equi7grid-lite

No one will drive us from the paradise which Equi7Grid created for us

The equi7grid-lite package implements a user-friendly Python interface to interact with the Equi7Grid grid system.

equi7grid-lite is an unofficial Python implementation of Equi7Grid. With this package, users can convert geographic coordinates to Equi7Grid tiles and vice versa. This implementation differs from the official version in tree key ways:

• Quad-Tree Grid Splitting: Users are required to split the grid in a Quad-Tree fashion, meaning each grid level is divided into four tiles. For example, transitioning from level 1 to level 0 involves splitting each tile into four regular smaller tiles.

• Revised Grid ID Encoding: The grid ID is always encoded in meters, and the reference to the tile system (e.g., "T1", "T3", "T6") is removed. Instead, tiles are dynamically defined by the min_grid_size parameter. Here is a comparison between the original Equi7Grid and equi7grid-lite name conventions:

  • 'EU500M_E036N006T6' -> 'EU2560_E4521N3011'

  Where 'EU' is the Equi7Grid zone, '2560' is the min_grid_size, 'E4521' is the position in the x tile grid, and 'N3011' is the position in the y tile grid.

• Upper Bound Level: The maximum grid level is determined as the nearest lower distance to 2_500_000 meters. This threshold serves as a limit to create the Quad-Tree grid structure.

Please refer to the Equi7Grid repository for more information of the official implementation.

Installation

The equi7grid-lite package is available on PyPI and can be installed using pip:

pip install equi7grid-lite

Usage

The equi7grid-lite package provides a single class, Equi7Grid, which can be used to convert between geographic coordinates and Equi7Grid tiles.

from equi7grid_lite import Equi7Grid

grid_system = Equi7Grid(min_grid_size=2560)
# Equi7Grid(min_grid_size=2560)
# ----------------
# levels: 0, 1, ... , 7, 8
# zones: AN, NA, OC, SA, AF, EU, AS
# min_grid_size: 2560 meters
# max_grid_size: 1310720 meters

To convert between geographic coordinates and Equi7Grid tiles, use the lonlat2grid method.

lon, lat = -79.5, -5.49
grid_system.lonlat2grid(lon=lon, lat=lat)
#                  id        lon       lat          x          y zone level  land             geometry
#0  SA2560_E2009N2525 -79.507568 -5.485739  5144320.0  6465280.0   SA    Z0  True  POLYGON ((514560...

Use the grid2lonlat method to convert from Equi7Grid tile id to geographic coordinates.

grid_system.grid2lonlat(grid_id="SA2560_E2009N2525")
#                  id        lon       lat          x          y zone level  land             geometry
#0  SA2560_E2009N2525 -79.507568 -5.485739  5144320.0  6465280.0   SA    Z0  True  POLYGON ((514560...

The Equi7Grid class also provides a method to create a grid of Equi7Grid upper-level tiles that cover a given bounding box.

import geopandas as gpd

from equi7grid_lite import Equi7Grid

# Define a POLYGON geometry
world_filepath = gpd.datasets.get_path('naturalearth_lowres')
world = gpd.read_file(world_filepath)
country = world[world.name == "Peru"].geometry

# Create a grid of Equi7Grid tiles that cover the bounding box of the POLYGON geometry
grid = grid_system.create_grid(
    level=4,
    zone="SA",
    mask=country # Only include tiles that intersect the polygon
)

# Export the grid to a GeoDataFrame
grid.to_file("grid.shp")

By running create_grid with different levels, you can obtain its corresponding Equi7Grid Quad-Tree grid structure for any region.

Obtain the metadata of each Equi7Grid zone:

from equi7grid_lite import Equi7Grid

# Zones: SA, EU, AF, AS, NA, AU
Equi7Grid.SA

Each zone has the following attributes:

• id: The zone ID code.
• crs: The WKT representation of the CRS.
• geometry_geo: The geometry of the zone in EPSG:4326.
• geometry_equi7grid: The geometry of the zone in the Equi7Grid CRS.
• bbox_geo: The bounding box of the zone in EPSG:4326.
• bbox_equi7grid: The bounding box of the zone in the Equi7Grid CRS.
• landmasses_equi7grid: The landmasses of the zone in the Equi7Grid CRS.
• origin: The central meridian and the latitude of origin.

