No subject
Thu Oct 30 11:44:21 EDT 2014
>>> > what I understand is that<br>
>>> ><br>
>>> > tend_heat_surf =3D=C2=A0 tend_advection +<br>
>>> >=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =
=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 tend_diffusion +<br>
>>> >=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =
=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 tend_kpp +<br>
>>> >=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =
=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 (TFLUX-oceQsw)/(rhoConst*Cp*DRF(1=
)*hFacC(i,j,1))<br>
>>> > +<br>
>>> > =E2=80=8B<br>
>>> >=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =
=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=
=A0 =C2=A0tend_Qsw -<br>
>>> >=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =
=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0(WTHMASS(i,j,1) -<br=
>
>>> > TsurfCor)/(DRF(1)*hFacC(i,j,1))<br>
>>> > where<br>
>>> > tend_Qsw =3D oceQsw/(rhoConst*Cp)/(DRF(1)*hFacC(i,j,1))*(=
swfrac(1)-swfrac(2))<br>
>>> > and<br>
>>> > TsurfCor =3D SUM( WTHMASS(i,j,1)*RAC(i,j) ) /=C2=A0 globa=
lArea<br>
>>> ><br>
>>> > and "swfrac" is already defined in our earlier =
communications.<br>
>>> ><br>
>>> > =E2=80=8BAm I correct?<br>
>>> ><br>
>>> > Thanking in advance,<br>
>>> > Abhisek=E2=80=8B<br>
>>> ><br>
>>> ><br>
>>> ><br>
>>> ><br>
>>> > *--------------------------------------------------------=
------------------------ABHISEK<br>
>>> > CHAKRABORTY*<br>
>>> ><br>
>>> > *Scientist - "SD"Oceanic Sciences Division (*<b=
r>
>>> > *AOSG/EPSA)*<br>
>>> ><br>
>>> > *Space Applications Centre*<br>
>>> ><br>
>>> > *Indian Space Research Organization*<br>
>>> ><br>
>>> > *Ahmedabad - 380 015*<br>
>>> ><br>
>>> > *Gujarat, INDIA*<br>
>>> > *Contact: +91-79-2691-6054 (O), +91-79-2686-1929(R)*<br>
>>> ><br>
>>> > On Tue, Nov 25, 2014 at 2:21 AM, Jean-Michel Campin <<=
a href=3D"mailto:jmc at ocean.mit.edu">jmc at ocean.mit.edu</a>><br>
>>> > wrote:<br>
>>> ><br>
>>> > > Hi Abhisek,<br>
>>> > ><br>
>>> > > > Following your suggestions, I think I am now ab=
le to close the heat<br>
>>> > > budget<br>
>>> > > > for the subsurface levels.<br>
>>> > > This is good news.<br>
>>> > ><br>
>>> > > > I have attached two figures (one for subsurface=
at 20 m depth and another<br>
>>> > > > for surface).<br>
>>> > > Could you clarify what is represented on these 2 fig=
s: it is for a<br>
>>> > > single grid point (but where ?) or the horizontal me=
an ?<br>
>>> > ><br>
>>> > > Regarding the surface forcing, which diagnostics are=
you using:<br>
>>> > >=C2=A0 surForcT ? TFLUX ? or oceQnet and TRELAX ?<br>
>>> > ><br>
>>> > > If you are using TFLUX (which contains the verticall=
y integrated SW<br>
>>> > > heating)<br>
>>> > > for the surface budget, you may want to remove oceQs=
w and then<br>
>>> > > add back tend_Qsw(k=3D1) (as you wrote below).<br>
>>> > ><br>
>>> > > Cheers,<br>
>>> > > Jean-Michel<br>
>>> > ><br>
>>> > > On Fri, Nov 21, 2014 at 07:13:31PM +0530, Abhisek Ch=
akraborty wrote:<br>
>>> > > > Dear Jean-Michel,<br>
>>> > > ><br>
>>> > > > To close the surface heat budget, I followed yo=
ur suggestion:<br>
>>> > > ><br>
>>> > > ><br>
>>> > > =E2=80=8B=E2=80=8B<br>
>>> > > TsurfCor =3D SUM( WTHMASS(i,j,1)*RAC(i,j) ) /=C2=A0 =
globalArea<br>
>>> > > >=C2=A0 and finally:<br>
>>> > > >=C2=A0 tend_T =3D (...) - ( WTHMASS(i,j,1) - Tsu=
rfCor )/(drF(1)*hFacC(i,j,1))<br>
>>> > > ><br>
>>> > > > Also, the shortwave tendency terms is computed =
as:<br>
>>> > > ><br>
>>> > > > tend_Qsw =3D oceQsw/(rhoConst*Cp)/(drF(k)*hFacC=
(I,j,k))<br>
>>> > > >=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0=
* ( swfrac(k) - swfrac(k+1) )<br>
>>> > > > where for surface, k=3D1<br>
>>> > > ><br>
>>> > > > But in this way, the surface tendency term is b=
ecoming one order higher<br>
>>> > > in<br>
>>> > > > magnitude than other terms (advection, diffusio=
n, surface correction.<br>
>>> > > Pls.<br>
>>> > > > see the figure in my earlier mail).<br>
>>> > > ><br>
>>> > > > So, please guide me how to close the budget at =
surface.<br>
>>> > > ><br>
>>> > > > regards,<br>
>>> > > > Abhisek<br>
>>> > > ><br>
>>> > > ><br>
>>> > > ><br>
>>> > > ><br>
>>> > > ><br>
