search

NOAA-CEFI-Regional-Ocean-Modeling / ocean_BGC

3 stars 4 forks source link

jprod_lithdet update for COBALTv3 #41

Closed feida6996 closed 2 months ago

feida6996 commented 2 months ago

This addresses discussion point https://github.com/NOAA-CEFI-Regional-Ocean-Modeling/ocean_BGC/issues/35 by adjusting the production of lithogenic detritus and conducting tests.

The tests show that the new expression has eliminated the occurrence of abnormally high values of lithogenic detritus and the subsequent flux crash. Instead, the flux stabilizes and reaches equilibrium after approximately 10 years.

We will also verify other key variables to ensure that they exhibit expected behavior after this change.

Picture1
yichengt900 commented 2 months ago

Ok here are some updated preliminary results (1993-1997) from NWA12 with the new expression:

surface alk, dic, and omega: alk_dic_omega

Surface chl: regional_chl

So far, everything looks ok to me.

yichengt900 commented 2 months ago

@feida6996, based on the short-term results from the regional domain (see above Figs), I can let your PR sneak in this time if you can make the following changes:

  1. Follow this instruction to merge the recent updates from main branch to your feature branch;

  2. After 1, copy the following and update the baseline answers in your feature branch:

    Step,       Day,  Truncs,      Energy/Mass,      Maximum CFL,  Mean Sea Level,  Total Mass,  Mean Salin, Mean Temp, Frac Mass Err,   Salin Err,    Temp Err
        [days]                 [m2 s-2]           [Nondim]       [m]             [kg]         [PSU]      [degC]       [Nondim]        [PSU]        [degC]
 0,  731595.000,     0, En 4.0531850015098383E-27, CFL  0.00000, SL -0.0000E+00, M 6.97073E+15, S 35.2427, T  6.2462, Me  0.00E+00, Se  0.00E+00, Te  0.00E+00
 6,  731595.250,     0, En 2.0223578954766472E-06, CFL  0.00281, SL -3.4291E-06, M 6.97073E+15, S 35.2427, T  6.2460, Me -8.57E-10, Se -2.85E-07, Te -2.66E-06
12,  731595.500,     0, En 4.2897044161823591E-06, CFL  0.00294, SL -6.4127E-06, M 6.97073E+15, S 35.2427, T  6.2457, Me -7.46E-10, Se -6.81E-07, Te -6.81E-06
18,  731595.750,     0, En 1.2185516288182668E-05, CFL  0.00794, SL -9.2141E-06, M 6.97073E+15, S 35.2427, T  6.2459, Me -7.00E-10, Se -9.89E-07, Te -1.00E-05
24,  731596.000,     0, En 2.4391883629687393E-05, CFL  0.00870, SL -1.3774E-05, M 6.97073E+15, S 35.2427, T  6.2457, Me -1.14E-09, Se -1.40E-06, Te -1.42E-05
30,  731596.250,     0, En 2.9283640524189851E-05, CFL  0.00724, SL -1.8458E-05, M 6.97073E+15, S 35.2427, T  6.2454, Me -1.17E-09, Se -1.93E-06, Te -1.96E-05
36,  731596.500,     0, En 3.6652617806009578E-05, CFL  0.00746, SL -2.3211E-05, M 6.97073E+15, S 35.2427, T  6.2451, Me -1.19E-09, Se -2.37E-06, Te -2.42E-05
42,  731596.750,     0, En 4.2253748566466643E-05, CFL  0.00812, SL -2.7653E-05, M 6.97073E+15, S 35.2427, T  6.2454, Me -1.11E-09, Se -2.79E-06, Te -2.86E-05
48,  731597.000,     0, En 4.7781350583834568E-05, CFL  0.00941, SL -3.1143E-05, M 6.97073E+15, S 35.2427, T  6.2453, Me -8.73E-10, Se -3.05E-06, Te -3.14E-05

  3. Add a commit for unit correction if you have one.

CC @charliestock , please consider adding a comment for this new expression.

charliestock commented 2 months ago

Hi Yi-Cheng, Fei,

Sorry for the long response time on this. I was pulled away on a few other things this week. The solution looks good to me. While the jprod_lithdet fix will change answers, I expect it to be small.

