relhum output: liquid only or wrt ice?

[MicroTed]

The relative humidity subroutine compute_relhum in driver_microphysics switches to RH_ice for T < 0C, but I personally find it much more useful as always being RH_w (as typically done for soundings). What do people think about changing it to output RH_w? (As is done in WRF?) Maybe add a namelist switch in case somebody wants RH_ice?

Furthermore, the diagnostic could be enhanced to check for a particular microphysics scheme and use a consistent calculation. (That could be up to the MP developers to add.)

 do j = jts,jte
 do i = its,ite
    do k = kts,kte
       p1d(k)  = pres_p(i,k,j)
       t1d(k)  = th_p(i,k,j) * pi_p(i,k,j)
       tempc   = t1d(k) - 273.16_RKIND
       qvs1d(k) = rslf(p1d(k),t1d(k))
       if(tempc .le. 0._RKIND) qvs1d(k) = rsif(p1d(k),t1d(k)) ! <--- comment out to make it always RH_w
       qv1d(k) = qv_p(i,k,j)
       relhum(k,i) = qv1d(k) / qvs1d(k)
       relhum(k,i) = relhum(k,i) * 100._RKIND
    enddo
 enddo
 enddo