>>> > > ><br>
>>> > > *---------------------------------------------------=
-----------------------------ABHISEK<br>
>>> > > > CHAKRABORTY*<br>
>>> > > ><br>
>>> > > > *Scientist - "SD"Oceanic Sciences Div=
ision (*<br>
>>> > > > *AOSG/EPSA)*<br>
>>> > > ><br>
>>> > > > *Space Applications Centre*<br>
>>> > > ><br>
>>> > > > *Indian Space Research Organization*<br>
>>> > > ><br>
>>> > > > *Ahmedabad - 380 015*<br>
>>> > > ><br>
>>> > > > *Gujarat, INDIA*<br>
>>> > > > *Contact: +91-79-2691-6054 (O), +91-79-2686-192=
9(R)*<br>
>>> > > ><br>
>>> > > > On Thu, Nov 20, 2014 at 6:59 PM, Abhisek Chakra=
borty <<br>
>>> > > <a href=3D"mailto:abhisek.sac at gmail.com">abhisek.sac=
@gmail.com</a>><br>
>>> > > > wrote:<br>
>>> > > ><br>
>>> > > > > Hi Jean-Michel,<br>
>>> > > > ><br>
>>> > > > > Sorry for delayed reply.<br>
>>> > > > ><br>
>>> > > > > Following your suggestions, I think I am n=
ow able to close the heat<br>
>>> > > budget<br>
>>> > > > > for the subsurface levels.<br>
>>> > > > ><br>
>>> > > > > However, in spite of my different attempts=
I am not able to close the<br>
>>> > > > > budget at surface where I am finding the s=
urface correction term (as<br>
>>> > > > > specified in your mail) along with other a=
dvection-diffusion terms are<br>
>>> > > too<br>
>>> > > > > small to balance the shortwave heating ter=
m.<br>
>>> > > > ><br>
>>> > > > > I have attached two figures (one for subsu=
rface at 20 m depth and<br>
>>> > > another<br>
>>> > > > > for surface).<br>
>>> > > > ><br>
>>> > > > > Pls. let me know if I am missing something=
.<br>
>>> > > > ><br>
>>> > > > > Thanking in advance,<br>
>>> > > > > Abhisek<br>
>>> > > > ><br>
>>> > > > ><br>
>>> > > > ><br>
>>> > > > ><br>
>>> > > *---------------------------------------------------=
-----------------------------ABHISEK<br>
>>> > > > > CHAKRABORTY*<br>
>>> > > > ><br>
>>> > > > > *Scientist - "SD"Oceanic Science=
s Division (*<br>
>>> > > > > *AOSG/EPSA)*<br>
>>> > > > ><br>
>>> > > > > *Space Applications Centre*<br>
>>> > > > ><br>
>>> > > > > *Indian Space Research Organization*<br>
>>> > > > ><br>
>>> > > > > *Ahmedabad - 380 015*<br>
>>> > > > ><br>
>>> > > > > *Gujarat, INDIA*<br>
>>> > > > > *Contact: +91-79-2691-6054 (O), +91-79-268=
6-1929(R)*<br>
>>> > > > ><br>
>>> > > > > On Wed, Nov 19, 2014 at 1:19 AM, Jean-Mich=
el Campin <<a href=3D"mailto:jmc at ocean.mit.edu">jmc at ocean.mit.edu</a><br=
>
>>> > > ><br>
>>> > > > > wrote:<br>
>>> > > > ><br>
>>> > > > >> Hi Abhisek,<br>
>>> > > > >><br>
>>> > > > >> This budget looks better. Just few rem=
arks (see below).<br>
>>> > > > >><br>
>>> > > > >> On Tue, Nov 18, 2014 at 11:20:46PM +05=
30, Abhisek Chakraborty wrote:<br>
>>> > > > >> > Dear Jean-Michel,<br>
>>> > > > >> ><br>
>>> > > > >> > Thanks for your prompt suggestion=
.<br>
>>> > > > >> ><br>
>>> > > > >> ><br>
>>> > > > >> > (1) the advection term for k>1=
will be<br>
>>> > > > >> ><br>
>>> > > > >> >=C2=A0 =C2=A0 - [ (ADVx_TH(i+1,j,k=
) =E2=80=93 ADVx_TH(I,j,k))/CV +<br>
>>> > > > >> ><br>
>>> > > > >> >=C2=A0 =C2=A0 =C2=A0(ADVy_TH(I,j+1=
,k) =E2=80=93 ADVy_TH(I,j,k))/CV +<br>
>>> > > > >> ><br>
>>> > > > >> >=C2=A0 =C2=A0 (ADVr_TH(I,j,k+1) =
=E2=80=93 ADVr_TH(I,j,k))/CV ]<br>
>>> > > > >><br>
>>> > > > >>=C2=A0 The vertical advective & dif=
fusive transport/flux are (I think)<br>
>>> > > > >>=C2=A0 positive in the direction of the=
vertical coordinate, i.e.,<br>
>>> > > > >>=C2=A0 when using z-coords, positive up=
ward. So you will have to reverse<br>
>>> > > > >>=C2=A0 (or add a minus sign) in front o=
f:<br>
>>> > > > >>=C2=A0 =C2=A0 =C2=A0 (ADVr_TH(I,j,k+1) =
=E2=80=93 ADVr_TH(I,j,k))/CV<br>
>>> > > > >><br>
>>> > > > >> ><br>
>>> > > > >> >=C2=A0 Where CV=3DAc * del RF * Hc=
=3D RAC(I,j) * (RF(K+1)-RF(K))=C2=A0 *<br>
>>> > > hFacC(I,j,k)<br>
>>> > > > >> ><br>
>>> > > > >> >=C2=A0 (<br>
>>> > > > >><br>
>>> > > <a href=3D"http://mitgcm.org/public/r2_manual/latest=
/online_documents/node73.html">http://mitgcm.org/public/r2_manual/latest/on=
line_documents/node73.html</a>)<br>
>>> > > > >> ><br>
>>> > > > >> ><br>
>>> > > > >> ><br>
>>> > > > >> > And similarly the diffusion and K=
PP terms. Correct?<br>
>>> > > > >> ><br>
>>> > > > >> ><br>
