:

# create and prepare a collection (with two images and three bands)
coll = ee.ImageCollection('COPERNICUS/S2_SR_HARMONIZED')
region = ee.Geometry.Rectangle(24.35, -33.75, 24.45, -33.65)
coll = coll.filterBounds(region).limit(2)
prep_coll = coll.gd.prepareForExport(
    region=region, scale=30, dtype='uint16', bands=['B4', 'B3', 'B2']
)

# export (4D array with bands along the third, and images along the fourth dimension)
array = prep_coll.gd.toNumPy(split='bands')

# display array format
print(type(array))
print(array.shape)
print(array.dtype)

<class 'numpy.ndarray'>
(379, 320, 3, 2)
uint16

Masking and data type¶

Both ee.Image.gd.toNumPy() and ee.ImageCollection.gd.toNumPy() have masked and structured parameters. The masked parameter controls whether the exported array has masked pixels set to nodata, or is a MaskedArray. The structured parameter controls whether the exported array has a numerical or structured data type. E.g.:

# export (2D masked array with a structured dtype representing the bands)
array = prep_im.gd.toNumPy(masked=True, structured=True)

# display array format
print(type(array))
print(array.shape)
print(array.dtype)

<class 'numpy.ma.MaskedArray'>
(379, 320)
[('B4', '<u2'), ('B3', '<u2'), ('B2', '<u2')]

Xarray¶

Image¶

ee.Image.gd.toXarray() exports an image to a Xarray DataArray:

# create and prepare image
im = ee.Image('COPERNICUS/S2_SR_HARMONIZED/20211220T080341_20211220T082827_T35HKC')
region = ee.Geometry.Rectangle(24.35, -33.75, 24.45, -33.65)
prep_im = im.gd.prepareForExport(
    region=region, scale=30, dtype='uint16', bands=['B4', 'B3', 'B2']
)

# export (3D DataArray)
da = prep_im.gd.toXarray()

da

<xarray.DataArray (y: 379, x: 320, band: 3)> Size: 728kB
427 450 343 991 867 532 913 1058 730 ... 964 882 650 998 918 756 1033 996 797
Coordinates:
  * y        (y) float64 3kB 6.274e+06 6.274e+06 ... 6.262e+06 6.262e+06
  * x        (x) float64 3kB 2.542e+05 2.543e+05 ... 2.638e+05 2.638e+05
  * band     (band) <U2 24B 'B4' 'B3' 'B2'
Attributes: (7)

xarray.DataArray

y: 379
x: 320
band: 3

427 450 343 991 867 532 913 1058 ... 882 650 998 918 756 1033 996 797

array([[[ 427,  450,  343],
        [ 991,  867,  532],
        [ 913, 1058,  730],
        ...,
        [1200, 1003,  753],
        [1160, 1002,  728],
        [1026,  893,  640]],

       [[ 406,  518,  285],
        [ 959,  900,  546],
        [1077, 1152,  878],
        ...,
        [1047,  905,  693],
        [1027,  901,  672],
        [1172, 1022,  750]],

       [[ 338,  430,  227],
        [ 937,  721,  409],
        [1205, 1147,  666],
        ...,
...
        ...,
        [1058,  895,  608],
        [1076,  899,  614],
        [1221,  957,  643]],

       [[ 227,  243,  190],
        [ 304,  348,  278],
        [ 437,  462,  317],
        ...,
        [1042,  956,  731],
        [ 984,  896,  667],
        [1182, 1127,  896]],

       [[ 233,  259,  209],
        [ 477,  498,  370],
        [ 427,  469,  326],
        ...,
        [ 964,  882,  650],
        [ 998,  918,  756],
        [1033,  996,  797]]], shape=(379, 320, 3), dtype=uint16)

Coordinates: (3)

(y)

float64

6.274e+06 6.274e+06 ... 6.262e+06

array([6273745., 6273715., 6273685., ..., 6262465., 6262435., 6262405.],
      shape=(379,))

(x)

float64

2.542e+05 2.543e+05 ... 2.638e+05

array([254235., 254265., 254295., ..., 263745., 263775., 263805.], shape=(320,))

