search

open-geocomputing / OpenEarthEngineLibrary

https://www.open-geocomputing.org/OpenEarthEngineLibrary/
125 stars 38 forks source link

Savitzky-Golay #10

Closed Hisschao closed 1 year ago

Hisschao commented 1 year ago

The d_1_ndvi value after sg filtering is very small, what is the reason for this?

mgravey commented 1 year ago

can you send a GEE link ?

Hisschao commented 1 year ago

https://code.earthengine.google.com/a781f2d3b5242c6fe849602b254d3a21

mgravey commented 1 year ago

oh, I see what you mean by very low then. I fact this is due to the use of "system:time_start" that is the time encoded in millisecond, if you want the derivate in second simply multiply by 1000, same for per days multiply by 8.64e7, etc.

Hisschao commented 1 year ago

Thank you for your answer.

Hisschao commented 1 year ago

Thank you for your patience with my question. Another question I want to ask is: How do I get the first derivative of the VI time series when I use a filter other than SG (eg harmonic model)?

mgravey commented 1 year ago

d1 IS the first derivative !!

Hisschao commented 1 year ago

Thank you, if SG is not used as a filter, can d1 be used as the first derivative?

mgravey commented 1 year ago

sure but it's not the derivate of the raw signal, it's the derivate of the filtered signal