How can I zoom in on the Full Disk Images

[blaylockbk]

This was a question asked in an email

How can I zoom in on the Full Disk Images? I have tried using the Rocket Plume example with no good results

[blaylockbk]

Plotting the GOES data with Cartopy involves several steps. You will need MetPy installed to get the projection information.

The easyier way

from goes2go.data import goes_nearesttime
import matplotlib.pyplot as plt
import cartopy.crs as ccrs
import cartopy.feature as cfeature

import metpy

# Download some data for the full disk
ds = goes_nearesttime(
    '2021-10-18 18:00',
    satellite='goes16',
    domain='F',
)

# Parse out the cartopy coordinate reference system information with MetPy
ds = ds.metpy.parse_cf()
crs = ds.metpy_crs.item().to_cartopy()

# Make a figure. Use `set_extent` to zoom on an area
# Plot the data for ABI channel 1 in projection coordinates
fig = plt.figure()
ax = fig.add_subplot(1, 1, 1, projection=crs)
ax.pcolormesh(ds.x, ds.y, ds.CMI_C01)

# Add coastline and country boarders
ax.coastlines(color='dodgerblue')
ax.add_feature(cfeature.BORDERS, color='dodgerblue')

# Zoom (a bit tricky to fine tune because of different projections.)
ax.set_extent([-80, -30, -60, -22], ccrs.PlateCarree())

(this took ~1 minute to complete)

Setting the extent with lat/lon coordinates might not be what you want because the coordinate reference system is not in lat/lon coordinates. Instead, you can fine tune the zoom area better if you use the GOES coordinate reference system

# Plot the data in projection coordinates
fig = plt.figure()
ax = fig.add_subplot(1, 1, 1, projection=crs)
ax.coastlines(color='dodgerblue')
ax.add_feature(cfeature.BORDERS, color='dodgerblue')

# Zoom (a bit better to tune in the GOES coordinate reference system)
# [left_limit, right_limit, bottom_limit, top_limit]
ax.set_extent([0, 2_300_000, -5_400_000, -2_000_000], crs)

You can do use the set_extent method when plotting an RGB image

ax = plt.subplot(projection=crs)
ax.imshow(ds.rgb.TrueColor(), **ds.rgb.imshow_kwargs)
ax.set_extent([0, 2_300_000, -5_400_000, -2_000_000], crs)

(Hmmm, I don't know why this takes >15 mins to plot)

---

The more-involved way

1. Download GOES ABI data and read with xarray, with goes2go 😁

from goes2go.data import goes_nearesttime
import matplotlib.pyplot as plt

ds = goes_nearesttime('2021-10-18 18:00',
                      satellite='goes16',
                      domain='F',
                      save_dir='$TMPDIR')

2. Parse the CF grid projection information using MetPy accessor parse_cf

ds = ds.metpy.parse_cf()

3. Convert the x/y radian coordinate system to the map projection using Metpy accessor assign_y_x

ds = ds.metpy.assign_y_x()

4. Trim the dataset for a region of interest, transforming limits from lat/lon extent to the data's map projection units.

extent = [-80, -40, -60, -20]

center_lon = ds.goes_imager_projection.longitude_of_projection_origin

min_lon, _ = goes_crs.transform_point(extent[0], 0, ccrs.PlateCarree())
max_lon, _ = goes_crs.transform_point(extent[1], 0, ccrs.PlateCarree())
_, min_lat = goes_crs.transform_point(center_lon, extent[2], ccrs.PlateCarree())
_, max_lat = goes_crs.transform_point(center_lon, extent[3], ccrs.PlateCarree())

5. Plot the data with cartopy.

from toolbox.cartopy_tools import common_features, pc, ccrs

ax = common_features(crs=ds.metpy_crs.item().to_cartopy()).BORDERS().ax
ax.pcolormesh(zoom_ds.x, zoom_ds.y, zoom_ds.CMI_C01)
ax.set_extent(extent, pc)

(this seems to take ~2 minutes to complete)

[blaylockbk]

Zoom on a CONUS image to show west coast

An example for zooming on CONUS images. Again, use the ax.set_extent() method.

from goes2go.data import goes_nearesttime
import matplotlib.pyplot as plt
import cartopy.crs as ccrs

G1 = goes_nearesttime('2021-11-03', satellite='G17', product="ABI")

Plot full figure

fig = plt.figure(figsize=[10,10])
ax = fig.add_subplot(1,1,1, projection=G1.rgb.crs)
ax.coastlines(color='y')
ax.imshow(G1.rgb.NaturalColor(), **G1.rgb.imshow_kwargs)

Plot zoomed figure

fig = plt.figure(figsize=[10,10])
ax = fig.add_subplot(1,1,1, projection=G1.rgb.crs)
ax.coastlines(color='y')
ax.imshow(G1.rgb.NaturalColor(), **G1.rgb.imshow_kwargs)
ax.set_extent([-130, -115, 30, 50], crs=ccrs.PlateCarree()) #<--- This is the important line

Setting the extent with lat/lon coordinates (and specifying the crs) bounds the domain by the upper-left and bottom-right corners.

fig = plt.figure(figsize=[10,10])
ax = fig.add_subplot(1,1,1, projection=G1.rgb.crs)
ax.coastlines(color='y')
ax.imshow(G1.rgb.NaturalColor(), **G1.rgb.imshow_kwargs)
ax.set_extent([-130, -115, 30, 50], crs=ccrs.PlateCarree())

ax.gridlines(draw_labels=True, xlocs=range(-140, 0, 2), ylocs=range(30,56,2))

ax.scatter(-130, 50, transform=ccrs.PlateCarree(), s=300)
ax.scatter(-115, 30, transform=ccrs.PlateCarree(), s=300)