band
(band)
<U2
'B4' 'B3' 'B2'
```
array(['B4', 'B3', 'B2'], dtype='<U2')
```

Attributes: (7)
id :
COPERNICUS/S2_SR_HARMONIZED/20211220T080341_20211220T082827_T35HKC
date :
2021-12-20T08:29:42.907+00:00
crs :
EPSG:32735
transform :
(30.0, 0.0, 254220.0, 0.0, -30.0, 6273760.0)
nodata :
0
ee :
{"system:footprint": {"geodesic": false, "crs": {"type": "name", "properties": {"name": "EPSG:32735"}}, "type": "Polygon", "coordinates": [[[254220, 6262390], [263820, 6262390], [263820, 6273760], [254220, 6273760], [254220, 6262390]]]}, "DATATAKE_IDENTIFIER": "GS2A_20211220T080341_033923_N03.01", "AOT_RETRIEVAL_ACCURACY": 0, "SPACECRAFT_NAME": "Sentinel-2A", "SATURATED_DEFECTIVE_PIXEL_PERCENTAGE": 0, "MEAN_INCIDENCE_AZIMUTH_ANGLE_B8A": 182.941314717219, "CLOUD_SHADOW_PERCENTAGE": 0.476946, "MEAN_SOLAR_AZIMUTH_ANGLE": 73.8073718283396, "VEGETATION_PERCENTAGE": 21.298875, "SOLAR_IRRADIANCE_B12": 85.25, "SOLAR_IRRADIANCE_B10": 367.15, "SENSOR_QUALITY": "PASSED", "SOLAR_IRRADIANCE_B11": 245.59, "GENERATION_TIME": 1639994549000, "SOLAR_IRRADIANCE_B8A": 955.32, "FORMAT_CORRECTNESS": "PASSED", "CLOUD_COVERAGE_ASSESSMENT": 7.464998, "THIN_CIRRUS_PERCENTAGE": 0.606607, "system:time_end": 1639988982907, "WATER_VAPOUR_RETRIEVAL_ACCURACY": 0, "system:time_start": 1639988982907, "DATASTRIP_ID": "S2A_OPER_MSI_L2A_DS_VGS2_20211220T100229_S20211220T082827_N03.01", "PROCESSING_BASELINE": "03.01", "SENSING_ORBIT_NUMBER": 35, "NODATA_PIXEL_PERCENTAGE": 0, "SENSING_ORBIT_DIRECTION": "DESCENDING", "GENERAL_QUALITY": "PASSED", "GRANULE_ID": "L2A_T35HKC_A033923_20211220T082827", "REFLECTANCE_CONVERSION_CORRECTION": 1.0326712354468, "MEDIUM_PROBA_CLOUDS_PERCENTAGE": 2.397248, "MEAN_INCIDENCE_AZIMUTH_ANGLE_B8": 186.357005411029, "DATATAKE_TYPE": "INS-NOBS", "MEAN_INCIDENCE_AZIMUTH_ANGLE_B9": 182.928253148734, "MEAN_INCIDENCE_AZIMUTH_ANGLE_B6": 183.3563508309, "MEAN_INCIDENCE_AZIMUTH_ANGLE_B7": 183.109691838117, "MEAN_INCIDENCE_AZIMUTH_ANGLE_B4": 184.132241490031, "MEAN_INCIDENCE_ZENITH_ANGLE_B1": 3.06615278603653, "NOT_VEGETATED_PERCENTAGE": 39.013728, "MEAN_INCIDENCE_AZIMUTH_ANGLE_B5": 183.682922936926, "RADIOMETRIC_QUALITY": "PASSED", "MEAN_INCIDENCE_AZIMUTH_ANGLE_B2": 187.678466476237, "MEAN_INCIDENCE_AZIMUTH_ANGLE_B3": 185.366615351959, "MEAN_INCIDENCE_ZENITH_ANGLE_B5": 2.72988386474952, "MEAN_INCIDENCE_AZIMUTH_ANGLE_B1": 183.002238919314, "MEAN_INCIDENCE_ZENITH_ANGLE_B4": 2.6563004578057, "MEAN_INCIDENCE_ZENITH_ANGLE_B3": 2.53509223726786, "MEAN_INCIDENCE_ZENITH_ANGLE_B2": 2.4284683919248, "MEAN_INCIDENCE_ZENITH_ANGLE_B9": 3.16193398356126, "MEAN_INCIDENCE_ZENITH_ANGLE_B8": 2.47809796571547, "MEAN_INCIDENCE_ZENITH_ANGLE_B7": 2.89096735679217, "DARK_FEATURES_PERCENTAGE": 2.535131, "HIGH_PROBA_CLOUDS_PERCENTAGE": 4.461143, "MEAN_INCIDENCE_ZENITH_ANGLE_B6": 2.80815177009197, "UNCLASSIFIED_PERCENTAGE": 1.75378, "MEAN_SOLAR_ZENITH_ANGLE": 26.2914700823507, "MEAN_INCIDENCE_ZENITH_ANGLE_B8A": 2.97743523242124, "RADIATIVE_TRANSFER_ACCURACY": 0, "MGRS_TILE": "35HKC", "CLOUDY_PIXEL_PERCENTAGE": 7.464998, "PRODUCT_ID": "S2A_MSIL2A_20211220T080341_N0301_R035_T35HKC_20211220T100229", "MEAN_INCIDENCE_ZENITH_ANGLE_B10": 2.61190347479859, "SOLAR_IRRADIANCE_B9": 812.92, "SNOW_ICE_PERCENTAGE": 0.00505, "DEGRADED_MSI_DATA_PERCENTAGE": 0, "MEAN_INCIDENCE_ZENITH_ANGLE_B11": 2.79184501565815, "MEAN_INCIDENCE_ZENITH_ANGLE_B12": 2.99404018653664, "SOLAR_IRRADIANCE_B6": 1287.61, "MEAN_INCIDENCE_AZIMUTH_ANGLE_B10": 183.209778422443, "SOLAR_IRRADIANCE_B5": 1424.64, "MEAN_INCIDENCE_AZIMUTH_ANGLE_B11": 182.210025828693, "SOLAR_IRRADIANCE_B8": 1041.63, "MEAN_INCIDENCE_AZIMUTH_ANGLE_B12": 181.730903604473, "SOLAR_IRRADIANCE_B7": 1162.08, "SOLAR_IRRADIANCE_B2": 1959.66, "SOLAR_IRRADIANCE_B1": 1884.69, "SOLAR_IRRADIANCE_B4": 1512.06, "GEOMETRIC_QUALITY": "PASSED", "SOLAR_IRRADIANCE_B3": 1823.24, "system:asset_size": 1735937022, "WATER_PERCENTAGE": 27.451491, "system:index": "20211220T080341_20211220T082827_T35HKC"}
stac :
{"description": "After 2022-01-25, Sentinel-2 scenes with PROCESSING_BASELINE '04.00' or above\nhave their DN (value) range shifted by 1000. The HARMONIZED collection\nshifts data in newer scenes to be in the same range as in older scenes.\n\nSentinel-2 is a wide-swath, high-resolution, multi-spectral\nimaging mission supporting Copernicus Land Monitoring studies,\nincluding the monitoring of vegetation, soil and water cover,\nas well as observation of inland waterways and coastal areas.\n\nThe Sentinel-2 L2 data are downloaded from CDSE. They were\ncomputed by running sen2cor. WARNING: 2017-2018 L2 coverage\nin the EE collection is not yet global.\n\nThe assets contain\n12 UINT16 spectral bands representing SR scaled by 10000 (unlike in L1 data,\nthere is no B10). There are also several more L2-specific bands (see band\nlist for details). See the\n[Sentinel-2 User Handbook](https://sentinel.esa.int/documents/247904/685211/Sentinel-2_User_Handbook)\nfor details.\n\nQA60 is a bitmask band that contained rasterized cloud mask\npolygons until 2022-01-25, when these polygons stopped being produced.