Polymer for Sentinel-2

Post Reply
avea
Posts: 1
Joined: Thu Mar 07, 2019 7:37 am
company / institution: University of Tartu, Tartu Observatory
Location: Estonia

Polymer for Sentinel-2

Post by avea » Fri Mar 08, 2019 6:12 am

Dear Steinmetz,

Thank you for Polymer.
I would like to process S2 images with Polymer using 10 m resolution. However, it takes forever because of the resolution (but I really need to use 10 m resolution). Is it possible to fasten the process defining the needed area of the image (subset, location etc)) or using some other approach (masking etc.).

Thank you
User avatar
fsteinmetz
Site Admin
Posts: 32
Joined: Fri Sep 07, 2018 1:34 pm
company / institution: Hygeos
Location: Lille, France
Contact:

Re: Polymer for Sentinel-2

Post by fsteinmetz » Fri Mar 08, 2019 7:06 am

Hello,

Yes there are three options that you can use to make it faster :

Code: Select all

from polymer.gsw import GSW

run_atm_corr(
	Level1_MSI(name, resolution='10',
		sline=1000, eline=2000, scol=1000,  ecol=2000,  # (1) process only a part of the image (coordinates within the 10980x10980 image because we are at 10m resolution)
		landmask=GSW()  # (2) apply a land mask, otherwise all land is processed (incorrectly)
		),
	Level2_nc(...),
	multiprocessing=-1,   # (3) use multiple cores of your processor in parallel
	)
Note that there is some documentation about these options in the python docstrings.
Post Reply