plot data on equal-area grid (Berkeley Earth)

[mktippett]

Berkeley Earth provides data on an equal-area grid. How can I plot that?

For instance,

import xarray as xr
import fsspec

url = 'https://berkeley-earth-temperature.s3.us-west-1.amazonaws.com/Global/Gridded/Complete_TAVG_Daily_EqualArea.nc'
with fsspec.open(url) as fobj:
    ds = xr.open_dataset(fobj)
    print(ds.temperature)

reads daily temperature, which has time and map_points dimensions.

<xarray.DataArray 'temperature' (time: 52077, map_points: 5498)>
[286319346 values with dtype=float32]
Coordinates:
    longitude  (map_points) float32 ...
    latitude   (map_points) float32 ...
Dimensions without coordinates: time, map_points
Attributes:
    units:          degree C
    standard_name:  surface_temperature_anomaly
    long_name:      Air Surface Temperature Anomaly
    valid_min:      -33.149067
    valid_max:      31.330288

How do I get a lat-on plot instead of this? Thanks!

ds.temperature.isel(time=0).plot()

EDIT chatGPT doesn't know either. Or rather, it thinks it does but it doesn't 🤓

[kmuehlbauer]

Try to use scatter plot with x="longitude" and y=latitude. I'll look up the correct incantation, when I'm back at my machine and if no-one beats me to it.

[mktippett]

With details

fig = plt.figure(figsize=(12, 5))
ax = plt.axes(projection=ccrs.PlateCarree())

xr.plot.scatter(ds.isel(time=-1), x='longitude', y='latitude', hue='temperature')
ax.coastlines()
plt.show()

Or in matplotlib

data = ds["temperature"].isel(time=-1)
lats = ds["latitude"]
lons = ds["longitude"]

fig = plt.figure(figsize=(12, 5))
ax = plt.axes(projection=ccrs.PlateCarree())

plt.scatter(x=lons, y=lats, c=data, cmap='RdBu_r', vmin=-12, vmax=12)
ax.coastlines()

plt.colorbar()
plt.show()