\nStarting 2024-02-28, legacy-consistent QA60 bands are constructed from\nthe MSK_CLASSI cloud classification bands.\nFor more details,\n[see the full explanation of how cloud masks are computed.](https://sentinel.esa.int/web/sentinel/technical-guides/sentinel-2-msi/level-1c/cloud-masks)\n\nEE asset ids for Sentinel-2 L2 assets have the following format:\nCOPERNICUS/S2_SR/20151128T002653_20151128T102149_T56MNN. Here the\nfirst numeric part represents the sensing date and time, the\nsecond numeric part represents the product generation date and\ntime, and the final 6-character string is a unique granule identifier\nindicating its UTM grid reference (see [MGRS](https://en.wikipedia.org/wiki/Military_Grid_Reference_System)).\n\nFor datasets to assist with cloud and/or cloud shadow detection, see [COPERNICUS/S2_CLOUD_PROBABILITY](COPERNICUS_S2_CLOUD_PROBABILITY)\nand [GOOGLE/CLOUD_SCORE_PLUS/V1/S2_HARMONIZED](GOOGLE_CLOUD_SCORE_PLUS_V1_S2_HARMONIZED).\n\nFor more details on Sentinel-2 radiometric resolution, [see this page](https://sentinel.esa.int/web/sentinel/user-guides/sentinel-2-msi/resolutions/radiometric).\n", "extent": {"spatial": {"bbox": [[-180, -56, 180, 83]]}, "temporal": {"interval": [["2017-03-28T00:00:00Z", "2026-06-08T03:22:49.740000Z"]]}}, "gee:categories": ["satellite-imagery"], "gee:interval": {"interval": 5, "type": "revisit_interval", "unit": "day"}, "gee:terms_of_use": "The use of Sentinel data is governed by the [Copernicus\nSentinel Data Terms and Conditions.](https://scihub.copernicus.eu/twiki/pub/SciHubWebPortal/TermsConditions/Sentinel_Data_Terms_and_Conditions.pdf)\n", "gee:type": "image_collection", "id": "COPERNICUS/S2_SR_HARMONIZED", "keywords": ["copernicus", "esa", "eu", "msi", "reflectance", "satellite-imagery", "sentinel", "sr"], "license": "proprietary", "links": [{"href": "https://storage.googleapis.com/earthengine-stac/catalog/COPERNICUS/COPERNICUS_S2_SR_HARMONIZED.json", "rel": "self", "type": "application/json"}, {"href": "https://storage.googleapis.com/earthengine-stac/catalog/COPERNICUS/catalog.json", "rel": "parent", "type": "application/json"}, {"href": "https://storage.googleapis.com/earthengine-stac/catalog/catalog.json", "rel": "root", "type": "application/json"}, {"code": "JavaScript", "href": "https://code.earthengine.google.com/?scriptPath=Examples:Datasets/COPERNICUS/COPERNICUS_S2_SR_HARMONIZED", "rel": "related", "title": "Run the example for COPERNICUS/S2_SR_HARMONIZED in the Earth Engine Code Editor", "type": "text/html"}, {"href": "https://developers.google.com/earth-engine/datasets/images/COPERNICUS/COPERNICUS_S2_SR_HARMONIZED_sample.png", "rel": "preview", "type": "image/png"}, {"href": "https://developers.google.com/earth-engine/datasets/catalog/COPERNICUS_S2_SR_HARMONIZED#terms-of-use", "rel": "license", "type": "text/html"}, {"href": "https://storage.googleapis.com/earthengine-stac/catalog/COPERNICUS/COPERNICUS_S2_SR_HARMONIZED.json", "rel": "latest-version", "title": "COPERNICUS/S2_SR_HARMONIZED", "type": "application/json"}, {"href": "https://storage.googleapis.com/earthengine-stac/catalog/COPERNICUS/COPERNICUS_S2_SR.json", "rel": "predecessor-version", "title": "COPERNICUS/S2_SR", "type": "application/json"}], "providers": [{"name": "European Union/ESA/Copernicus", "roles": ["licensor", "producer"], "url": "https://sentinel.esa.int/web/sentinel/user-guides/sentinel-2-msi/processing-levels/level-2"}, {"name": "Google Earth Engine", "roles": ["host"], "url": "https://developers.google.com/earth-engine/datasets/catalog/COPERNICUS_S2_SR_HARMONIZED"}], "stac_extensions": ["https://stac-extensions.github.io/eo/v1.0.0/schema.json", "https://stac-extensions.github.io/version/v1.0.0/schema.json"], "stac_version": "1.0.0", "summaries": {"MSK_CLDPRB": {"gee:estimated_range": false, "maximum": 100, "minimum": 0}, "MSK_SNWPRB": {"gee:estimated_range": false, "maximum": 100, "minimum": 0}, "SCL": {"gee:estimated_range": false, "maximum": 11, "minimum": 1}, "eo:bands": [{"center_wavelength": 0.4439, "description": "Aerosols", "gee:scale": 0.0001, "gee:wavelength": "443.9nm (S2A) / 442.3nm (S2B)", "gsd": 60, "name": "B1"}, {"center_wavelength": 0.4966, "description": "Blue", "gee:scale": 0.0001, "gee:wavelength": "496.6nm (S2A) / 492.1nm (S2B)", "gsd": 10, "name": "B2"}, {"center_wavelength": 0.56, "description": "Green", "gee:scale": 0.0001, "gee:wavelength": "560nm (S2A) / 559nm (S2B)", "gsd": 10, "name": "B3"}, {"center_wavelength": 0.6645, "description": "Red", "gee:scale": 0.0001, "gee:wavelength": "664.5nm (S2A) / 665nm (S2B)", "gsd": 10, "name": "B4"}, {"center_wavelength": 0.7039, "description": "Red Edge 1", "gee:scale": 0.0001, "gee:wavelength": "703.9nm (S2A) / 703.8nm (S2B)", "gsd": 20, "name": "B5"}, {"center_wavelength": 0.7402, "description": "Red Edge 2", "gee:scale": 0.0001, "gee:wavelength": "740.2nm (S2A) / 739.1nm (S2B)", "gsd": 20, "name": "B6"}, {"center_wavelength": 0.7825, "description": "Red Edge 3", "gee:scale": 0.0001, "gee:wavelength": "782.5nm (S2A) / 779.7nm (S2B)", "gsd": 