gorelick / fmask

Automatically exported from code.google.com/p/fmask
0 stars 0 forks source link

TWO DIFFERENT RESULTS FOR 2 DIFFERENT SEASON DATA, LESS ACCURACY IN PRE MONSOON SEASON AS COMPARE TO MONSOON SEASON. #18

Open GoogleCodeExporter opened 8 years ago

GoogleCodeExporter commented 8 years ago
What steps will reproduce the problem?
1.
2.
3.

What is the expected output? What do you see instead?

What version of the product are you using? On what operating system?
version 3.2 and windows 8 64 bit.

Please provide any additional information below.

Hello Sir,

I have used fmask for Landsat 8 OLI data and my study area is in Punjab, India. 
I have observed two different things in fmask results as my study include two 
season data, Pre Monsoon (May month) and Monsoon (August month). Both season 
data were covered with cloud and shadow and for May month (Pre monsoon season) 
data only 60 % accuracy reported, even cloud free pixels also mapped as cloud 
which include waterbodies, crop land and vegetation. But for august month 
(monsoon season) the accuracy were around 90%, means it worked correctly for 
this season. I don't know why it happened, is there any seasonal affect on 
cloud condition. How can I achieve more accuracy for May month data. I have 
used same command (default as suggested in tutor) for both season data to run 
fmask.
My other question is related to DN and TOA. Why it requires to convert DN to 
TOA for further data processing, what difference it make as compare to DN in 
results ?

Original issue reported on code.google.com by vivekkan...@gmail.com on 5 Feb 2015 at 3:57