Skip to content

Adjust airplane altitude calculation in LUT generation#789

Merged
unbohn merged 2 commits into
isofit:devfrom
WonseokChoi0:patch-1
Feb 23, 2026
Merged

Adjust airplane altitude calculation in LUT generation#789
unbohn merged 2 commits into
isofit:devfrom
WonseokChoi0:patch-1

Conversation

@WonseokChoi0

Copy link
Copy Markdown
Contributor

This change is about the definition of "airplane altitude".

isofit intended to generate LUT with a fixed airplane altitude (asl) which is described as H1ALT in MODTRAN template file.
However, 6S defines the airplane altitude as "above ground level" and internally add ground altitude to the airplane altitude.
This caused simulating atmospheric transfer functions at different airplane altitudes as follows:
image
In the above cases, the H1ATL was set to 1.17 km in isofit, but we found that 6S simulated spectra with different airplane altitudes like 1.17, 2.13, and 3.33.

To correct this issue, I have added one line of code in sixs_s.py calculating above ground level by subtracting surface_elevation_km from observer_altitude_km. When the surface_elevation_km was greater than the observer_altitude_km, I set the AGL to its minimum value of 0.01 km.

@pgbrodrick

Copy link
Copy Markdown
Collaborator

This is definitely a correct identified bug and a good patch - thanks Wonseok! We need to think carefully before merging for historic sRTMnet trainings (we might need to add something into the sRTMnet class too).

@pgbrodrick

Copy link
Copy Markdown
Collaborator

Rebased and brought back up to speed. I fully expect all checks to pass. Given that this is a clear correction to the physics, I think we can merge. Previous hesitancy stemmed from backwards compatibility...but given how quickly we're iterating on dev on the 6c physics, I'm inclined to just bring this in. Also, testing with the 3c emulator shows that there are not deleterious consequences from incorporating.

Any resistance to merging this in?

@unbohn

unbohn commented Feb 23, 2026

Copy link
Copy Markdown
Collaborator

Thanks, @WonseokChoi0! Finally merging this one in.

@unbohn unbohn merged commit 911d26e into isofit:dev Feb 23, 2026
23 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants