Fix bug in implementation of LUO_WETDEP option

Following the benchmarking of GEOS-Chem 13.2.0 with the Luo et al. (2020)
wet deposition scheme, Gan Luo noticed a bug in the code. He wrote:

   Simulated sulfate mass concentration looks lower than what I expected.

   I checked https://github.com/geoschem/geos-chem/blob/13.2.0-beta.0/GeosCore/sulfate_mod.F90
   It seems that we missed some switches for Non-volatile aerosol
   concentration calculation:

   (1) At line 5142 D = (2.e+0_fp*SO4nss) - (TNA+2.e+0_fp*TDCA), it should be:

       #ifdef LUO_WETDEP
          D = (1.5e+0_fp*SO4nss) - (TNA+2.e+0_fp*TDCA)
       #else
          D = (2.e+0_fp*SO4nss) - (TNA+2.e+0_fp*TDCA)
       #endif

   (2) At line 5156 D = (2.e+0_fp * SO4nss) - TNA), it should be:

       #ifdef LUO_WETDEP
          D = (1.5e+0_fp * SO4nss) - TNA
       #else
          D = (2.e+0_fp * SO4nss) - TNA
       #endif

   Switching from 2.e+0_fp * SO4nss to 1.5e+0_fp * SO4nss is following the
   reasons introduced in Luo et al. (2020).

Signed-off-by: Melissa Sulprizio <[email protected]>

GeosCore/sulfate_mod.F90

@@ -5139,7 +5139,11 @@ SUBROUTINE GET_HPLUS( SO4nss, TNH3, TNO3, SO2, CL, TNA, TDCA, TFA, &
        ! then all Ca is dissolved else [Ca2+] varies with [H+]
        IF ( fCa .ge. TDCA ) THEN
           ! Non-volatile aerosol concentration [M]
+#ifdef LUO_WETDEP
+          D = (1.5e+0_fp*SO4nss) - (TNA+2.e+0_fp*TDCA)
+#else
           D = (2.e+0_fp*SO4nss) - (TNA+2.e+0_fp*TDCA)
+#endif
 
           ! Define f(x)
           f = D - nHPLUS + Kw/nHPLUS + fHCO3 + 2.e+0_fp * &
@@ -5153,7 +5157,11 @@ SUBROUTINE GET_HPLUS( SO4nss, TNH3, TNO3, SO2, CL, TNA, TDCA, TFA, &
 
        ELSE
           ! Non-volatile aerosol concentration [M]
+#ifdef LUO_WETDEP
+          D = (1.5e+0_fp * SO4nss) - TNA
+#else
           D = (2.e+0_fp * SO4nss) - TNA
+#endif
 
           ! Define f(x)
           f = D - nHPLUS + Kw/nHPLUS + fHCO3 + 2.e+0_fp * fCO3 + &