Basic plots and analyses of the simulations
###########################################

In this page, we provide python scripts to make first plots for the analysis of the simulation (forward or inverse) outputs.
These are only basic plots and should be adapted/extended based on the scientific questions and focuses of the study.

1. Forward simulation
=====================

This script provides basic analyses of the results of a forward simulation of TROPOMI satellite observations, within the CIF-CHIMERE framework. 
The example data is for a simulation in the South American domain.

The script loads the output monitor file, then makes 5 plots: 
    - **Fig 1**: histogram of obs-sim differences
    - **Fig 2**: time series of daily/monthly observed and simulated XCH4
    - **Fig 3**: maps of TROPOMI obs count, observed XCH4 and simulated XCH4 per lat/lon pixel
    - **Fig 4**: maps of the obs-sim XCH4 difference and the relative difference per lat/lon pixel
    - **Fig 5**: seasonal maps of obs-sim, obs and sim XCH4 per lat/lon pixel

.. container:: toggle
    .. literalinclude:: ./scripts/plot_fwd.py
              :linenos:
              :language: python


2. Inverse simulation
=====================

This script provides basic analyses of the results of an inverse simulation assimilating TROPOMI satellite observations, within the CIF-CHIMERE framework. 
The example data is for a simulation in the South American domain.

It makes 7 plots: 
    - **Fig 1**: plots of J and grad(J) along the optimization process
    - **Fig 2**: histograms of prior and posterior differences obs-sim XCH4
    - **Fig 3**: time series + maps of observed, prior simulated and posterior simulated XCH4
    - **Fig 4**: time series of prior and posterior CH4 emissions per sector
    - **Fig 5**: maps of increments (posterior-prior) of CH4 emissions per sector
    - **Fig 6**: budgets of prior and posterior CH4 emissions per sector
    - **Fig 7**: budgets of prior and posterior CH4 emissions per country [optional, if mask]

.. warning::
    The loading of the control vectors for each sector can be long (several minutes) if there are many sectors.

.. container:: toggle
    .. literalinclude:: ./scripts/plot_inv.py
              :linenos:
              :language: python