In this page, you will be able to consult the code documentation of the main modules of the library

geesarfetcher

The main module of interest, where you will find the fetch functions, for different areas.

geesarfetcher

geesarfetcher

geesarfetcher.fetch(top_left=None, bottom_right=None, coords=None, start_date: datetime = datetime.date(2021, 6, 20), end_date: datetime = datetime.date(2022, 6, 20), ascending: bool = True, orbit_number: Optional[object] = None, scale: int = 20, n_jobs: int = 8, verbose: int = 0)

Fetches SAR data in the form of a dictionnary with image data as well as timestamps

Parameters

top_left (tuple of float, optional) – Top left coordinates (lon, lat) of the Region
bottom_right (tuple of float, optional) – Bottom right coordinates (lon, lat) of the Region
coords (tuple of tuple of float or list of list of float, optional) – If top_left and bottom_right are not specified, we expect coords to be a list (resp. tuple) of the form [top_left, bottom_right] (resp. (top_left, bottom_right))
start_date (datetime.datetime, optional) – First date of the time interval
end_date (datetime.datetime, optional) – Last date of the time interval
ascending (boolean, optional) – The trajectory to use when selecting data
orbit_number (int or str, optional) –
The orbit number to restrict the download to. If provided with an integer, the S1 temporal stack is filtered using the provided orbit number. If provided with a string value, we expect one of these keywords:
- ”max” for the orbit number with the highest number of image in the stack
- ”min” for the orbit number with the smallest number of image in the stack
If None, then no filter over the orbit number is applied.
scale (int, optional) – Scale parameters of the getRegion() function. Defaulting at 20, change it to change the scale of the final data points. The highest, the lower the spatial resolution. Should be at least 10.
n_jobs (int, optional) – Set the parallelisation factor (number of threads) for the GEE data access process. Set to 1 if no parallelisation required.
verbose (int, optional) – Verbosity mode (0: No info, 1: Info, 2: Detailed info, with added timestamp)

Returns

Dictionnary with four keys:

"stacks"
4-D array containing db intensity measure (numpy.ndarray), (height, width, pol_count, time_series_length)

"coordinates"
3-D array containg coordinates where [:,:,0] provides access to latitude and [:,:,1] provides access to longitude, (numpy.ndarray), (height, width, 2)

"timestamps"
list of acquisition timestamps of size (time_series_length,) (list of str)

"metadata"
Dictionnary describing data for each axis of the stack and the coordinates as well as the properties (orbit number, slice, acquisition time…) of each image of the temporal stack

Return type

dict

geesarfetcher.fetch_and_save(save_dir: Optional[str] = None, top_left=None, bottom_right=None, coords=None, start_date: datetime = datetime.date(2021, 6, 20), end_date: datetime = datetime.date(2022, 6, 20), ascending: bool = True, orbit_number: Optional[object] = None, scale: int = 20, n_jobs: int = 8, verbose: int = 0)

Fetches SAR data by looping other each timestep and each generated subregion and saves extracted images as GeoTIFF in the supplied save_dir folder

Parameters

save_dir (str) – Path toward an existing directory where to save the images. If non-existing, an Exception is raised.
top_left (tuple of float, optional) – Top left coordinates (lon, lat) of the Region
bottom_right (tuple of float, optional) – Bottom right coordinates (lon, lat) of the Region
coords (tuple of tuple of float or list of list of float, optional) – If top_left and bottom_right are not specified, we expect coords to be a list (resp. tuple) of the form [top_left, bottom_right] (resp. (top_left, bottom_right))
start_date (datetime.datetime, optional) – First date of the time interval
end_date (datetime.datetime, optional) – Last date of the time interval
ascending (boolean, optional) – The trajectory to use when selecting data
orbit_number (int or str, optional) –
The orbit number to restrict the download to. If provided with an integer, the S1 temporal stack is filtered using the provided orbit number. If provided with a string value, we expect one of these keywords:
- ”max” for the orbit number with the highest number of image in the stack
- ”min” for the orbit number with the smallest number of image in the stack
If None, then no filter over the orbit number is applied.
scale (int, optional) – Scale parameters of the getRegion() function. Defaulting at 20, change it to change the scale of the final data points. The highest, the lower the spatial resolution. Should be at least 10.
n_jobs (int, optional) – Set the parallelisation factor (number of threads) for the GEE data access process. Set to 1 if no parallelisation required.
verbose (int, optional) – Verbosity mode (0: No info, 1: Info, 2: Detailed info, with added timestamp)