20, "name": "B7"}, {"center_wavelength": 0.8351, "description": "NIR", "gee:scale": 0.0001, "gee:wavelength": "835.1nm (S2A) / 833nm (S2B)", "gsd": 10, "name": "B8"}, {"center_wavelength": 0.8648, "description": "Red Edge 4", "gee:scale": 0.0001, "gee:wavelength": "864.8nm (S2A) / 864nm (S2B)", "gsd": 20, "name": "B8A"}, {"center_wavelength": 0.945, "description": "Water vapor", "gee:scale": 0.0001, "gee:wavelength": "945nm (S2A) / 943.2nm (S2B)", "gsd": 60, "name": "B9"}, {"center_wavelength": 1.6137, "description": "SWIR 1", "gee:scale": 0.0001, "gee:wavelength": "1613.7nm (S2A) / 1610.4nm (S2B)", "gsd": 20, "name": "B11"}, {"center_wavelength": 2.2024, "description": "SWIR 2", "gee:scale": 0.0001, "gee:wavelength": "2202.4nm (S2A) / 2185.7nm (S2B)", "gsd": 20, "name": "B12"}, {"description": "Aerosol Optical Thickness", "gee:scale": 0.001, "gsd": 10, "name": "AOT"}, {"description": "Water Vapor Pressure. The height the water would occupy if the vapor were condensed into\nliquid and spread evenly across the column.\n", "gee:scale": 0.001, "gee:units": "cm", "gsd": 10, "name": "WVP"}, {"description": "Scene Classification Map (The \"No Data\" value of 0 is masked out)", "gee:classes": [{"color": "ff0004", "description": "Saturated or defective", "value": 1}, {"color": "868686", "description": "Dark Area Pixels", "value": 2}, {"color": "774b0a", "description": "Cloud Shadows", "value": 3}, {"color": "10d22c", "description": "Vegetation", "value": 4}, {"color": "ffff52", "description": "Bare Soils", "value": 5}, {"color": "0000ff", "description": "Water", "value": 6}, {"color": "818181", "description": "Clouds Low Probability / Unclassified", "value": 7}, {"color": "c0c0c0", "description": "Clouds Medium Probability", "value": 8}, {"color": "f1f1f1", "description": "Clouds High Probability", "value": 9}, {"color": "bac5eb", "description": "Cirrus", "value": 10}, {"color": "52fff9", "description": "Snow / Ice", "value": 11}], "gsd": 20, "name": "SCL"}, {"description": "True Color Image, Red channel", "gsd": 10, "name": "TCI_R"}, {"description": "True Color Image, Green channel", "gsd": 10, "name": "TCI_G"}, {"description": "True Color Image, Blue channel", "gsd": 10, "name": "TCI_B"}, {"description": "Cloud Probability Map (missing in some products)", "gsd": 20, "name": "MSK_CLDPRB"}, {"description": "Snow Probability Map (missing in some products)", "gsd": 10, "name": "MSK_SNWPRB"}, {"description": "Always empty", "gsd": 10, "name": "QA10"}, {"description": "Always empty", "gsd": 20, "name": "QA20"}, {"description": "Cloud mask. Masked out between 2022-01-25 to 2024-02-28 inclusive.", "gee:bitmask": {"bitmask_parts": [{"bit_count": 10, "description": "Unused", "first_bit": 0}, {"bit_count": 1, "description": "Opaque clouds", "first_bit": 10, "values": [{"description": "No opaque clouds", "value": 0}, {"description": "Opaque clouds present", "value": 1}]}, {"bit_count": 1, "description": "Cirrus clouds", "first_bit": 11, "values": [{"description": "No cirrus clouds", "value": 0}, {"description": "Cirrus clouds present", "value": 1}]}], "total_bit_count": 12}, "gsd": 60, "name": "QA60"}, {"description": "Opaque clouds classification band (0=no clouds, 1=clouds). Masked out before February 2024.", "gsd": 60, "name": "MSK_CLASSI_OPAQUE"}, {"description": "Cirrus clouds classification band (0=no clouds, 1=clouds). Masked out before February 2024.", "gsd": 60, "name": "MSK_CLASSI_CIRRUS"}, {"description": "Snow/ice classification band (0=no snow/ice, 1=snow/ice). Masked out before February 2024.", "gsd": 60, "name": "MSK_CLASSI_SNOW_ICE"}], "gee:schema": [{"description": "Accuracy of Aerosol Optical thickness model", "name": "AOT_RETRIEVAL_ACCURACY", "type": "DOUBLE"}, {"description": "Granule-specific cloudy pixel percentage taken from the original metadata", "name": "CLOUDY_PIXEL_PERCENTAGE", "type": "DOUBLE"}, {"description": "Cloudy pixel percentage for the whole archive that\ncontains this granule. Taken from the original metadata\n", "name": "CLOUD_COVERAGE_ASSESSMENT", "type": "DOUBLE"}, {"description": "Percentage of pixels classified as cloud shadow", "name": "CLOUDY_SHADOW_PERCENTAGE", "type": "DOUBLE"}, {"description": "Percentage of pixels classified as dark features or shadows", "name": "DARK_FEATURES_PERCENTAGE", "type": "DOUBLE"}, {"description": "Unique identifier of the datastrip Product Data Item (PDI)", "name": "DATASTRIP_ID", "type": "STRING"}, {"description": "Uniquely identifies a given Datatake. The ID contains\nthe Sentinel-2 satellite, start date and time, absolute orbit\nnumber, and processing baseline.