Make distinction between particle and molecular backscatter in variable naming convention

[svniemeijer]

This is of importance for data from e.g. AEOLUS and EarthCARE.

We currently only define backscatter_coeffient and (<aerosol_size>_)aerosol_backscatter_coefficient. But we can more generally distinguish particle (Mie) backscatter and molecular (Rayleigh) backscatter. Particle backscatter not only comprises aerosol backscatter, but any particle backscatter (e.g. clouds).

We should thus at least change (<aerosol_size>_)aerosol_backscatter_coefficient to the more generic <particle>_backscatter_coefficient. And the same applies for the extinction coefficient. But should we also introduce a molecular_backscatter_coefficient? Does this need to be <species> specific?

Linked to this, there is also the distinction between a molecular and particle depolarization ratio, where the sum of both is identified as 'volume depolarization ratio'. We currently have circular_depolarization_ratio and linear_depolarization_ratio, so we should see how the molecular/particle distinction can coexist with the circular/linear distinction.

[svniemeijer]

In EartCARE for ATL_AER_2A the variables in the product used are particle_extinction_coefficient_355nm, particle_backscatter_coefficient_355nm, and particle_linear_depolarization_ratio_355nm.

With the change to <particle>_backscatter_coefficient, we should thus probably also rename the existing HARP variables for these products from aerosol_* to particle_*.

And the ATL_AER_2A example also shows how to combine the molecular/particle distinction with the circular/linear distinction in HARP.

[svniemeijer]

This has been implemented in V1.22 of HARP.