>>> > > > >> > (2) The source term (shortwave) f=
or k>1 will be the following<br>
>>> > > > >> ><br>
>>> > > > >> >=C2=A0 =C2=A0 depth=3DRF(k)<br>
>>> > > > >> ><br>
>>> > > > >> >=C2=A0 =C2=A0 swfrac=3D0.62*exp(de=
pth/0.6) + (1.0-0.62) * exp(depth/20)<br>
>>> > > > >> ><br>
>>> > > > >> >=C2=A0 =C2=A0if(depth < -200.0)=
swfrac=3D0.0<br>
>>> > > > >> ><br>
>>> > > > >> >=C2=A0 =C2=A0source term =3D=C2=A0=
=C2=A0(oceQsw/(rhoConst*Cp)) * swfrac * hFacC(I,j,k)<br>
>>> > > > >> ><br>
>>> > > > >> >=C2=A0 =C2=A0Is this correct for k=
>1 ?<br>
>>> > > > >><br>
>>> > > > >>=C2=A0 This does not look right:<br>
>>> > > > >>=C2=A0 =C2=A0 from depth=3DRF(k)=C2=A0 =
, we compute swfrac(k) as above<br>
>>> > > > >>=C2=A0 =C2=A0 from depth=3DRF(k+1), we =
compute swfrac(k+1) as above except that<br>
>>> > > > >>=C2=A0 =C2=A0 =C2=A0at the bottom (deep=
est level of this column) swfrac(k+1)=3D0.<br>
>>> > > > >>=C2=A0 and the tendency contribution fr=
om shortwave heating should be<br>
>>> > > something<br>
>>> > > > >> like:<br>
>>> > > > >>=C2=A0 =C2=A0 tend_Qsw =3D oceQsw/(rhoC=
onst*Cp)/(drF(k)*hFacC(I,j,k))<br>
>>> > > > >>=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=
=A0 =C2=A0* ( swfrac(k) - swfrac(k+1) )<br>
>>> > > > >> ><br>
>>> > > > >> > (3) I am using the tendency as TO=
TTTEND/86400. Is it correct?<br>
>>> > > > >><br>
>>> > > > >>=C2=A0 yes, providing you are using lin=
ear free-surface.<br>
>>> > > > >><br>
>>> > > > >> ><br>
>>> > > > >> > (4) I am using linFSConserveTr=3D=
TRUE. Yet I have to use<br>
>>> > > > >> > WTHMASS(i,j,1)*RAC(i,j) to correc=
t the heat budget at the surface?<br>
>>> > > > >><br>
>>> > > > >>=C2=A0 Yes, but since you are using lin=
FSConserveTr=3DTRUE., the global mean<br>
>>> > > > >>=C2=A0 =C2=A0value of WTHMASS has to be=
removed in the budget (like it is in the<br>
>>> > > > >> model):<br>
>>> > > > >>=C2=A0 =C2=A0 TsurfCor =3D SUM( WTHMASS=
(i,j,1)*RAC(i,j) ) /=C2=A0 globalArea<br>
>>> > > > >>=C2=A0 and finally:<br>
>>> > > > >>=C2=A0 tend_T =3D (...) - ( WTHMASS(i,j=
,1) - TsurfCor )/(drF(1)*hFacC(i,j,1))<br>
>>> > > > >><br>
>>> > > > >> Cheers,<br>
>>> > > > >> Jean-Michel<br>
>>> > > > >><br>
>>> > > > >> ><br>
>>> > > > >> > Thanking in advance,<br>
>>> > > > >> ><br>
>>> > > > >> > Abhisek<br>
>>> > > > >> ><br>
>>> > > > >> ><br>
>>> > > > >> ><br>
>>> > > > >> ><br>
>>> > > > >><br>
>>> > > *---------------------------------------------------=
-----------------------------ABHISEK<br>
>>> > > > >> > CHAKRABORTY*<br>
>>> > > > >> ><br>
>>> > > > >> > *Scientist - "SD"Oceani=
c Sciences Division (*<br>
>>> > > > >> > *AOSG/EPSA)*<br>
>>> > > > >> ><br>
>>> > > > >> > *Space Applications Centre*<br>
>>> > > > >> ><br>
>>> > > > >> > *Indian Space Research Organizati=
on*<br>
>>> > > > >> ><br>
>>> > > > >> > *Ahmedabad - 380 015*<br>
>>> > > > >> ><br>
>>> > > > >> > *Gujarat, INDIA*<br>
>>> > > > >> > *Contact: +91-79-2691-6054 (O), +=
91-79-2686-1929(R)*<br>
>>> > > > >> ><br>
>>> > > > >> > On Tue, Nov 18, 2014 at 10:42 PM,=
Jean-Michel Campin <<br>
>>> > > <a href=3D"mailto:jmc at ocean.mit.edu">jmc at ocean.mit.e=
du</a><br>
>>> > > > >> ><br>
>>> > > > >> > wrote:<br>
>>> > > > >> ><br>
>>> > > > >> > > Hi Abhisek,<br>
>>> > > > >> > ><br>
>>> > > > >> > > I would recommand to concent=
rate first on any level except the<br>
>>> > > surface<br>
>>> > > > >> > > (k=3D1),<br>
>>> > > > >> > > and try to close the budget =
; then, once you get a clean closed<br>
>>> > > budget<br>
>>> > > > >> > > for levels at depth, you can=
revisit the surface level problem.<br>
>>> > > > >> > ><br>
>>> > > > >> > > at depth (k>1):<br>
>>> > > > >> > > the way you estimate the ten=
dency is not what is done in the<br>
>>> > > model,<br>
>>> > > > >> > > see e.g.:<br>
>>> > > > >> > ><br>
>>> > > > >><br>
>>> > > <a href=3D"http://mitgcm.org/public/r2_manual/latest=
/online_documents/node73.html">http://mitgcm.org/public/r2_manual/latest/on=
line_documents/node73.html</a><br>
>>> > > > >> > ><br>
>>> > > > >> > > The diagnostics for the KPP =
non-local term is like a vertical flux<br>
>>> > > > >> > > (added to fVerT), so it can =
be treated (for budget purpose) the<br>
>>> > > same<br>