Returns

Dictionnary with four keys:

"stack"
4-D array containing db intensity measure (numpy.ndarray), (height, width, pol_count, time_series_length)

"coordinates"
3-D array containg coordinates where [:,:,0] provides access to latitude and [:,:,1] provides access to longitude, (numpy.ndarray), (height, width, 2)

"timestamps"
list of acquisition timestamps of size (time_series_length,) (list of str)

"metadata"
Dictionnary describing data for each axis of the stack, the coordinates as well as the properties (orbit number, slice, acquisition time….) of each image of the temporal stack

Return type

dict

geesarfetcher.fetch_point(coords, start_date: datetime = datetime.date(2021, 6, 20), end_date: datetime = datetime.date(2022, 6, 20), ascending: bool = True, orbit_number: Optional[object] = None, scale: int = 20, n_jobs: int = 8, verbose: int = 0)

Fetches SAR data from a single coordinate point in the form of a dictionnary with image data as well as timestamps

Parameters

coords (tuple of float) – Coordinates (lon, lat) of the point of interest
start_date (datetime.datetime, optional) – First date of the time interval
end_date (datetime.datetime, optional) – Last date of the time interval
ascending (boolean, optional) – The trajectory to use when selecting data
orbit_number (int or str, optional) –
The orbit number to restrict the download to. If provided with an integer, the S1 temporal stack is filtered using the provided orbit number. If provided with a string value, we expect one of these keywords:
- ”max” for the orbit number with the highest number of image in the stack
- ”min” for the orbit number with the smallest number of image in the stack
If None, then no filter over the orbit number is applied.
scale (int, optional) – Scale parameters of the getRegion() function. Defaulting at 20, change it to change the scale of the final data points. The highest, the lower the spatial resolution. Should be at least 10.
n_jobs (int, optional) – Set the parallelisation factor (number of threads) for the GEE data access process. Set to 1 if no parallelisation required.
verbose (int, optional) – Verbosity mode (0: No info, 1: Info, 2: Detailed info, with added timestamp)

Returns

Dictionnary with four keys:

"stack"

4-D array containing db intensity measure (numpy.ndarray),

(1, 1, pol_count, time_series_length)

"coordinates"
3-D array containg coordinates where [:,:,0] provides access to latitude and [:,:,1] provides access to longitude, (numpy.ndarray), (1, 1, 2)

"timestamps"
list of acquisition timestamps of size (time_series_length,) (list of str)

"metadata"
Dictionnary describing data for each axis of the stack and the coordinates as well as the properties (orbit number, slice, acquisition tim….) of each point of the temporal stack

Return type

dict

geesarfetcher.filter

geesarfetcher.filter.filter_sentinel1_data(start_date, end_date, geometry, orbit='ASCENDING', orbit_number=None)

Filters Sentinel-1 products to get images collected in interferometric wide swath mode (IW) and on i) a date range, ii) a geometry and iii) ascending or descending orbit.

Parameters

start_date (str) – str following the pattern 'yyyy-mm-dd' describing the start date of the time interval
end_date (str) – str following the pattern 'yyyy-mm-dd' describing the end date of the time interval
geometry (ee.Geometry) – Geometry object defining the area of process
orbit (str, optional) – Defines the orbit to set for the data retrieval process
Returns –
-------- –
ee.ImageCollection – Filtered ImageCollection left to be queried

geesarfetcher.fetcher

geesarfetcher.fetcher.fetch_sentinel1_data(start_date, end_date, geometry, scale, orbit='ASCENDING', orbit_number=None)

Retrieves and queries ImageCollection using input parameters and return data as a tuple of header and values.

