Get Sentinel-2 (Cloud Optimized Geotiffs) COG files from AWS S3
You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
 
 

5.8 KiB

s2froms3

Get Sentinel-2 Cloud Optimized Geotiffs (COG) files from AWS S3.

Installation

pip install s2froms3

Dependencies

mgrs, s3fs

Tutorial

The library is small and limited. It could be used to download Cloud Optimized Geotiff files from AWS S3.

First of all we should import the library:

import s2froms3

To know what you can download from S3 you can use:

for item in s2froms3.products.Properties:
    print(item.value)
    description = item.describe()
    for k, v in description.items():
        print(f'    {k}: {v}')
    print()

The previous code will print the available options to download COGs:

TCI
    resolution: 10
    title: True color image
    center wavelength: None

B01
    resolution: 60
    title: Band 1 (coastal)
    center wavelength: 0.4439

B02
    resolution: 10
    title: Band 2 (blue)
    center wavelength: 0.4966

B03
    resolution: 10
    title: Band 3 (green)
    center wavelength: 0.56

B04
    resolution: 10
    title: Band 4 (red)
    center wavelength: 0.6645

B05
    resolution: 20
    title: Band 5
    center wavelength: 0.7039

B06
    resolution: 20
    title: Band 6
    center wavelength: 0.7402

B07
    resolution: 20
    title: Band 7
    center wavelength: 0.7825

B08
    resolution: 10
    title: Band 8 (nir)
    center wavelength: 0.8351

B8A
    resolution: 20
    title: Band 8A
    center wavelength: 0.8648

B09
    resolution: 60
    title: Band 9
    center wavelength: 0.945

B11
    resolution: 20
    title: Band 11 (swir16)
    center wavelength: 1.6137

B12
    resolution: 20
    title: Band 12 (swir22)
    center wavelength: 2.22024

AOT
    resolution: 60
    title: Aerosol Optical Thickness (AOT)
    center wavelength: None

WVP
    resolution: 10
    title: Water Vapour (WVP)
    center wavelength: None

SCL
    resolution: 20
    title: Scene Classification Map (SCL)
    center wavelength: None

Once you know the different possibilities you can start to download COGs using the following:

import datetime as dt
from pathlib import Path

lon = 10 # Longitude of interest
lat = 10 # Latitude of interest
start_date = dt.date(2020, 8, 1) # Start date to search images
end_date = dt.date(2020, 8, 15) # End date to search images
what = ['B02', 'B03', 'B04'] # What we want to download
cc = 25 # Minimum cloud cover on each image, 25 is 25%
folder = Path('.') # Where the files will be downloaded

downloaded = s2froms3.download_S2(
    lon=lon,
    lat=lat,
    start_date=start_date,
    end_date=end_date,
    what=what,
    cloud_cover_le=cc,
    folder=folder
)

After a while, you will find several COG files in the same folder where you were running the code. The function above will return the S3 file paths of the downloaded files:

print(downloaded)

The previous code will show:

['sentinel-cogs/sentinel-s2-l2a-cogs/32/P/PS/2020/8/S2A_32PPS_20200804_0_L2A/B02.tif',
 'sentinel-cogs/sentinel-s2-l2a-cogs/32/P/PS/2020/8/S2A_32PPS_20200804_0_L2A/B03.tif',
 'sentinel-cogs/sentinel-s2-l2a-cogs/32/P/PS/2020/8/S2A_32PPS_20200804_0_L2A/B04.tif',
 'sentinel-cogs/sentinel-s2-l2a-cogs/32/P/PS/2020/8/S2B_32PPS_20200809_0_L2A/B02.tif',
 'sentinel-cogs/sentinel-s2-l2a-cogs/32/P/PS/2020/8/S2B_32PPS_20200809_0_L2A/B03.tif',
 'sentinel-cogs/sentinel-s2-l2a-cogs/32/P/PS/2020/8/S2B_32PPS_20200809_0_L2A/B04.tif',
 'sentinel-cogs/sentinel-s2-l2a-cogs/32/P/PS/2020/8/S2B_32PPS_20200812_0_L2A/B02.tif',
 'sentinel-cogs/sentinel-s2-l2a-cogs/32/P/PS/2020/8/S2B_32PPS_20200812_0_L2A/B03.tif',
 'sentinel-cogs/sentinel-s2-l2a-cogs/32/P/PS/2020/8/S2B_32PPS_20200812_0_L2A/B04.tif']

Sometimes the location you have in mind is in a border of a tile (COG) and the scene you want to analyze is not complete. You can check if the location is close to a border/corner using the following function:

from s2froms3 import point_in_tile

print(s2froms3.point_in_tile(2.98279, 39.73887))

The previous code will show


+ + + + + + + + + + + 
+                 O + 
+                   + 
+                   + 
+                   + 
+                   + 
+                   + 
+                   + 
+                   + 
+                   + 
+ + + + + + + + + + + 

This way you can better understand if, in this case, you will need to download also the COGs in the East, the Northeast and/or the North.

if you want to download also COG files adjacent to the target location you could use the also option. For instance, in the previous case you maybe also want to download the COGs located to the East (E), to the North (N) and to the Northeast (NE). You could do so with the following code:

lon, lat = 2.98279, 39.73887

downloaded = s2froms3.download_S2(
    lon=lon,
    lat=lat,
    start_date=start_date,
    end_date=end_date,
    what=what,
    cloud_cover_le=cc,
    folder=folder,
    also=['N', 'NE', 'E']
)

Now, you will get several COGs but also the COGs located to the N, NE and E.

You can see a practical example in the notebooks folder of this repo (in spanish and english).

TO-DO

  • Use threading to download the files (done ✔️).
  • Add entry point / cli command (I'm thinking about it. Let's see).

Issues / Feature requests

File an issue here or start a discussion here.

Code of conduct

Any interaction you have with me or others must be guided by the highest standards of politeness and respect.

Releases

version 0.2.0

  • Added function point_in_tile to check where is located the requested location within the COG tile.
  • Added ssl_option keyword in download_S2 function.
  • Use threading to download COG files.
  • Added also and download keywords in download_S2 function.

version 0.1.0

  • Initial version with basic machinery to download Sentinel-2 COGs from S3.