Change cinitime timestamp for scalars of initial mixing ratios to account for length of back trajectory

rename izwork to cova in mod_analytic for clarity and added check to mod_settings for use of const_out

changed read_obs.f90 to deal with different length of satellite name and in mod_settings.f90 updated default values and tests for parameters

Fix to average.f90 to include new retrieval OCO-2 and S5P-BLND and inclusion of routines and SETTINGS for S5P-BLND

Bug fix to prep_satellite for case when no retrievals found within +/-2 SD of mean and update to prep_regions for satellites for compatability with new version of FPv10.4 for satellites

Change to calculation of eigendecomposition of prior error covariance to only calculate eigenvectors for the kept eigenvalues

changes to find median solution for log-normal PDF in state space and changes to reading of prior fluxes

updates for lognormal inversion to include influence on mixing ratio from ocean fluxes when these are not optimized