Use Equi7Grid with cubo

The equi7grid-lite package can be used in conjunction with the cubo to retrieve Earth Observation (EO) data.

import cubo
import matplotlib.pyplot as plt
import numpy as np
import rioxarray
from rasterio.enums import Resampling

from equi7grid_lite import Equi7Grid

# Initialize Equi7Grid system
grid_system = Equi7Grid(min_grid_size=2560)

# Specify the center coordinates
lon, lat = -122.4194, 37.7749

# Retrieve parameters for the CUBO request
cubo_parameters = grid_system.cubo_utm_parameters(lon=lon, lat=lat)

# Define the cube request using CUBO
da = cubo.create(
    lat=cubo_parameters["lat"],
    lon=cubo_parameters["lon"],
    collection="sentinel-2-l2a",  # Name of the STAC collection
    bands=["B04", "B03", "B02"],   # Bands to retrieve
    start_date="2021-08-01",       # Start date of the cube
    end_date="2021-10-30",         # End date of the cube
    edge_size=cubo_parameters["distance"] // 10,  # Distance in pixels
    resolution=10,                 # Pixel size of the cube (m)
    query={"eo:cloud_cover": {"lt": 50}}  # Query parameters
)

# Add the CRS to the cube
da = da.rio.write_crs(f"epsg:{da.attrs['epsg']}")
da = da.drop_vars("cubo:distance_from_center")

# Convert the cube to a dataset and compute median over time
image = da.to_dataset("band").median("time", skipna=True)

# Increase the resolution of the cube with Lanczos resampling
image_reprojected = image.rio.reproject(
    cubo_parameters["crs"],
    resolution=2.5,
    resampling=Resampling.lanczos
)

# Downsample the cube with nearest neighbor resampling
image_reprojected = image_reprojected.rio.reproject(
    cubo_parameters["crs"],
    resolution=10,
    resampling=Resampling.nearest
)

# Clip the cube to the specified polygon
composite_e7g = image_reprojected.rio.clip([cubo_parameters["polygon"]]).to_array()

# Save the images in UTM and E7G projections
composite_e7g.rio.to_raster("composite_e7g.tif")
image.to_array().rio.to_raster("composite_utm.tif")

License

This package is released under the MIT License. For more information, see the LICENSE file.

Contributing

Contributions are welcome! For bug reports or feature requests, please open an issue on GitHub. For contributions, please submit a pull request with a detailed description of the changes.

Citation

This is a simple adaptation of the Equi7Grid paper and code. If you use this package in your research, please consider citing the original Equi7Grid package and paper.

Package:

@software{bernhard_bm_2023_8252376,
  author       = {Bernhard BM and
                  Sebastian Hahn and
                  actions-user and
                  cnavacch and
                  Manuel Schmitzer and
                  shochsto and
                  Senmao Cao},
  title        = {TUW-GEO/Equi7Grid: v0.2.4},
  month        = aug,
  year         = 2023,
  publisher    = {Zenodo},
  version      = {v0.2.4},
  doi          = {10.5281/zenodo.8252376},
  url          = {https://doi.org/10.5281/zenodo.8252376}
}

Paper:

@article{BAUERMARSCHALLINGER201484,
title = {Optimisation of global grids for high-resolution remote sensing data},
journal = {Computers & Geosciences},
volume = {72},
pages = {84-93},
year = {2014},
issn = {0098-3004},
doi = {https://doi.org/10.1016/j.cageo.2014.07.005},
url = {https://www.sciencedirect.com/science/article/pii/S0098300414001629},
author = {Bernhard Bauer-Marschallinger and Daniel Sabel and Wolfgang Wagner},
keywords = {Remote sensing, High resolution, Big data, Global grid, Projection, Sampling, Equi7 Grid}
}