>>> > > > >> way<br>
>>> > > > >> > > as DIFrI_TH.<br>
>>> > > > >> > ><br>
>>> > > > >> > > at the surface, you can sear=
ch the mitgcm-support archive and find<br>
>>> > > > >> some<br>
>>> > > > >> > > information there, e.g.:<br>
>>> > > > >> > > <a href=3D"http://mitgcm.org=
/pipermail/mitgcm-support/2014-April/009093.html">http://mitgcm.org/piperma=
il/mitgcm-support/2014-April/009093.html</a><br>
>>> > > > >> > > but it's likely that you=
would need additional diagnostics<br>
>>> > > output,i<br>
>>> > > > >> > > e.g. WTHMASS & WSLTMASS =
at k=3D1.<br>
>>> > > > >> > ><br>
>>> > > > >> > > Cheers,<br>
>>> > > > >> > > Jean-Michel<br>
>>> > > > >> > ><br>
>>> > > > >> > > On Tue, Nov 18, 2014 at 09:4=
9:34PM +0530, Abhisek Chakraborty<br>
>>> > > wrote:<br>
>>> > > > >> > > > Many thanks Ryan, Dimit=
ris, John and Gael for your suggestions.<br>
>>> > > > >> > > ><br>
>>> > > > >> > > ><br>
>>> > > > >> > > > I am not much familiar =
with MATLAB, so for me it was too much<br>
>>> > > > >> difficult<br>
>>> > > > >> > > to<br>
>>> > > > >> > > > understand the code sha=
red by Gael, though I tried to look into<br>
>>> > > it.<br>
>>> > > > >> > > ><br>
>>> > > > >> > > ><br>
>>> > > > >> > > > As a matter of fact is =
that I am using is the ECCO1 version (1<br>
>>> > > deg<br>
>>> > > > >> > > global,<br>
>>> > > > >> > > > 50 levels, excluding po=
les; i=3D1:360; j=3D1:160; k=3D1:50) of MITGCM<br>
>>> > > with<br>
>>> > > > >> > > linear<br>
>>> > > > >> > > > free surface and linFSC=
onserveTR=3DTrue. The diagnostic<br>
>>> > > parameters are<br>
>>> > > > >> > > > written in every day wi=
th diagnostic timephase=3D0. Model<br>
>>> > > > >> parameterizations<br>
>>> > > > >> > > > are KPP and GM-Redi.<br=
>
>>> > > > >> > > ><br>
>>> > > > >> > > ><br>
>>> > > > >> > > > I=E2=80=99d like to ana=
lyze the heat budget at a particular location,<br>
>>> > > say,<br>
>>> > > > >> > > (i,j,k).<br>
>>> > > > >> > > > So, from the suggestion=
s by Ryan, Dimitris and looking into the<br>
>>> > > > >> Gael=E2=80=99s<br>
>>> > > > >> > > > code, I am trying the f=
ollowing:<br>
>>> > > > >> > > ><br>
>>> > > > >> > > ><br>
>>> > > > >> > > > !!Cell volumes (center =
difference)<br>
>>> > > > >> > > ><br>
>>> > > > >> > > > Cvx =3D (RAC(i+1,j) - R=
AC(i-1,j)) * RF(k) * hFacC(i,j,k)<br>
>>> > > > >> > > ><br>
>>> > > > >> > > > Cvy =3D (RAC(i,j+1) - R=
AC(i,j-1)) * RF(k) * hFacC(i,j,k)<br>
>>> > > > >> > > ><br>
>>> > > > >> > > > Cvz =3D RAC(i,j) * (RF(=
k+1) =E2=80=93 RF(k)) * hFacC(i,j,k)<br>
>>> > > > >> > > ><br>
>>> > > > >> > > ><br>
>>> > > > >> > > ><br>
>>> > > > >> > > > !!Divergence of advecti=
on & diffusion by center difference<br>
>>> > > > >> > > ><br>
>>> > > > >> > > ><br>
>>> > > > >> > > > if (k>1) then<br>
>>> > > > >> > > ><br>
>>> > > > >> > > >=C2=A0 =C2=A0 =C2=A0 =C2=
=A0 =C2=A0 adc_dif_vert_k =3D (ADVr_TH(i,j,k+1) =E2=80=93<br>
>>> > > ADVr_TH(I,j,k-1))/cvz<br>
>>> > > > >> +<br>
>>> > > > >> > > ><br>
>>> > > > >> > > >=C2=A0 =C2=A0 =C2=A0 =C2=
=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =
=C2=A0 =C2=A0(DIFrE_TH(I,j,k+1) =E2=80=93<br>
>>> > > > >> > > DIFrE_TH(I,j,k-1))/cvz +<br>
>>> > > > >> > > ><br>
>>> > > > >> > > >=C2=A0 =C2=A0 =C2=A0 =C2=
=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =
=C2=A0 =C2=A0(DIFrI_TH(I,j,k+1) =E2=80=93<br>
>>> > > > >> DIFrI_TH(I,j,k-1))/cvz<br>
>>> > > > >> > > ><br>
>>> > > > >> > > > else<br>
>>> > > > >> > > ><br>
>>> > > > >> > > >=C2=A0 =C2=A0 =C2=A0 =C2=
=A0 =C2=A0 adc_dif_vert_k =3D 0.0<br>
>>> > > > >> > > ><br>
>>> > > > >> > > > endif<br>
>>> > > > >> > > ><br>
>>> > > > >> > > ><br>
>>> > > > >> > > ><br>
>>> > > > >> > > > adc_dif_hori_k =3D (ADV=
x_TH(i+1,j,k) =E2=80=93 ADVx_TH(i-1,j,k))/cvx +<br>
>>> > > > >> > > ><br>
>>> > > > >> > > >=C2=A0 =C2=A0 =C2=A0 =C2=
=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 (ADVy_TH=
(i,j+1,k) =E2=80=93<br>
>>> > > ADVy_TH(i,j-1,k))/cvy +<br>
>>> > > > >> > > ><br>
>>> > > > >> > > >=C2=A0 =C2=A0 =C2=A0 =C2=
=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 (DIFxE_T=
H(i+1,j,k) =E2=80=93<br>
>>> > > > >> DIFxE_TH(i-1,j,k))/cvx +<br>
>>> > > > >> > > ><br>
>>> > > > >> > > >=C2=A0 =C2=A0 =C2=A0 =C2=
