Extracting countries from NetCDF data using geopandas

Question

I am trying to extract countries from NetCDF3 data using the pdsi monthly mean calibrate data from: https://psl.noaa.gov/data/gridded/data.pdsi.html . I am using the following code which performs a spatial merge of coordinates and identifies countries based on a shapefile of the world.

PDSI data format

# Import shapefile from geopandas
path_to_data = geopandas.datasets.get_path("naturalearth_lowres")
world_shp = geopandas.read_file(path_to_data)
    
world_shp.head()

# Import netCDF file
ncs = "pdsi.mon.mean.selfcalibrated.nc"

# Read in netCDF as a pandas dataframe
# Xarray provides a simple method of opening netCDF files, and converting them to pandas dataframes
ds = xr.open_dataset(ncs)
pdsi = ds.to_dataframe()

# the index in the df is a Pandas.MultiIndex. To reset it, use df.reset_index()
pdsi = pdsi.reset_index()

# quick check for shpfile plotting
world_shp.plot(figsize=(12, 8));

# use geopandas points_from_xy() to transform Longitude and Latitude into a list of shapely.Point objects and set it as a geometry while creating the GeoDataFrame
pdsi_gdf = geopandas.GeoDataFrame(pdsi, geometry=geopandas.points_from_xy(pdsi.lon, pdsi.lat))

print(pdsi_gdf.head())

# check CRS coordinates
world_shp.crs #shapefile
pdsi_gdf.crs #geodataframe netcdf

# set coordinates equal to each other
# PointsGeodataframe.crs = PolygonsGeodataframe.crs
pdsi_gdf.crs = world_shp.crs

# check coordinates after setting coordinates equal to each other
pdsi_gdf.crs #geodataframe netcdf


#spatial join
join_inner_df = geopandas.sjoin(pdsi_gdf, world_shp, how="inner")
join_inner_df

The problem I am having is that the original data in the NetCDF format consists of spatial coverage/gridded data where the values of the key variable (pdsi) represents the area within each shaded squares (see image below). So far, only the coordinate points in the middle of the squares are being matches, and I would like each shaded square to match to each country that it is inside. For example, if the area of the shaded squares are within the boundaries of Germany and Netherlands, then the key variable should be attributed to both countries. Any help on this issue would be greatly appreciated.

NetCDF gridded data example

Answer 1

• have sourced data that you referenced to ensure this can be re-run on any machine
• core solution, a square buffer around the point https://gis.stackexchange.com/questions/314949/creating-square-buffers-around-points-using-shapely
• have analysed data to ensure value used for buffer is appropriate and calculated from data

# make sure that data supports using a buffer...
assert (
    gdf["lat"].diff().loc[lambda s: s.ne(0)].mode()
    == gdf["lon"].diff().loc[lambda s: s.ne(0)].mode()
).all()
# how big should the square buffer be around the point??
buffer = gdf["lat"].diff().loc[lambda s: s.ne(0)].mode().values[0] / 2
gdf["geometry"] = gdf["geometry"].buffer(buffer, cap_style=3)

• the remaining solution is now a spatial join

# the solution... spatial join buffered polygons to countries
# comma separate associated countries
gdf = gdf.join(
    world_shp.sjoin(gdf.set_crs("EPSG:4326"))
    .groupby("index_right")["name"]
    .agg(",".join)
)

• have used plotly to visualise. From image you can see that multiple countries have been associated with a bounding box.

complete code

import geopandas as gpd
import numpy as np
import plotly.express as px
import requests
from pathlib import Path
from zipfile import ZipFile
import urllib
import geopandas as gpd
import shapely.geometry
import xarray as xr

# download NetCDF data...
# fmt: off
url = "https://psl.noaa.gov/repository/entry/get/pdsi.mon.mean.selfcalibrated.nc?entryid=synth%3Ae570c8f9-ec09-4e89-93b4-babd5651e7a9%3AL2RhaV9wZHNpL3Bkc2kubW9uLm1lYW4uc2VsZmNhbGlicmF0ZWQubmM%3D"
f = Path.cwd().joinpath(Path(urllib.parse.urlparse(url).path).name)
# fmt: on

if not f.exists():
    r = requests.get(url, stream=True, headers={"User-Agent": "XY"})
    with open(f, "wb") as fd:
        for chunk in r.iter_content(chunk_size=128):
            fd.write(chunk)
ds = xr.open_dataset(f)
pdsi = ds.to_dataframe()
pdsi = pdsi.reset_index().dropna()  # don't care about places in oceans...

# use subset for testing... last 5 times...
pdsim = pdsi.loc[pdsi["time"].isin(pdsi.groupby("time").size().index[-5:])]

# create geopandas dataframe
gdf = gpd.GeoDataFrame(
    pdsim, geometry=pdsim.loc[:, ["lon", "lat"]].apply(shapely.geometry.Point, axis=1)
)

# make sure that data supports using a buffer...
assert (
    gdf["lat"].diff().loc[lambda s: s.ne(0)].mode()
    == gdf["lon"].diff().loc[lambda s: s.ne(0)].mode()
).all()
# how big should the square buffer be around the point??
buffer = gdf["lat"].diff().loc[lambda s: s.ne(0)].mode().values[0] / 2
gdf["geometry"] = gdf["geometry"].buffer(buffer, cap_style=3)

# Import shapefile from geopandas
path_to_data = gpd.datasets.get_path("naturalearth_lowres")
world_shp = gpd.read_file(path_to_data)

# the solution... spatial join buffered polygons to countries
# comma separate associated countries
gdf = gdf.join(
    world_shp.sjoin(gdf.set_crs("EPSG:4326"))
    .groupby("index_right")["name"]
    .agg(",".join)
)
gdf["time_a"] = gdf["time"].dt.strftime("%Y-%b-%d")

# simplest way to test is visualise...
px.choropleth_mapbox(
    gdf,
    geojson=gdf.geometry,
    locations=gdf.index,
    color="pdsi",
    hover_data=["name"],
    animation_frame="time_a",
    opacity=.3
).update_layout(
    mapbox={"style": "carto-positron", "zoom": 1},
    margin={"l": 0, "r": 0, "t": 0, "b": 0},
)