Light attenuation in Fasham

[LauraB]

Hello all,

My question now is about light attenuation, which is given by an exponential: exp[-(kw + kchl . CHL) z],
where CHL is the integrated chlorophyll concentration in the column from surface to level z.

In BioFasham.in, kchl (AttChl) is given in units of 1/(mgChl m^2). In the code, the computation of light attenuation is

Code: Select all

Att=EXP(-0.5_r8*(AttSW(ng)+AttChl(ng)*Bio(i,k,iChlo))*  &
     &                  (z_w(i,j,k)-z_w(i,j,k-1)))

Bio(i,k,iChlo) units should be mgChl*m, and should represent the integrated chlorophyll in a grid cell... Does Bio(i,k,iChlo) have these characteristics? As far as I understand fasham.h, Bio(i,k,iChlo) is chlorophyll in mg/m^3 (somewhere a comment says that "all the tracer at input have transport units: m Tunits....", and then creates Bio(i,k,indx) by dividing by Hz)

Thanks a lot for the input,
Laura

[kfennel]

Hi Laura,

In the equation exp[-(kw + kchl . CHL) z] CHL represents the mean chlorophyll concentration between z and the surface (the chl integral divided by z).

The correct unit for kCHL is 1/(mg CHL m^{-2}).

Also note that the factor 1/z in front of the chl integral is missing in equation 5 in Fennel, Wilkin et al. (GBC 2006).

Hope this helps!
-Katja

[longmtm]

Hi Laura,

I want to confirm that Katja is right. The Unit of AttChl in the bioFasham.in file should be[m2/(mg_Chl) rather than [1/(mg_Chl m2).This should be corrected in the file ROMS/External/bioFasham.in

and also Bio(i,k,iChlo) should have unit [milligram CHL/m3]

Wen

[pmaccc]

I think that the units of AttChl are still given incorrectly in the latest version of ROMS/External/bio_Fennel.in:

! Light attenuation by chlorophyll [1/(mg_Chl m2)], {0.02486d0}.

AttChl == 0.02486d0

Should be [m2 / (mg Chl)] right?

[wilkin]

You are correct. The units of AttChl are m^2 / mg_Chl. This is because they are inverse meters per chlorophyll concentration.

At line 711 in fennel.h where AttChl is multiplied by the chlorophyll concentration (mg_Chl / m^3) it has to give units of inverse length ... to match units of AttSW ... and so that when Att (the sum of all attenuation factors) is multiplied by delta_z we have a nondimensional quantity to pass to the EXP( ) function.

Code: Select all

                Att=(AttSW(ng)+                                         &
     &               AttChl(ng)*Bio(i,k,iChlo)+                         &
     &               AttFac)*                                           &
     &               (z_w(i,j,k)-z_w(i,j,k-1))
                ExpAtt=EXP(-Att)