\n", "name": "DATATAKE_IDENTIFIER", "type": "STRING"}, {"description": "MSI operation mode", "name": "DATATAKE_TYPE", "type": "STRING"}, {"description": "Percentage of degraded MSI and ancillary data", "name": "DEGRADED_MSI_DATA_PERCENTAGE", "type": "DOUBLE"}, {"description": "Synthesis of the On-Line Quality Control (OLQC) checks\nperformed at granule (Product_Syntax) and datastrip (Product\nSyntax and DS_Consistency) levels\n", "name": "FORMAT_CORRECTNESS", "type": "STRING"}, {"description": "Synthesis of the OLQC checks performed at the datastrip level (Relative_Orbit_Number)", "name": "GENERAL_QUALITY", "type": "STRING"}, {"description": "Product generation time", "name": "GENERATION_TIME", "type": "DOUBLE"}, {"description": "Synthesis of the OLQC checks performed at the datastrip level (Attitude_Quality_Indicator)", "name": "GEOMETRIC_QUALITY", "type": "STRING"}, {"description": "Unique identifier of the granule PDI (PDI_ID)", "name": "GRANULE_ID", "type": "STRING"}, {"description": "Percentage of pixels classified as high probability clouds", "name": "HIGH_PROBA_CLOUDS_PERCENTAGE", "type": "DOUBLE"}, {"description": "Mean value containing viewing incidence azimuth angle average for band B1 and for all detectors", "name": "MEAN_INCIDENCE_AZIMUTH_ANGLE_B1", "type": "DOUBLE"}, {"description": "Mean value containing viewing incidence azimuth angle average for band B2 and for all detectors", "name": "MEAN_INCIDENCE_AZIMUTH_ANGLE_B2", "type": "DOUBLE"}, {"description": "Mean value containing viewing incidence azimuth angle average for band B3 and for all detectors", "name": "MEAN_INCIDENCE_AZIMUTH_ANGLE_B3", "type": "DOUBLE"}, {"description": "Mean value containing viewing incidence azimuth angle average for band B4 and for all detectors", "name": "MEAN_INCIDENCE_AZIMUTH_ANGLE_B4", "type": "DOUBLE"}, {"description": "Mean value containing viewing incidence azimuth angle average for band B5 and for all detectors", "name": "MEAN_INCIDENCE_AZIMUTH_ANGLE_B5", "type": "DOUBLE"}, {"description": "Mean value containing viewing incidence azimuth angle average for band B6 and for all detectors", "name": "MEAN_INCIDENCE_AZIMUTH_ANGLE_B6", "type": "DOUBLE"}, {"description": "Mean value containing viewing incidence azimuth angle average for band B7 and for all detectors", "name": "MEAN_INCIDENCE_AZIMUTH_ANGLE_B7", "type": "DOUBLE"}, {"description": "Mean value containing viewing incidence azimuth angle average for band B8 and for all detectors", "name": "MEAN_INCIDENCE_AZIMUTH_ANGLE_B8", "type": "DOUBLE"}, {"description": "Mean value containing viewing incidence azimuth angle average for band B8a and for all detectors", "name": "MEAN_INCIDENCE_AZIMUTH_ANGLE_B8A", "type": "DOUBLE"}, {"description": "Mean value containing viewing incidence azimuth angle average for band B9 and for all detectors", "name": "MEAN_INCIDENCE_AZIMUTH_ANGLE_B9", "type": "DOUBLE"}, {"description": "Mean value containing viewing incidence azimuth angle average for band B10 and for all detectors", "name": "MEAN_INCIDENCE_AZIMUTH_ANGLE_B10", "type": "DOUBLE"}, {"description": "Mean value containing viewing incidence azimuth angle average for band B11 and for all detectors", "name": "MEAN_INCIDENCE_AZIMUTH_ANGLE_B11", "type": "DOUBLE"}, {"description": "Mean value containing viewing incidence azimuth angle average for band B12 and for all detectors", "name": "MEAN_INCIDENCE_AZIMUTH_ANGLE_B12", "type": "DOUBLE"}, {"description": "Mean value containing viewing incidence zenith angle average for band B1 and for all detectors", "name": "MEAN_INCIDENCE_ZENITH_ANGLE_B1", "type": "DOUBLE"}, {"description": "Mean value containing viewing incidence zenith angle average for band B2 and for all detectors", "name": "MEAN_INCIDENCE_ZENITH_ANGLE_B2", "type": "DOUBLE"}, {"description": "Mean value containing viewing incidence zenith angle average for band B3 and for all detectors", "name": "MEAN_INCIDENCE_ZENITH_ANGLE_B3", "type": "DOUBLE"}, {"description": "Mean value containing viewing incidence zenith angle average for band B4 and for all detectors", "name": "MEAN_INCIDENCE_ZENITH_ANGLE_B4", "type": "DOUBLE"}, {"description": "Mean value containing viewing incidence zenith angle average for band B5 and for all detectors", "name": "MEAN_INCIDENCE_ZENITH_ANGLE_B5", "type": "DOUBLE"}, {"description": "Mean value containing viewing incidence zenith angle average for band B6 and for all detectors", "name": "MEAN_INCIDENCE_ZENITH_ANGLE_B6", "type": "DOUBLE"}, {"description": "Mean value containing viewing incidence zenith angle average for band B7 and for all detectors", "name": "MEAN_INCIDENCE_ZENITH_ANGLE_B7", "type": "DOUBLE"}, {"description": "Mean value containing viewing incidence zenith angle average for band B8 and for all detectors", "name": "MEAN_INCIDENCE_ZENITH_ANGLE_B8", "type": "DOUBLE"}, {"description": "Mean value containing viewing incidence zenith angle average for band B8a and for all detectors", "name": "MEAN_INCIDENCE_ZENITH_ANGLE_B8A", "type": "DOUBLE"}, {"description": "Mean value containing viewing incidence zenith angle average for band B9 and for all detectors", "name": "MEAN_INCIDENCE_ZENITH_ANGLE_B9", "type": "DOUBLE"}, {"description": "Mean value containing viewing incidence zenith angle average for band B10 and for all detectors", "name": "MEAN_INCIDENCE_ZENITH_ANGLE_B10", "type": "DOUBLE"}, {"description": "Mean value containing viewing incidence zenith angle average for band B11 and for all