=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 (DIFyE_T=
H(i,j+1,k) =E2=80=93<br>
>>> > > DIFyE_TH(i,j-1,k))/cvy<br>
>>> > > > >> > > ><br>
>>> > > > >> > > ><br>
>>> > > > >> > > > !!To account for the=C2=
=A0 shortwave heating<br>
>>> > > > >> > > ><br>
>>> > > > >> > > > depth=3DRF(k)<br>
>>> > > > >> > > ><br>
>>> > > > >> > > > swfrac=3D0.62*exp(depth=
/0.6) + (1.0-0.62) * exp(depth/20)<br>
>>> > > > >> > > ><br>
>>> > > > >> > > > if(depth < -200.0) t=
hen<br>
>>> > > > >> > > ><br>
>>> > > > >> > > > swfrac=3D0.0<br>
>>> > > > >> > > ><br>
>>> > > > >> > > > endif<br>
>>> > > > >> > > ><br>
>>> > > > >> > > > if (k=3D1)then<br>
>>> > > > >> > > ><br>
>>> > > > >> > > >=C2=A0 =C2=A0 =C2=A0 =C2=
=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 swfrac=3D1.0<br>
>>> > > > >> > > ><br>
>>> > > > >> > > > endif<br>
>>> > > > >> > > ><br>
>>> > > > >> > > ><br>
>>> > > > >> > > > !!Finally balance<br>
>>> > > > >> > > ><br>
>>> > > > >> > > ><br>
>>> > > > >> > > > TOTTTEND/86400.0 =3D - =
adc_dif_vert_k -=C2=A0 adc_dif_hori_k +<br>
>>> > > > >> > > > (oceQsw/(rhoConst*Cp)) =
* swfrac * hFacC(I,j,k) + KPPg_TH(i,j,k)<br>
>>> > > > >> > > ><br>
>>> > > > >> > > ><br>
>>> > > > >> > > > Please let me know whet=
her I am doing properly or not.<br>
>>> > > > >> > > ><br>
>>> > > > >> > > ><br>
>>> > > > >> > > ><br>
>>> > > > >> > > > Thanking in advance,<br=
>
>>> > > > >> > > ><br>
>>> > > > >> > > > Abhisek<br>
>>> > > > >> > > ><br>
>>> > > > >> > > ><br>
>>> > > > >> > > ><br>
>>> > > > >> > > ><br>
>>> > > > >> > > ><br>
>>> > > > >> > > ><br>
>>> > > > >> > > ><br>
>>> > > > >> > > ><br>
>>> > > > >> > > ><br>
>>> > > > >> > > ><br>
>>> > > > >> > > ><br>
>>> > > > >> > > ><br>
>>> > > > >> > ><br>
>>> > > > >><br>
>>> > > *---------------------------------------------------=
-----------------------------ABHISEK<br>
>>> > > > >> > > > CHAKRABORTY*<br>
>>> > > > >> > > ><br>
>>> > > > >> > > > *Scientist - "SD&q=
uot;Oceanic Sciences Division (*<br>
>>> > > > >> > > > *AOSG/EPSA)*<br>
>>> > > > >> > > ><br>
>>> > > > >> > > > *Space Applications Cen=
tre*<br>
>>> > > > >> > > ><br>
>>> > > > >> > > > *Indian Space Research =
Organization*<br>
>>> > > > >> > > ><br>
>>> > > > >> > > > *Ahmedabad - 380 015*<b=
r>
>>> > > > >> > > ><br>
>>> > > > >> > > > *Gujarat, INDIA*<br>
>>> > > > >> > > > *Contact: +91-79-2691-6=
054 (O), +91-79-2686-1929(R)*<br>
>>> > > > >> > > ><br>
>>> > > > >> > > > On Tue, Nov 18, 2014 at=
5:06 AM, gael forget <<a href=3D"mailto:gforget at mit.edu">gforget at mit.ed=
u</a>><br>
>>> > > > >> wrote:<br>
>>> > > > >> > > ><br>
>>> > > > >> > > > > Hi Abhisek, et al,=
<br>
>>> > > > >> > > > ><br>
>>> > > > >> > > > > the following matl=
ab code may provide some guidance :<br>
>>> > > > >> > > > ><br>
>>> > > > >> > > > ><br>
>>> > > > >> > ><br>
>>> > > > >><br>
>>> > > <a href=3D"http://mitgcm.org/viewvc/MITgcm/MITgcm_co=
ntrib/gael/matlab_class/gcmfaces_diags/diags_set_D.m">http://mitgcm.org/vie=
wvc/MITgcm/MITgcm_contrib/gael/matlab_class/gcmfaces_diags/diags_set_D.m</a=
><br>
>>> > > > >> > > > ><br>
>>> > > > >> > > > > It does not cover =
every possible combination of model options<br>
>>> > > > >> > > > > (e.g., no kpp or t=
hsice) of course, but treats a few. In that<br>
>>> > > > >> code,<br>
>>> > > > >> > > > > it is assumed that=
<br>
>>> > > > >> > > > > 1) the setting for=
the diagnostics package was according to<br>
>>> > > > >> > > > ><br>
>>> > > > >> > > > ><br>
>>> > > > >> > ><br>
>>> > > > >><br>
>>> > > <a href=3D"http://mitgcm.org/viewvc/MITgcm/MITgcm_co=
ntrib/gael/verification/global_oce_llc90/input/data.diagnostics">http://mit=
gcm.org/viewvc/MITgcm/MITgcm_contrib/gael/verification/global_oce_llc90/inp=
ut/data.diagnostics</a><br>
>>> > > > >> > > > >=C2=A0 =C2=A0 =C2=
=A0 which includes vertical integration. The relevant<br>
>>> > > diagnostics<br>
>>> > > > >> > > > >=C2=A0 =C2=A0 =C2=
=A0 for budget are those with filename like<br>
>>> > > > >> 'diags/budg2d_snap_set1',<br>
>>> > > > >> > > > > 2) tendencies have=
been pre-computed by differencing between<br>
>>> > > > >> > > > >=C2=A0 =C2=A0 =C2=
=A0consecutive (monthly) snap shots =E2=80=94 so that ETAN<br>
