= Atmospheric correction with ATCOR =

The latest version of our atmospheric correction procedure is being developed in a private GitHub repository - email us if you require access. The process is still being refined to cope with the variety of environments and conditions flown by NERC-ARF and feedback is welcome.

== NERC-ARF Specific Instructions ==

For running on the NERC-ARF systems there are a number of utilities to make batch processing easier. 

**Work in progress - more documentation will be added as it is written **

=== Generating the .inn file ===

To generate the .inn file for ATCOR use the `gen_inn_atcor.py` script.

* `--calfile` needs to point to the .cal file ATCOR users, this wil be `~arsf/usr/opt/atcor_4/sensor/SENSOR/*.cal` where `SENSOR` is the name used for the sensor definition in ATCOR (e.g., f17_b2).
*  `inputbil` needs to point to the bil file in the delivery folder directory.
* `--outputdir` should be the directory used or ATCOR processing.

If microtops data has been provided pass this in using `--microtops` to read using the https://github.com/robintw/PyMicrotops library and set a constant visibility based on this. Note if the microtops data has a number of header lines these need to be removed before PyMicrotops can read the file, there should just be one header line with the column names.

After generating the .inn file it should be renamed to use the same name as the image (without  the .bsq extension). For example if the file being processed in ATCOR is 'f168041b_masked.bsq' name 'f168041b_masked.inn'.

Currently ATCOR doesn't always read all the parameters correctly open the .inn file (plain text) and update the values in the ATCOR GUI if needed.

=== Batch running .inn files ===

Once .inn files have been generated and checked using the ATCOR GUI it is possible to run in batch using the script `batch_run_atcor_inn.py`. To do so ssh onto 'rsg-atcor' and run:
{{{
batch_run_atcor_inn.py /path/to/files/to/process
}}}
This will recursively search for .inn files and run them through ATCOR.