Discover and mosaic Sentinel-2 tiles¶

In [1]:

%load_ext autoreload
%autoreload 2

import geopandas as gpd
import folium
import folium.plugins
import logging
from eo_tools.util import explore_products
from eo_tools.S2 import process_s2_tiles, make_s2_color
import rioxarray as riox


from eodag import EODataAccessGateway

# credentials need to be stored in the following file (see EODAG docs)
confpath = "/data/eodag_config.yml"
dag = EODataAccessGateway(user_conf_file_path=confpath)
# make sure cop_dataspace will be used 
dag.set_preferred_provider("cop_dataspace")
logging.basicConfig(level=logging.INFO)

%load_ext autoreload %autoreload 2 import geopandas as gpd import folium import folium.plugins import logging from eo_tools.util import explore_products from eo_tools.S2 import process_s2_tiles, make_s2_color import rioxarray as riox from eodag import EODataAccessGateway # credentials need to be stored in the following file (see EODAG docs) confpath = "/data/eodag_config.yml" dag = EODataAccessGateway(user_conf_file_path=confpath) # make sure cop_dataspace will be used dag.set_preferred_provider("cop_dataspace") logging.basicConfig(level=logging.INFO)

Directory setup¶

In [ ]:

# change to your data directories
path_data = "/data"
path_s2 = f"{path_data}/S2"
path_processed = f"{path_data}/res"

# change to your data directories path_data = "/data" path_s2 = f"{path_data}/S2" path_processed = f"{path_data}/res"

Search¶

In [ ]:

# load a geometry
aoi_file = "../data/Bretagne_AOI.geojson"
shp = gpd.read_file(aoi_file).geometry[0]
search_criteria = {
    "productType": "S2_MSI_L1C",
    "start": "2023-06-01",
    "end": "2023-08-31",
    "cloudCover": 10,
    "geom": shp
}
results = dag.search_all(**search_criteria)

# load a geometry aoi_file = "../data/Bretagne_AOI.geojson" shp = gpd.read_file(aoi_file).geometry[0] search_criteria = { "productType": "S2_MSI_L1C", "start": "2023-06-01", "end": "2023-08-31", "cloudCover": 10, "geom": shp } results = dag.search_all(**search_criteria)

Getting unique dates to help with product exploration

In [ ]:

def get_unique_dates(results):
    import numpy as np

    dates = []
    for p in results:
        dates.append(p.properties["startTimeFromAscendingNode"])
    return np.unique(dates).tolist()

dates = get_unique_dates(results)
print(f"unique dates:")
print(", \n".join(dates))

def get_unique_dates(results): import numpy as np dates = [] for p in results: dates.append(p.properties["startTimeFromAscendingNode"]) return np.unique(dates).tolist() dates = get_unique_dates(results) print(f"unique dates:") print(", \n".join(dates))

Visualize product footprints on a map¶

In [ ]:

# show all results
# explore_products(results, shp)

# filter by date
results_filtered = results.filter_date("2023-06-05", "2023-06-06") 
explore_products(results_filtered, shp)

# show all results # explore_products(results, shp) # filter by date results_filtered = results.filter_date("2023-06-05", "2023-06-06") explore_products(results_filtered, shp)

Download¶

In [ ]:

# re-run to get only links
dl = dag.download_all(results_filtered, output_dir=path_s2)

# re-run to get only links dl = dag.download_all(results_filtered, output_dir=path_s2)

Merge downloaded tiles¶

In [ ]:

out_dirs = process_s2_tiles(
    dl,
    bands=["B2", "B3", "B4", "B8"],
    shp=shp,
    aoi_name="aoi_bretagne",
    output_dir=path_processed,
)

out_dirs = process_s2_tiles( dl, bands=["B2", "B3", "B4", "B8"], shp=shp, aoi_name="aoi_bretagne", output_dir=path_processed, )

Make CIR color composite¶

In [ ]:

make_s2_color(out_dirs[0], name="CIR")

make_s2_color(out_dirs[0], name="CIR")

Visualize¶

In [17]:

pth_arr = f"{out_dirs[0]}/CIR.tif"
arr = riox.open_rasterio(pth_arr)
arr.plot.imshow(cmap="gray", vmax=200)

pth_arr = f"{out_dirs[0]}/CIR.tif" arr = riox.open_rasterio(pth_arr) arr.plot.imshow(cmap="gray", vmax=200)

Out[17]:

<matplotlib.image.AxesImage at 0x7f0f90fce390>