As we discussed on Tuesday, it is possible that the coefficients will change if the model is reformulated. The current rationale is as follows: The filter feeders responsible for converting lithogenic material to lithogenic detritus are either medium or large zooplankton (but not small). Most of the small phytoplankton are consumed by microzooplankton. The value of 0.3 thus reflects the relative availability of small phytoplankton to medium zooplankton (i.e., filter feeders) rather than small zooplankton. This roughly comes out to 0.4/(0.4 + 1.0) = 0.29, which I rounded to 0.3 to not give a sense of false precision.

For medium phytoplankton, the analogous calculation is: (1.0 + 0.4)/(0.4 + 1.0 + 0.4) = 0.78, again rounded to 0.8 to not give a sense of false precision.

I'd also emphasize that the production of lithdet from lith is a coarse scaling that basically ensures that lithogenic material is acting to preserve detritus near regions with high dust and river inputs. There is room for improvement and sensitivities here, but it is really just a first order transfer function.

Once this pull request is in, I will add a comment to clarify the rationale for these numbers.

feida6996 commented 2 months ago

@feida6996, based on the short-term results from the regional domain (see above Figs), I can let your PR sneak in this time if you can make the following changes:

  1. Follow this instruction to merge the recent updates from main branch to your feature branch;
  2. After 1, copy the following and update the baseline answers in your feature branch:
  Step,       Day,  Truncs,      Energy/Mass,      Maximum CFL,  Mean Sea Level,  Total Mass,  Mean Salin, Mean Temp, Frac Mass Err,   Salin Err,    Temp Err
            [days]                 [m2 s-2]           [Nondim]       [m]             [kg]         [PSU]      [degC]       [Nondim]        [PSU]        [degC]
     0,  731595.000,     0, En 4.0531850015098383E-27, CFL  0.00000, SL -0.0000E+00, M 6.97073E+15, S 35.2427, T  6.2462, Me  0.00E+00, Se  0.00E+00, Te  0.00E+00
     6,  731595.250,     0, En 2.0223578954766472E-06, CFL  0.00281, SL -3.4291E-06, M 6.97073E+15, S 35.2427, T  6.2460, Me -8.57E-10, Se -2.85E-07, Te -2.66E-06
    12,  731595.500,     0, En 4.2897044161823591E-06, CFL  0.00294, SL -6.4127E-06, M 6.97073E+15, S 35.2427, T  6.2457, Me -7.46E-10, Se -6.81E-07, Te -6.81E-06
    18,  731595.750,     0, En 1.2185516288182668E-05, CFL  0.00794, SL -9.2141E-06, M 6.97073E+15, S 35.2427, T  6.2459, Me -7.00E-10, Se -9.89E-07, Te -1.00E-05
    24,  731596.000,     0, En 2.4391883629687393E-05, CFL  0.00870, SL -1.3774E-05, M 6.97073E+15, S 35.2427, T  6.2457, Me -1.14E-09, Se -1.40E-06, Te -1.42E-05
    30,  731596.250,     0, En 2.9283640524189851E-05, CFL  0.00724, SL -1.8458E-05, M 6.97073E+15, S 35.2427, T  6.2454, Me -1.17E-09, Se -1.93E-06, Te -1.96E-05
    36,  731596.500,     0, En 3.6652617806009578E-05, CFL  0.00746, SL -2.3211E-05, M 6.97073E+15, S 35.2427, T  6.2451, Me -1.19E-09, Se -2.37E-06, Te -2.42E-05
    42,  731596.750,     0, En 4.2253748566466643E-05, CFL  0.00812, SL -2.7653E-05, M 6.97073E+15, S 35.2427, T  6.2454, Me -1.11E-09, Se -2.79E-06, Te -2.86E-05
    48,  731597.000,     0, En 4.7781350583834568E-05, CFL  0.00941, SL -3.1143E-05, M 6.97073E+15, S 35.2427, T  6.2453, Me -8.73E-10, Se -3.05E-06, Te -3.14E-05
  1. Add a commit for unit correction if you have one.

CC @charliestock , please consider adding a comment for this new expression.

Hi @yichengt900,

I double checked and the feature branch is up to date with the main branch. The baseline answer was updated. Two unit changes were also added.