>>> > > > >> > > > >=C2=A0 =C2=A0 =C2=
=A0really means dETAN/dt in this context, and similarly for<br>
>>> > > > >> > > > >=C2=A0 =C2=A0 =C2=
=A0THETA, SALT, etc. The other variable names used in<br>
>>> > > > >> > > > >=C2=A0 =C2=A0 =C2=
=A0diags_set_D.m<br>
>>> > > > >> > > > > <<br>
>>> > > > >> > ><br>
>>> > > > >><br>
>>> > > <a href=3D"http://mitgcm.org/viewvc/MITgcm/MITgcm_co=
ntrib/gael/matlab_class/gcmfaces_diags/diags_set_D.m">http://mitgcm.org/vie=
wvc/MITgcm/MITgcm_contrib/gael/matlab_class/gcmfaces_diags/diags_set_D.m</a=
><br>
>>> > > > >> ><br>
>>> > > > >> > > match<br>
>>> > > > >> > > > > those used in pkg/=
diagnostics<br>
>>> > > > >> > > > ><br>
>>> > > > >> > > > > I should also ment=
ion that diags_set_D.m<br>
>>> > > > >> > > > > <<br>
>>> > > > >> > ><br>
>>> > > > >><br>
>>> > > <a href=3D"http://mitgcm.org/viewvc/MITgcm/MITgcm_co=
ntrib/gael/matlab_class/gcmfaces_diags/diags_set_D.m">http://mitgcm.org/vie=
wvc/MITgcm/MITgcm_contrib/gael/matlab_class/gcmfaces_diags/diags_set_D.m</a=
><br>
>>> > > > >> ><br>
>>> > > > >> > > is<br>
>>> > > > >> > > > > part of a<br>
>>> > > > >> > > > > broader framework =
(gcmfaces) which is documented in<br>
>>> > > > >> > > > ><br>
>>> > > > >> > > > ><br>
>>> > > > >> > ><br>
>>> > > > >><br>
>>> > > <a href=3D"http://mitgcm.org/viewvc/*checkout*/MITgc=
m/MITgcm_contrib/gael/matlab_class/gcmfaces.pd">http://mitgcm.org/viewvc/*c=
heckout*/MITgcm/MITgcm_contrib/gael/matlab_class/gcmfaces.pd</a><br>
>>> > > > >> > > > > <<br>
>>> > > > >> > ><br>
>>> > > > >><br>
>>> > > <a href=3D"http://mitgcm.org/viewvc/MITgcm/MITgcm_co=
ntrib/gael/matlab_class/gcmfaces.pd">http://mitgcm.org/viewvc/MITgcm/MITgcm=
_contrib/gael/matlab_class/gcmfaces.pd</a><br>
>>> > > > >> > > ><br>
>>> > > > >> > > > > f<br>
>>> > > > >> > > > > =E2=80=A6 although=
I am afraid this documentation is also<br>
>>> > > > >> > > > > lacking wrt the bu=
dget diagnostic codes.<br>
>>> > > > >> > > > ><br>
>>> > > > >> > > > > Cheers,<br>
>>> > > > >> > > > > Gael<br>
>>> > > > >> > > > ><br>
>>> > > > >> > > > > On Nov 17, 2014, a=
t 5:53 PM, Dimitris Menemenlis <<br>
>>> > > > >> > > <a href=3D"mailto:dmenemenli=
s at gmail.com">dmenemenlis at gmail.com</a>><br>
>>> > > > >> > > > > wrote:<br>
>>> > > > >> > > > ><br>
>>> > > > >> > > > > =E2=80=A6 and you =
probably already know this, but not clear from your<br>
>>> > > > >> email:<br>
>>> > > > >> > > > > shortwave is not d=
umped in surface level only, but rather is<br>
>>> > > > >> > > distributed<br>
>>> > > > >> > > > > with exponentially=
<br>
>>> > > > >> > > > > decaying profile i=
n top 200-m, with subtle modifications if<br>
>>> > > this<br>
>>> > > > >> > > heating<br>
>>> > > > >> > > > > is occurring withi=
n<br>
>>> > > > >> > > > > the KPP mixing lay=
er depth.<br>
>>> > > > >> > > > ><br>
>>> > > > >> > > > > On Nov 17, 2014, a=
t 2:44 PM, Ryan Abernathey <<br>
>>> > > > >> > > <a href=3D"mailto:ryan.abern=
athey at gmail.com">ryan.abernathey at gmail.com</a>><br>
>>> > > > >> > > > > wrote:<br>
>>> > > > >> > > > ><br>
>>> > > > >> > > > > Abhisek,<br>
>>> > > > >> > > > ><br>
>>> > > > >> > > > > This is a very com=
mon question that unfortunately is not<br>
>>> > > > >> documented<br>
>>> > > > >> > > very<br>
>>> > > > >> > > > > well. A couple of =
suggestions...<br>
>>> > > > >> > > > ><br>
>>> > > > >> > > > > For the transport =
terms, (ADV* and DF*), keep in mind that<br>
>>> > > these<br>
>>> > > > >> are<br>
>>> > > > >> > > FLUX<br>
>>> > > > >> > > > > terms, defined on =
the cell boundaries. In order to calculate a<br>
>>> > > > >> heat<br>
>>> > > > >> > > budget,<br>
>>> > > > >> > > > > you need to calcul=
ate the DIVERGENCE of those fluxes.<br>
>>> > > > >> > > > ><br>
>>> > > > >><br>
>>> > > <a href=3D"http://mitgcm.org/public/r2_manual/latest=
/online_documents/node71.html">http://mitgcm.org/public/r2_manual/latest/on=
line_documents/node71.html</a><br>
>>> > > > >> > > > > Maybe this is obvi=
ous, but your email did not make it clear<br>
>>> > > > >> whether you<br>
>>> > > > >> > > > > were doing this or=
not.<br>
>>> > > > >> > > > ><br>
>>> > > > >> > > > > Likewise, TFLUX is=
the downward flux at the surface, in<br>
>>> > > W/m^2. In<br>
