Python

Reading the regions with geopandas, and assigning regions to data

The examples use geopandas. Code on this page is not run when the site is built.

Reading

GeoParquet, from a URL (needs fsspec and aiohttp) or a local file:

import geopandas as gpd

base = "https://australianantarcticdivision.github.io/measoshapes/data/v06"
regions = gpd.read_parquet(f"{base}/measo_regions06.parquet")
regions_ll = gpd.read_parquet(f"{base}/measo_regions06_ll.parquet")
regions.crs

FlatGeobuf or GeoJSON are read with GDAL (via pyogrio), which reads URLs directly:

regions = gpd.read_file(f"{base}/measo_regions06.fgb")

Assigning regions to data

geopandas works in the plane, which matches the regions’ definition (boundaries straight in longitude/latitude), so longitude/latitude points can be joined directly to measo_regions06_ll:

import pandas as pd

pts = pd.DataFrame({"lon": [147.3, 100.5, -10.5, -82.5, 171.5, -175.5],
                    "lat": [-50.2, -60.5, -70.5, -45.5, -65.5, -65.5]})
pts = gpd.GeoDataFrame(pts, geometry=gpd.points_from_xy(pts.lon, pts.lat), crs="OGC:CRS84")
gpd.sjoin(pts, regions_ll, how="left", predicate="within")[["lon", "lat", "name", "label"]]

     lon   lat name         sector       zone
0  147.3 -50.2  EIN     EastIndian   Northern
1  100.5 -60.5  CIA  CentralIndian  Antarctic
2  -10.5 -70.5  AOA       Atlantic  Antarctic
3  -82.5 -45.5  EPN    EastPacific   Northern
4  171.5 -65.5  WPA    WestPacific  Antarctic
5 -175.5 -65.5  WPA    WestPacific  Antarctic

A point exactly on a boundary (common for gridded data, as many boundaries run along whole or half degrees) is within neither region; use predicate="intersects" to get both.

Areas

Use the polar equal area version:

regions.assign(area_km2=regions.area / 1e6)[["name", "area_km2"]]

Maps

The fill column has the colours:

ax = regions.plot(color=regions["fill"].fillna("#FFFFFF"),
                  edgecolor="grey", linewidth=0.5)
ax.set_axis_off()