detectors", "name": "MEAN_INCIDENCE_ZENITH_ANGLE_B11", "type": "DOUBLE"}, {"description": "Mean value containing viewing incidence zenith angle average for band B12 and for all detectors", "name": "MEAN_INCIDENCE_ZENITH_ANGLE_B12", "type": "DOUBLE"}, {"description": "Mean value containing sun azimuth angle average for all bands and detectors", "name": "MEAN_SOLAR_AZIMUTH_ANGLE", "type": "DOUBLE"}, {"description": "Mean value containing sun zenith angle average for all bands and detectors", "name": "MEAN_SOLAR_ZENITH_ANGLE", "type": "DOUBLE"}, {"description": "Percentage of pixels classified as medium probability clouds", "name": "MEDIUM_PROBA_CLOUDS_PERCENTAGE", "type": "DOUBLE"}, {"description": "US-Military Grid Reference System (MGRS) tile", "name": "MGRS_TILE", "type": "STRING"}, {"description": "Percentage of No Data pixels", "name": "NODATA_PIXEL_PERCENTAGE", "type": "DOUBLE"}, {"description": "Percentage of pixels classified as non-vegetated", "name": "NOT_VEGETATED_PERCENTAGE", "type": "DOUBLE"}, {"description": "Configuration baseline used at the time of the product\ngeneration in terms of processor software version and major Ground\nImage Processing Parameters (GIPP) version\n", "name": "PROCESSING_BASELINE", "type": "STRING"}, {"description": "The full id of the original Sentinel-2 product", "name": "PRODUCT_ID", "type": "STRING"}, {"description": "Accuracy of radiative transfer model", "name": "RADIATIVE_TRANSFER_ACCURACY", "type": "DOUBLE"}, {"description": "Based on the OLQC reports contained in the Datastrips/QI_DATA with RADIOMETRIC_QUALITY checklist name", "name": "RADIOMETRIC_QUALITY", "type": "STRING"}, {"description": "Earth-Sun distance correction factor", "name": "REFLECTANCE_CONVERSION_CORRECTION", "type": "DOUBLE"}, {"description": "Percentage of saturated or defective pixels", "name": "SATURATED_DEFECTIVE_PIXEL_PERCENTAGE", "type": "DOUBLE"}, {"description": "Imaging orbit direction", "name": "SENSING_ORBIT_DIRECTION", "type": "STRING"}, {"description": "Imaging orbit number", "name": "SENSING_ORBIT_NUMBER", "type": "DOUBLE"}, {"description": "Synthesis of the OLQC checks performed at granule\n(Missing_Lines, Corrupted_ISP, and Sensing_Time) and datastrip\n(Degraded_SAD and Datation_Model) levels\n", "name": "SENSOR_QUALITY", "type": "STRING"}, {"description": "Mean solar exoatmospheric irradiance for band B1", "name": "SOLAR_IRRADIANCE_B1", "type": "DOUBLE"}, {"description": "Mean solar exoatmospheric irradiance for band B2", "name": "SOLAR_IRRADIANCE_B2", "type": "DOUBLE"}, {"description": "Mean solar exoatmospheric irradiance for band B3", "name": "SOLAR_IRRADIANCE_B3", "type": "DOUBLE"}, {"description": "Mean solar exoatmospheric irradiance for band B4", "name": "SOLAR_IRRADIANCE_B4", "type": "DOUBLE"}, {"description": "Mean solar exoatmospheric irradiance for band B5", "name": "SOLAR_IRRADIANCE_B5", "type": "DOUBLE"}, {"description": "Mean solar exoatmospheric irradiance for band B6", "name": "SOLAR_IRRADIANCE_B6", "type": "DOUBLE"}, {"description": "Mean solar exoatmospheric irradiance for band B7", "name": "SOLAR_IRRADIANCE_B7", "type": "DOUBLE"}, {"description": "Mean solar exoatmospheric irradiance for band B8", "name": "SOLAR_IRRADIANCE_B8", "type": "DOUBLE"}, {"description": "Mean solar exoatmospheric irradiance for band B8a", "name": "SOLAR_IRRADIANCE_B8A", "type": "DOUBLE"}, {"description": "Mean solar exoatmospheric irradiance for band B9", "name": "SOLAR_IRRADIANCE_B9", "type": "DOUBLE"}, {"description": "Mean solar exoatmospheric irradiance for band B10", "name": "SOLAR_IRRADIANCE_B10", "type": "DOUBLE"}, {"description": "Mean solar exoatmospheric irradiance for band B11", "name": "SOLAR_IRRADIANCE_B11", "type": "DOUBLE"}, {"description": "Mean solar exoatmospheric irradiance for band B12", "name": "SOLAR_IRRADIANCE_B12", "type": "DOUBLE"}, {"description": "Percentage of pixels classified as snow or ice", "name": "SNOW_ICE_PERCENTAGE", "type": "DOUBLE"}, {"description": "Sentinel-2 spacecraft name: Sentinel-2A, Sentinel-2B", "name": "SPACECRAFT_NAME", "type": "STRING"}, {"description": "Percentage of pixels classified as thin cirrus clouds", "name": "THIN_CIRRUS_PERCENTAGE", "type": "DOUBLE"}, {"description": "Percentage of unclassified pixels", "name": "UNCLASSIFIED_PERCENTAGE", "type": "DOUBLE"}, {"description": "Percentage of pixels classified as vegetation", "name": "VEGETATION_PERCENTAGE", "type": "DOUBLE"}, {"description": "Percentage of pixels classified as water", "name": "WATER_PERCENTAGE", "type": "DOUBLE"}, {"description": "Declared accuracy of the Water Vapor model", "name": "WATER_VAPOUR_RETRIEVAL_ACCURACY", "type": "DOUBLE"}], "gee:visualizations": [{"display_name": "RGB", "image_visualization": {"band_vis": {"bands": ["B4", "B3", "B2"], "max": [3000], "min": [0]}}, "lookat": {"lat": 17.7009, "lon": 83.277, "zoom": 12}}], "instruments": ["MSI"], "platform": ["Sentinel-2A", "Sentinel-2B"]}, "title": "Harmonized Sentinel-2 MSI: MultiSpectral Instrument, Level-2A (SR)", "type": "Collection", "version": "harmonized", "gee:status": "ready"}

