Sentinel
Sentinel 2 - AWS¶
L1C - JPEG2000¶
Warnings
JPEG2000 format is not Cloud Optimized, numerous GET requests will be needed to read such format
which could result in important cost.
bands: B01, B02, B03, B04, B05, B06, B07, B08, B09, B11, B12, B8A
import rasterio
from rio_tiler_pds.sentinel.aws import S2JP2Reader
# We use __enter__ context manager for the Reader.
# When creating the instance of `sentinel` the Reader will fetch the sentinel 2 TileInfo.json
# to retrieve the bounds of the dataset and other metadata available at `sentinel.tileInfo`.
with rasterio.Env(AWS_REQUEST_PAYER="requester"):
with S2JP2Reader("S2A_L1C_20170729_19UDP_0") as sentinel:
# bounds and metadata are derived from the tileInfo.json file stored with the JPEG2000
print(type(sentinel))
>>> <class 'rio_tiler_pds.sentinel.aws.sentinel2.S2L1CReader'>
print(type(sentinel.tileInfo))
>>> <class 'dict'>
print(type(sentinel.datageom))
>>> <class 'dict'>
print(sentinel.bands)
>>> ('B01',
'B02',
'B03',
'B04',
'B05',
'B06',
'B07',
'B08',
'B09',
'B11',
'B12',
'B8A')
print(sentinel.info(bands="B01").json(exclude_none=True))
>>> {
'bounds': [-69.98971880792764, 47.761069480166995, -68.86723101847079, 48.75300221903151],
'minzoom': 8,
'maxzoom': 14,
'band_metadata': [["B01", {}]],
'band_descriptions': [["B01", {}]],
'dtype': 'uint16',
'nodata_type': 'Nodata',
'colorinterp': ['gray']
}
print(sentinel.statistics(bands="B8A")["B8A"].json())
>>> {
'min': 1.0,
'max': 19753.0,
'mean': 2873.8173758756675,
'count': 653216.0,
'sum': 1877223491.0,
'std': 2680.2546389126283,
'median': 2919.0,
'majority': 117.0,
'minority': 9913.0,
'unique': 11767.0,
'histogram': [
[281576.0, 154185.0, 130600.0, 49758.0, 30001.0, 6851.0, 242.0, 1.0, 1.0, 1.0],
[1.0, 1976.2, 3951.4, 5926.6, 7901.8, 9877.0, 11852.2, 13827.4, 15802.6, 17777.8, 19753.0]
],
'valid_percent': 62.3,
'masked_pixels': 395360.0,
'valid_pixels': 653216.0,
'percentile_98': 9320.699999999953,
'percentile_2': 106.0
}
img = sentinel.tile(77, 89, 8, bands="B01")
assert img.data.shape == (1, 256, 256)
print(sentinel.point(-69.41, 48.25, bands=("B01", "B02")))
>> PointData(
array=masked_array(data=[1201, 843], mask=[False, False], fill_value=999999, dtype=uint16),
band_names=['B01', 'B02'],
coordinates=(-69.41, 48.25),
crs=CRS.from_epsg(4326),
assets=[
's3://sentinel-s2-l1c/tiles/19/U/DP/2017/7/29/0/B01.jp2',
's3://sentinel-s2-l1c/tiles/19/U/DP/2017/7/29/0/B02.jp2'
],
metadata={}
)
# Working with Expression
img = sentinel.tile(77, 89, 8, expression="B01/B02")
assert igm.data.shape == (1, 256, 256)
print(sentinel.point(-69.41, 48.25, expression="B01/B02"))
>> PointData(
array=masked_array(data=[1.424673784104389], mask=[False], fill_value=999999, dtype=float32),
band_names=['B01/B02'],
coordinates=(-69.41, 48.25),
crs=CRS.from_epsg(4326),
assets=[
's3://sentinel-s2-l1c/tiles/19/U/DP/2017/7/29/0/B01.jp2',
's3://sentinel-s2-l1c/tiles/19/U/DP/2017/7/29/0/B02.jp2'
],
metadata={}
)
L2A - JPEG2000¶
Warnings
JPEG2000 format is not Cloud Optimized, numerous GET requests will be needed to read such format
which could result in important cost.
bands: B01, B02, B03, B04, B05, B06, B07, B08, B09, B11, B12, B8A
Note: AOT, SCL, WVP bands are not supported.
import rasterio
from rio_tiler_pds.sentinel.aws import S2JP2Reader
with rasterio.Env(AWS_REQUEST_PAYER="requester"):
with S2JP2Reader("S2A_L2A_20170729_19UDP_0") as sentinel:
# bounds and metadata are derived from the tileInfo.json file stored with the JPEG2000
print(type(sentinel))
>>> <class 'rio_tiler_pds.sentinel.aws.sentinel2.S2L2AReader'>
print(type(sentinel.tileInfo))
>>> dict
print(type(sentinel.datageom))
>>> dict
print(sentinel.info(bands="B01").dict(exclude_none=True))
>>> {
"bounds": [-69.98831359398795, 47.7610811323474, -68.86723101847079, 48.75300225264652],
"minzoom": 8,
"maxzoom": 14,
"band_metadata": [["B01", {}]],
"band_descriptions": [["B01", ""]],
"dtype": "uint16",
"nodata_type": "Nodata",
"colorinterp": ["gray"]
}
COG (Only L2A available for now)¶
bands: B01, B02, B03, B04, B05, B06, B07, B08, B09, B11, B12, B8A
Note: AOT, SCL, WVP STAC assets are not supported.
Notes:
- the
B10band is the cirrus band and is only supported for L1C, not L2A. - the
sentinel-2-l2a-cogsis a public dataset, no AWS credential should be needed.
from rio_tiler_pds.sentinel.aws import S2COGReader
with S2COGReader("S2A_29RKH_20200219_0_L2A") as sentinel:
print(type(sentinel))
>>> <class 'rio_tiler_pds.sentinel.aws.sentinel2.S2L2ACOGReader'>
print(sentinel.bands)
>>> ('B01', 'B02', 'B03', 'B04', 'B05', 'B06', 'B07', 'B08', 'B09', 'B11', 'B12', 'B8A')
# bounds and metadata are derived from the STAC item stored with the COG
print(type(sentinel.stac_item))
>>> dict
Sentinel 1 - AWS¶
import rasterio
from rio_tiler_pds.sentinel.aws import S1L1CReader
with rasterio.Env(AWS_REQUEST_PAYER="requester"):
with S1L1CReader("S1A_IW_GRDH_1SDV_20180716T004042_20180716T004107_022812_02792A_FD5B") as sentinel:
print(sentinel.bands)
> ('vv', 'vh')
print(sentine.bounds)
> (75.605247, 9.225784, 78.203903, 11.190425)
print(type(sentinel.productInfo))
> dict
print(sentinel._get_band_url("vv"))
> 's3://sentinel-s1-l1c/GRD/2018/7/16/IW/DV/S1A_IW_GRDH_1SDV_20180716T004042_20180716T004107_022812_02792A_FD5B/measurement/iw-vv.tiff'
Important
For most dataset hosted on AWS you will need to have AWS credentials available in your environment.