Nice combo example about python-training

[deeplycloudy]

Here is code for adding GLM flash and group centroid locations, also using the NOAA S3 archive.

In the 5 min window I used, there's lots of flashes, so it swamps the MRMS display at this scale.

# Grab the corresponding GLM data in the 5 min window before the MD.
glm_window = timedelta(minutes=5) # We want this much data
glm_prod_length = timedelta(seconds=20) # LCFA files are always 20s long
n_glm = int(glm_window/glm_prod_length)
glm_prod_times = [md_time - ti*glm_prod_length for ti in range(n_glm)][::-1]
glm_prods = [GOESArchive(satellite=16).get_product('GLM-L2-LCFA', ti).parse() for ti in glm_prod_times]

# Plot GLM
for glm in glm_prods:
    ax.plot(glm.flash_lon, glm.flash_lat, marker='D', color='w', markersize=4, linewidth=0, transform=ccrs.PlateCarree())
    ax.plot(glm.group_lon, glm.group_lat, marker='s', color='y', markersize=2, linewidth=0, transform=ccrs.PlateCarree())