Masking¶

Both ee.Image.gd.toXarray() and ee.ImageCollection.gd.toXarray() have a masked parameter that controls whether exported masked pixels are set to nodata, or to NaN. If they are set to NaN, the export data type will be converted to a floating point type able to represent the data:

# create and prepare a cloud masked image
im = ee.Image('COPERNICUS/S2_SR_HARMONIZED/20211220T080341_20211220T082827_T35HKC')
im = im.gd.addMaskBands().gd.maskClouds()
region = ee.Geometry.Rectangle(24.35, -33.75, 24.45, -33.65)
prep_im = im.gd.prepareForExport(
    region=region, scale=30, dtype='uint16', bands=['B4', 'B3', 'B2']
)

# export, setting masked pixels to NaN
da = prep_im.gd.toXarray(masked=True)

# check for NaN pixels and floating point data type
print(da.isnull().any())
print(da.dtype)

<xarray.DataArray ()> Size: 1B
True
Attributes: (7)
float32

See the Xarray documentation on missing values for background.

Attributes¶

DataArray / Dataset attributes include crs, transform and nodata values for compatibility with rioxarray, as well as ee and stac JSON strings of the Earth Engine property and STAC dictionaries.

Google cloud¶

Image¶

ee.Image.gd.toGoogleCloud() exports an image to Google Drive, Earth Engine asset or Google Cloud Storage:

# create and prepare image
im = ee.Image('COPERNICUS/S2_SR_HARMONIZED/20211220T080341_20211220T082827_T35HKC')
region = ee.Geometry.Rectangle(24.35, -33.75, 24.45, -33.65)
prep_im = im.gd.prepareForExport(region=region, scale=30, dtype='uint16')

# export to Earth Engine asset 's2' in the 'geedim' project, waiting for completion
_ = prep_im.gd.toGoogleCloud('s2', type='asset', folder='geedim', wait=True)

# display asset image info
ee.Image('projects/geedim/assets/s2').getInfo()

{'type': 'Image', 'bands': [{'id': 'B1', 'data_type': {'type': 'PixelType', ...

Collection¶

ee.ImageCollection.gd.toGoogleCloud() exports a collection to Google Drive, Earth Engine asset or Google Cloud Storage. The split parameter controls whether exported files / assets correspond to collection bands or images:

# create and prepare a collection (with two images and three bands)
coll = ee.ImageCollection('COPERNICUS/S2_SR_HARMONIZED')
region = ee.Geometry.Rectangle(24.35, -33.75, 24.45, -33.65)
coll = coll.filterBounds(region).limit(2)
prep_coll = coll.gd.prepareForExport(
    region=region, scale=30, dtype='uint16', bands=['B4', 'B3', 'B2']
)

# export to Earth Engine assets in the 'geedim' project, waiting for completion
# (one asset for each collection band)
_ = prep_coll.gd.toGoogleCloud(type='asset', folder='geedim', wait=True, split='bands')

# display the info of the first asset image
ee.Image('projects/geedim/assets/B4').getInfo()

{'type': 'Image', 'bands': [{'id': 'B_20180510T075611_20180510T082300_T35HKC', ...

Additional arguments¶

Depending on the type parameter, toGoogleCloud() calls one of the Export.image.toDrive(), Export.image.toAsset() or Export.image.toCloudStorage() Earth Engine functions to perform the export. ee.Image.gd.toGoogleCloud() and ee.ImageCollection.gd.toGoogleCloud() allow additional keyword arguments to be passed to the type relevant Earth Engine function. See the Export.image.toDrive(), Export.image.toAsset() or Export.image.toCloudStorage() docs for supported parameters. E.g.

# export to Google Drive using the TFRecord format
_ = prep_im.gd.toGoogleCloud(
    's2',
    type='drive',
    folder='geedim',
    fileFormat='TFRecord',
    formatOptions={'patchDimensions': [256, 256], 'compressed': True},
)

Tiling¶

The toGeoTIFF(), toNumPy() and toXarray() methods divide images into tiles for export. Tiles are downloaded and decompressed concurrently, then reassembled into the target export format. Tile size can be controlled with the max_tile_size, max_tile_dim and max_tile_bands parameters. Download concurrency can be controlled with the max_requests, and decompress concurrency with the max_cpus parameter. Each parameter has an upper limit - see the toGeoTIFF(), toNumPy() or toXarray() reference docs for details. For most uses, the tiling parameters can be left on their default values.

User memory limit error¶

Exporting computed images with toGeoTIFF(), toNumPy() or toXarray() could raise a 'User memory limit exceeded' error in some unusual cases. Earth Engine raises this error if a computation exceeds the limit on user memory. E.g.:

# create a 2 year cloud-free median composite
coll = ee.ImageCollection('COPERNICUS/S2_SR_HARMONIZED')
region = ee.Geometry.Rectangle(24.35, -33.75, 24.45, -33.65)
coll = coll.gd.filter('2021-01-01', '2023-01-01', region=region)
comp_im = coll.gd.composite(method='median')

# prepare the composite for export
prep_im = comp_im.gd.prepareForExport(
    crs='EPSG:3857', region=region, scale=10, dtype='uint16'
)

# attempt export to NumPy array
array = prep_im.gd.toNumPy()

---------------------------------------------------------------------------
ClientResponseError                       Traceback (most recent call last)
Cell In[23], line 13
      9     crs='EPSG:3857', region=region, scale=10, dtype='uint16'
     10 )
     11 
     12 # attempt export to NumPy array
---> 13 array = prep_im.gd.toNumPy()

File ~/checkouts/readthedocs.org/user_builds/geedim/envs/latest/lib/python3.14/site-packages/geedim/image.py:1165, in ImageAccessor.toNumPy(self, masked, structured, max_tile_size, max_tile_dim, max_tile_bands, max_requests, max_cpus)
   1155     array[tile.slices.row, tile.slices.col, tile.slices.band] = tile_array
   1157 with Tiler(
   1158     self,
   1159     max_tile_size=max_tile_size,
   (...)   1163     max_cpus=max_cpus,
   1164 ) as tiler:
-> 1165     tiler.map_tiles(write_tile, masked=masked)
   1167 if structured:
   1168     dtype = np.dtype(
   1169         dict(names=self.bandNames, formats=[self.dtype] * len(self.bandNames))
   1170     )

File ~/checkouts/readthedocs.org/user_builds/geedim/envs/latest/lib/python3.14/site-packages/geedim/tile.py:426, in Tiler.map_tiles(self, func, masked)
    424 # download tiles using persistent session
    425 runner = utils.AsyncRunner()
--> 426 runner.run(_map_tiles(runner.session))

