Changes since last module_mp_thompson.F update:

In order of importance for these modifications:
1.) Created entirely new look-up tables for collision efficiencies of cloud water
    by rain and snow (see comments in each subroutine) and removed Stokes number
    relationships. Altered graupel collecting cloud water as well, but retains a
    Stokes number relationship.
2.) Reduced max rain median volume diameter (mvd_r) from 3 to 2.5 mm.
3.) Altered graupel collecting rain above melting to melt (instanteously) the
    collector species into rain.  Other half remains graupel.
4.) Switched to spherical cloud ice (am_i, bm_i) and increased min snow diameter
    from 125 to 200 microns (D0s). With larger min snow size, altered prefactor
    for v(D) relation for cloud ice (av_i).
5.) Constrained cloud ice number such that mass-weighted mean size remains between
    30 and 300 microns and fewer than 250 crystals per liter.
6.) Altered graupel intercept parameter relation with mixing ratio, allowed a
    larger upper limit of 5.E6 (was 1.E6), and increased min graupel diameter from
    150 to 250 microns.
7.) Altered requirements for ice nucleation from deposition & condensation
    freezing.  Now requires water saturation and T<-10C or 20% ice supersaturated.
8.) Increased max rain intercept parameter (ronv_max) from 2.E9 to 9.E9.
9.) Decreased max collision efficiencies between various species to max 95%, was 99%.
10.) Removed extraneous code for Koenig ice deposition (ABER1, ABER2, CBG).
11.) Created size bins for all hydrometeor distributions; formerly, these were
     created within look-up tables, now moved to top-level for use in more places.
     New size bins required changing a few variables names, like Dc changed to xDc
     and dts changed to dtsave among others.
12.) Some constants like cre(4), cge(4), cse(3), cse(15), and cse(17) were not
     used so they were altered in preparation for radar reflectivity code.