Parameters

start_date (str) – str following the pattern 'yyyy-mm-dd' describing the start date of the time interval
end_date (str) – str following the pattern 'yyyy-mm-dd' describing the end date of the time interval
geometry (ee.Geometry) – Geometry object defining the area of process
scale (int) – Scale parameters of the getRegion() function. Defaulting at 20, change it to change the scale of the final data points. The highest, the lower the spatial resolution. Should be at least 10.
orbit (str, optional) – Defines the orbit to set for the data retrieval process
orbit_number (int or str, optional) –
The orbit number to restrict the download to. If provided with an integer, the S1 temporal stack is filtered using the provided orbit number. If provided with a string value, we expect one of these keywords:
- ”max” for the orbit number with the highest number of image in the stack
- ”min” for the orbit number with the smallest number of image in the stack
If None, then no filter over the orbit number is applied.

Returns

Returns a dictionnary of the properties of the Image retrieved from GEE.

Return type

dict

geesarfetcher.fetcher.fetch_sentinel1_properties(start_date, end_date, geometry, orbit='ASCENDING', orbit_number=None)

Retrieves and queries ImageCollection using input parameters and return Image properties.

Parameters

start_date (str) – str following the pattern 'yyyy-mm-dd' describing the start date of the time interval
end_date (str) – str following the pattern 'yyyy-mm-dd' describing the end date of the time interval
geometry (ee.Geometry) – Geometry object defining the area of process
orbit (str, optional) – Defines the orbit to set for the data retrieval process
orbit_number (int or str, optional) –
The orbit number to restrict the download to. If provided with an integer, the S1 temporal stack is filtered using the provided orbit number. If provided with a string value, we expect one of these keywords:
- ”max” for the orbit number with the highest number of image in the stack
- ”min” for the orbit number with the smallest number of image in the stack
If None, then no filter over the orbit number is applied.

Returns

val_header corresponds to the list of str describing the fields of the val array. The val array is a list of data records, each represented as a list of the same size as the val_header array.

Return type

tuple

geesarfetcher.utils

The utils module contains functions used to prepare data before fetching sar images.

geesarfetcher.utils.cmp_coords(a, b)

Given two coordinates dict a and b, compare which one is closer to the North-Eastern direction

Parameters

a (dict) – dict with keys "lon" and "lat"
b (dict) – dict with keys "lon" and "lat"

Returns

-1 if a > b, 1 if a < b, 0 if a == b

Return type

int

geesarfetcher.utils.get_date_interval_array(start_date, end_date, day_timedelta=1)

Initialize a list of days interval of size day_timedelta iteratively created between start_date and end_date.

Parameters

start_date (datetime.datetime) – first date time of the array
end_date (datetime.datetime) – last date of the array
day_timedelta (int) – size, in days, of every interval

geesarfetcher.utils.make_polygon(top_left, bottom_right)

Given two (lon, lat) coordinates of both the top left and bottom right corner of a polygon, return the list of corner coordinates of this polygon

Parameters

top_left (list of int or tuple of int) – Top Left coordinates of the polygon
bottom_right (list of int or tuple of int) – Bottom right coordinates of the polygon

Returns

2-D list of the 5 coordinates need to create a Rectangular Polygon [top_left, top_right, bottom_right, bottom_left, top_left].

Return type

list

geesarfetcher.utils.retrieve_max_pixel_count_from_pattern(error_str)

Given an input getRegion error from GEE, extract the provided points count.

Parameters: error_str (str) – the str text of the GEE error (e.g. the function caled on "ImageCollection.getRegion: Too many values: x points ..." will output x)
Returns: Returns the number of points specified in the input image
Return type: int

geesarfetcher.utils.tile_coordinates(total_count_of_pixels, coordinates, max_gee=1048576)

Given a coordinates array describing a Polygon, a count of pixes within that polygons, tiles this polygon into a grid a sub-Polygons where each sub-Polygon size matches the max_gee pixel count given as a parameter.

Parameters

total_count_of_pixels (int) – Total number of pixels of the designated area
coordinates (array of array of floats) – Can be a 5-sized list of every coordinates defining the polygon [[long1, lat1],[long2, lat1]...,[long1, lat1]] or a 2-sized list of coordinates defining the top left and bottom right corner of the Polygon [[long1, lat1],[long2, lat2]]
max_gee_threshold (int, optional) – Total number of points allowed for one data query. Default: 1048576

Returns

3-dimensional list of coordinates with pixel count inferior or equal to the maximum GEE threshold (shape: (number of images, number of coordinates per image, 2))

Return type

list