>>> > > > >> > > order to<br>
>>> > > > >> > > > > convert to tempera=
ture units, you need to divide by<br>
>>> > > > >> > > > > HeatCapacity_Cp*rU=
nit2mass. For z coordinates, rUnit2mass is<br>
>>> > > equal<br>
>>> > > > >> > > > > to rhoConst.<br>
>>> > > > >> > > > ><br>
>>> > > > >> > > > > There are some sub=
tleties regarding tracer budgets under<br>
>>> > > > >> different free<br>
>>> > > > >> > > > > surface treatments=
. Maybe someone else on the list understands<br>
>>> > > > >> this<br>
>>> > > > >> > > issue<br>
>>> > > > >> > > > > better and is will=
ing to explain. Anyway, those should be<br>
>>> > > > >> second-order<br>
>>> > > > >> > > > > effects.<br>
>>> > > > >> > > > ><br>
>>> > > > >> > > > > Good luck with you=
r analysis.<br>
>>> > > > >> > > > ><br>
>>> > > > >> > > > > -Ryan<br>
>>> > > > >> > > > ><br>
>>> > > > >> > > > ><br>
>>> > > > >> > > > > On Sat, Nov 15, 20=
14 at 2:29 AM, Abhisek Chakraborty <<br>
>>> > > > >> > > > > <a href=3D"mailto:=
abhisek.sac at gmail.com">abhisek.sac at gmail.com</a>> wrote:<br>
>>> > > > >> > > > ><br>
>>> > > > >> > > > >> Dear Users,<br=
>
>>> > > > >> > > > >><br>
>>> > > > >> > > > >> I am trying to=
analyze heat budget in MITGCM. From the<br>
>>> > > diagnostic<br>
>>> > > > >> > > > >> outputs I have=
the following terms<br>
>>> > > > >> > > > >><br>
>>> > > > >> > > > >> TOTTTEND , ADV=
r_TH, ADVx_TH, ADVy_TH, DFrE_TH, DFxE_TH,<br>
>>> > > DFyE_TH,<br>
>>> > > > >> > > > >> DFrI_TH, KPPg_=
TH, TFLUX (i.e. all terms corresponding to<br>
>>> > > > >> tendency,<br>
>>> > > > >> > > > >> advection, dif=
fusion, KPP and total heat flux).<br>
>>> > > > >> > > > >><br>
>>> > > > >> > > > >> Apart from the=
se, I have other usual outputs like THETA, SALT<br>
>>> > > > >> etc.<br>
>>> > > > >> > > > >><br>
>>> > > > >> > > > >> The tendency t=
erm is in degC/day, the other terms are in<br>
>>> > > > >> degC*m^3/s<br>
>>> > > > >> > > > >> and TFLUX is i=
n W/m^2. Thus I have converted the tendency<br>
>>> > > term<br>
>>> > > > >> into<br>
>>> > > > >> > > > >> degC/s by divi=
ding by 86400.0 and the advection-diffusion<br>
>>> > > terms<br>
>>> > > > >> are<br>
>>> > > > >> > > > >> divided by cel=
l volume (=3DRAC*RF*HFACC).<br>
>>> > > > >> > > > >><br>
>>> > > > >> > > > >> I am using &qu=
ot;linFSconserveTr=3DTrue".<br>
>>> > > > >> > > > >><br>
>>> > > > >> > > > >> For the surfac=
e layer I have to consider TFLUX (watt/m^2),<br>
>>> > > but<br>
>>> > > > >> how to<br>
>>> > > > >> > > > >> convert it int=
o degC/s unit?<br>
>>> > > > >> > > > >><br>
>>> > > > >> > > > >> For subsurface=
levels, I am trying to equate tendency term<br>
>>> > > with<br>
>>> > > > >> > > > >> (advection + d=
iffusion + KPP term). But there is a mismatch.<br>
>>> > > > >> > > > >><br>
>>> > > > >> > > > >> Can somebody p=
lease guide me how to achieve the exact heat<br>
>>> > > > >> budget for<br>
>>> > > > >> > > > >> both surface a=
nd subsurface levels ?<br>
>>> > > > >> > > > >><br>
>>> > > > >> > > > >> Thanking in ad=
vance,<br>
>>> > > > >> > > > >> Abhisek<br>
>>> > > > >> > > > >><br>
>>> > > > >> > > > >> --<br>
>>> > > > >> > > > >><br>
>>> > > > >> > > > >><br>
>>> > > > >> > > > >> *ABHISEK CHAKR=
ABORTY*<br>
>>> > > > >> > > > >><br>
>>> > > > >> > > > >> *Scientist - &=
quot;SD"Oceanic Sciences Division (*AOSG/EPSA)*<br>
>>> > > > >> > > > >><br>
>>> > > > >> > > > >> *Space Applica=
tions Centre*<br>
>>> > > > >> > > > >><br>
>>> > > > >> > > > >> *Indian Space =
Research Organization*<br>
>>> > > > >> > > > >><br>
>>> > > > >> > > > >> *Ahmedabad - 3=
80 015*<br>
>>> > > > >> > > > >><br>
>>> > > > >> > > > >> *Gujarat, INDI=
A*<br>
>>> > > > >> > > > >> *Contact: +91-=
79-2691-6054 (O), +91-79-2686-1929(R)*<br>
>>> > > > >> > > > >><br>
>>> > > > >> > > > ><br>
>>> > > > >> > > > > __________________=
_____________________________<br>
>>> > > > >> > > > > MITgcm-support mai=
ling list<br>
>>> > > > >> > > > > <a href=3D"mailto:=
MITgcm-support at mitgcm.org">MITgcm-support at mitgcm.org</a><br>
>>> > > > >> > > > > <a href=3D"http://=
mitgcm.org/mailman/listinfo/mitgcm-support">http://mitgcm.org/mailman/listi=
nfo/mitgcm-support</a><br>
>>> > > > >> > > > ><br>