File ~/checkouts/readthedocs.org/user_builds/geedim/envs/latest/lib/python3.14/site-packages/geedim/utils.py:389, in AsyncRunner.run(self, coro, **kwargs)
    383 if loop and loop.is_running():
    384     # run in a separate thread if there is an existing loop (e.g. we are in a
    385     # jupyter notebook)
    386     self._executor = self._executor or ThreadPoolExecutor(max_workers=1)
    387     return self._executor.submit(
    388         lambda: self._runner.run(coro, **kwargs)
--> 389     ).result()
    390 else:
    391     # run in this thread if there is no existing loop
    392     return self._runner.run(coro, **kwargs)

File ~/.asdf/installs/python/3.14.0/lib/python3.14/concurrent/futures/_base.py:450, in Future.result(self, timeout)
    448     raise CancelledError()
    449 elif self._state == FINISHED:
--> 450     return self.__get_result()
    451 else:
    452     raise TimeoutError()

File ~/.asdf/installs/python/3.14.0/lib/python3.14/concurrent/futures/_base.py:395, in Future.__get_result(self)
    393 if self._exception is not None:
    394     try:
--> 395         raise self._exception
    396     finally:
    397         # Break a reference cycle with the exception in self._exception
    398         self = None

File ~/.asdf/installs/python/3.14.0/lib/python3.14/concurrent/futures/thread.py:86, in _WorkItem.run(self, ctx)
     83     return
     85 try:
---> 86     result = ctx.run(self.task)
     87 except BaseException as exc:
     88     self.future.set_exception(exc)

File ~/.asdf/installs/python/3.14.0/lib/python3.14/concurrent/futures/thread.py:73, in WorkerContext.run(self, task)
     71 def run(self, task):
     72     fn, args, kwargs = task
---> 73     return fn(*args, **kwargs)

File ~/checkouts/readthedocs.org/user_builds/geedim/envs/latest/lib/python3.14/site-packages/geedim/utils.py:388, in AsyncRunner.run.<locals>.<lambda>()
    383 if loop and loop.is_running():
    384     # run in a separate thread if there is an existing loop (e.g. we are in a
    385     # jupyter notebook)
    386     self._executor = self._executor or ThreadPoolExecutor(max_workers=1)
    387     return self._executor.submit(
--> 388         lambda: self._runner.run(coro, **kwargs)
    389     ).result()
    390 else:
    391     # run in this thread if there is no existing loop
    392     return self._runner.run(coro, **kwargs)

File ~/.asdf/installs/python/3.14.0/lib/python3.14/asyncio/runners.py:127, in Runner.run(self, coro, context)
    125 self._interrupt_count = 0
    126 try:
--> 127     return self._loop.run_until_complete(task)
    128 except exceptions.CancelledError:
    129     if self._interrupt_count > 0:

File ~/.asdf/installs/python/3.14.0/lib/python3.14/asyncio/base_events.py:719, in BaseEventLoop.run_until_complete(self, future)
    716 if not future.done():
    717     raise RuntimeError('Event loop stopped before Future completed.')
--> 719 return future.result()

File ~/checkouts/readthedocs.org/user_builds/geedim/envs/latest/lib/python3.14/site-packages/geedim/tile.py:414, in Tiler.map_tiles.<locals>._map_tiles(session)
    410     with utils.auto_leave_tqdm(
    411         asyncio.as_completed(tasks), total=len(tasks), **tqdm_kwargs
    412     ) as bar:
    413         for task in bar:
--> 414             await task
    416 except:
    417     # cancel and await any incomplete tasks
    418     logger.debug('Cleaning up export tasks...')

File ~/.asdf/installs/python/3.14.0/lib/python3.14/asyncio/tasks.py:618, in _AsCompletedIterator._wait_for_one(self, resolve)
    615 if f is None:
    616     # Dummy value from _handle_timeout().
    617     raise exceptions.TimeoutError
--> 618 return f.result() if resolve else f

File ~/checkouts/readthedocs.org/user_builds/geedim/envs/latest/lib/python3.14/site-packages/geedim/tile.py:356, in Tiler._map_tile(self, func, tile, masked, session, max_retries, backoff_factor)
    354     url = await loop.run_in_executor(self._executor, get_tile_url)
    355     logger.debug(f'Downloading {tile!r} from {url}.')
--> 356     mem_file = await download_url(url)
    358 # enter memory file context (it must be closed when done)
    359 with mem_file:
    360     # limit concurrent tile reads to leave CPU capacity for the event
    361     # loop

File ~/checkouts/readthedocs.org/user_builds/geedim/envs/latest/lib/python3.14/site-packages/geedim/tile.py:337, in Tiler._map_tile.<locals>.download_url(url)
    335     except aiohttp.ClientError:
    336         pass
--> 337     response.raise_for_status()
    339 async for data, _ in response.content.iter_chunks():
    340     mem_file.write(data)

File ~/checkouts/readthedocs.org/user_builds/geedim/envs/latest/lib/python3.14/site-packages/aiohttp/client_reqrep.py:655, in ClientResponse.raise_for_status(self)
    652 if not self._in_context:
    653     self.release()
--> 655 raise ClientResponseError(
    656     self.request_info,
    657     self.history,
    658     status=self.status,
    659     message=self.reason,
    660     headers=self.headers,
    661 )

ClientResponseError: 400, message='User memory limit exceeded.', url='https://earthengine-highvolume.googleapis.com/v1/projects/geedim/thumbnails/313e551072ecb05467cfc9af33db22ea-184799ab3aa113cb044b4fbf6885bed6:getPixels'

toGoogleCloud() is not subject to the limit and using it for export is recommended in this situation. Images can first be exported to Earth Engine asset with toGoogleCloud(), and then the computed assets exported to a target format with one of toGeoTIFF(), toNumPy() or toXarray(). E.g.:

# export composite to Earth Engine asset 's2-comp' in the 'geedim' project
_ = prep_im.gd.toGoogleCloud('s2-comp', type='asset', folder='geedim', wait=True)

# export the asset to a NumPy array
array = ee.Image('projects/geedim/assets/s2-comp').gd.toNumPy()