>>> > > > >> > > > ><br>
>>> > > > >> > > > ><br>
>>> > > > >> > > > > __________________=
_____________________________<br>
>>> > > > >> > > > > MITgcm-support mai=
ling list<br>
>>> > > > >> > > > > <a href=3D"mailto:=
MITgcm-support at mitgcm.org">MITgcm-support at mitgcm.org</a><br>
>>> > > > >> > > > > <a href=3D"http://=
mitgcm.org/mailman/listinfo/mitgcm-support">http://mitgcm.org/mailman/listi=
nfo/mitgcm-support</a><br>
>>> > > > >> > > > ><br>
>>> > > > >> > > > ><br>
>>> > > > >> > ><br>
>>> > > > >> > > > _______________________=
________________________<br>
>>> > > > >> > > > MITgcm-support mailing =
list<br>
>>> > > > >> > > > <a href=3D"mailto:MITgc=
m-support at mitgcm.org">MITgcm-support at mitgcm.org</a><br>
>>> > > > >> > > > <a href=3D"http://mitgc=
m.org/mailman/listinfo/mitgcm-support">http://mitgcm.org/mailman/listinfo/m=
itgcm-support</a><br>
>>> > > > >> > ><br>
>>> > > > >> > ><br>
>>> > > > >> > > ____________________________=
___________________<br>
>>> > > > >> > > MITgcm-support mailing list<=
br>
>>> > > > >> > > <a href=3D"mailto:MITgcm-sup=
port at mitgcm.org">MITgcm-support at mitgcm.org</a><br>
>>> > > > >> > > <a href=3D"http://mitgcm.org=
/mailman/listinfo/mitgcm-support">http://mitgcm.org/mailman/listinfo/mitgcm=
-support</a><br>
>>> > > > >> > ><br>
>>> > > > >><br>
>>> > > > >> > _________________________________=
______________<br>
>>> > > > >> > MITgcm-support mailing list<br>
>>> > > > >> > <a href=3D"mailto:MITgcm-support@=
mitgcm.org">MITgcm-support at mitgcm.org</a><br>
>>> > > > >> > <a href=3D"http://mitgcm.org/mail=
man/listinfo/mitgcm-support">http://mitgcm.org/mailman/listinfo/mitgcm-supp=
ort</a><br>
>>> > > > >><br>
>>> > > > >><br>
>>> > > > >> ______________________________________=
_________<br>
>>> > > > >> MITgcm-support mailing list<br>
>>> > > > >> <a href=3D"mailto:MITgcm-support at mitgc=
m.org">MITgcm-support at mitgcm.org</a><br>
>>> > > > >> <a href=3D"http://mitgcm.org/mailman/l=
istinfo/mitgcm-support">http://mitgcm.org/mailman/listinfo/mitgcm-support</=
a><br>
>>> > > > >><br>
>>> > > > ><br>
>>> > > > ><br>
>>> > ><br>
>>> > > > _______________________________________________=
<br>
>>> > > > MITgcm-support mailing list<br>
>>> > > > <a href=3D"mailto:MITgcm-support at mitgcm.org">MI=
Tgcm-support at mitgcm.org</a><br>
>>> > > > <a href=3D"http://mitgcm.org/mailman/listinfo/m=
itgcm-support">http://mitgcm.org/mailman/listinfo/mitgcm-support</a><br>
>>> > ><br>
>>> > ><br>
>>> > > _______________________________________________<br>
>>> > > MITgcm-support mailing list<br>
>>> > > <a href=3D"mailto:MITgcm-support at mitgcm.org">MITgcm-=
support at mitgcm.org</a><br>
>>> > > <a href=3D"http://mitgcm.org/mailman/listinfo/mitgcm=
-support">http://mitgcm.org/mailman/listinfo/mitgcm-support</a><br>
>>> > ><br>
>>><br>
>>> > _______________________________________________<br>
>>> > MITgcm-support mailing list<br>
>>> > <a href=3D"mailto:MITgcm-support at mitgcm.org">MITgcm-suppo=
rt at mitgcm.org</a><br>
>>> > <a href=3D"http://mitgcm.org/mailman/listinfo/mitgcm-supp=
ort">http://mitgcm.org/mailman/listinfo/mitgcm-support</a><br>
>>><br>
>>><br>
>>> _______________________________________________<br>
>>> MITgcm-support mailing list<br>
>>> <a href=3D"mailto:MITgcm-support at mitgcm.org">MITgcm-support at mi=
tgcm.org</a><br>
>>> <a href=3D"http://mitgcm.org/mailman/listinfo/mitgcm-support">=
http://mitgcm.org/mailman/listinfo/mitgcm-support</a><br>
>><br>
>><br>
>> _______________________________________________<br>
>> MITgcm-support mailing list<br>
>> <a href=3D"mailto:MITgcm-support at mitgcm.org">MITgcm-support at mitgcm=
.org</a><br>
>> <a href=3D"http://mitgcm.org/mailman/listinfo/mitgcm-support">http=
://mitgcm.org/mailman/listinfo/mitgcm-support</a><br>
>><br>
><br>
><br>
> _______________________________________________<br>
> MITgcm-support mailing list<br>
> <a href=3D"mailto:MITgcm-support at mitgcm.org">MITgcm-support at mitgcm.org=
</a><br>
> <a href=3D"http://mitgcm.org/mailman/listinfo/mitgcm-support">http://m=
itgcm.org/mailman/listinfo/mitgcm-support</a><br>
><br>
</p>
--089e013c6a7e5c5e6d0508d95e43--
More information about the MITgcm-support
mailing list