Hi all,
This problem has been confused me for a while and I am seeking for help here.
I was trying to configure a coupled 3D physical - fennel biology model for the MAB region, but the surface chlorophyll spatial distribution and the time of bloom always went wrong (comparing with satellite and in-situ observation). To be more specific, surprisingly the bloom shows during the summer months. So I decided to use the 1D bio_fennel to study the biological processes first, also, I want to see how vertical mixing in the simple 1D model could stimulate the fall bloom.
The model configuration is similar to bio_toy, i.e. I configured a 6*6 grid with vertical 30 layers for uniform water depth (I did a series of experiments by varying the depth from 30m to 100m); the boundaries are all periodic and the surface forcing I provided is daily atmospheric forcing; the initial conditions for NO3 is the empirical temperature-nitrate profile over continental shelf, based on the description in Fennel et. al. (2006) Auxiliary materials. Other biological variables are set small as 0.1 mmol N m-3.
For the biological parameters, I notice in Fennel et. al. (2008) Auxiliary materials there have been several modifications, so I just use the new sets.
After running this 1D simulation for 5 years, the results I expect to see should be 2 blooms in one year, one is in spring, another one starting from early fall. However, I always lost the fall bloom in the
simulations, however, the diffusivity coefficient are strong during all the winter months.
I am wondering can anyone give me some suggestions?
Questions about bio_fennel simulation
Re: Questions about bio_fennel simulation
Hi Shuwen,
It would be helpful if you could post some plots of your results. For the bio-toy case it would be useful to see time series of nutrient and phytoplankton profiles.
Cheers, Katja
It would be helpful if you could post some plots of your results. For the bio-toy case it would be useful to see time series of nutrient and phytoplankton profiles.
Cheers, Katja
Re: Questions about bio_fennel simulation
Hi Dr. Fennel,
Thank you so much for the quick reply. Actually I really need your suggestions. Since the region where I will at last configure for 3D coupled model has deepest water depth with only 100m, so at present the experiments I set up by running 1D bio-toy are with different water depth, 30m, 50m and 90m respectively.
The CPP I turned on are Bio_fennel, Bio_sediment, DENITRIFICATION, and due to the previous discussions in forum, I also considered DIURNAL_SRFLUX. The time series for six variables (I consider the vertical water column so chlorophyll won't be put here) during the 5-year experiments are as following:
Depth 30m
Depth 50m
Depth 90m
The white colors show when the variables are below their minimum thresholds, i.e. 1e-3, the xlabel are 12 months in each year and the ylabel are water depths. Two things surprised me:
1) With time interpolation, phytoplankton are almost depleted instead of showing an annual cycle, in all cases.
2) The deeper water depth, the more NO3 accumulates.
Because the final goal I want to achieve is a annual cycle, i.e. a spring bloom triggered by light availability and a fall-winter bloom due to strong mixing, now I turned off DENITRIFICATION, because for my understanding with denitrification the system is not conservative. The results for water depth 30m experiment is shown:
I am still confused by the vanish of the fall-winter bloom, if I suspect the vertical mixing is correct based on Mellor-Yamada 2.5 scheme.
Thanks in advance for any suggestions you give!
Shuwen
Thank you so much for the quick reply. Actually I really need your suggestions. Since the region where I will at last configure for 3D coupled model has deepest water depth with only 100m, so at present the experiments I set up by running 1D bio-toy are with different water depth, 30m, 50m and 90m respectively.
The CPP I turned on are Bio_fennel, Bio_sediment, DENITRIFICATION, and due to the previous discussions in forum, I also considered DIURNAL_SRFLUX. The time series for six variables (I consider the vertical water column so chlorophyll won't be put here) during the 5-year experiments are as following:
Depth 30m
Depth 50m
Depth 90m
The white colors show when the variables are below their minimum thresholds, i.e. 1e-3, the xlabel are 12 months in each year and the ylabel are water depths. Two things surprised me:
1) With time interpolation, phytoplankton are almost depleted instead of showing an annual cycle, in all cases.
2) The deeper water depth, the more NO3 accumulates.
Because the final goal I want to achieve is a annual cycle, i.e. a spring bloom triggered by light availability and a fall-winter bloom due to strong mixing, now I turned off DENITRIFICATION, because for my understanding with denitrification the system is not conservative. The results for water depth 30m experiment is shown:
I am still confused by the vanish of the fall-winter bloom, if I suspect the vertical mixing is correct based on Mellor-Yamada 2.5 scheme.
Thanks in advance for any suggestions you give!
Shuwen
Re: Questions about bio_fennel simulation
Dear Shuwen,
It looks like your fall mixing is not bringing enough nutrients into the euphotic zone to trigger a fall bloom. It may help to plot the mixed layer depth on top of your time series plots (at least to illustrate the reasons for the lack of a fall bloom). If this is meant to be just an idealized model case, you may want to modify your atmospheric forcing to drive deeper mixing earlier in the fall.
Also note, BIO_TOY is not very realistic in terms of the vertical nutrient profiles it produces. In the TOY model the nitrate profile (especially below the winter maximum depth of the mixed layer) is dominated by diffusion, while in a proper 3D model and in the real ocean lateral nutrient supply processes are really important producing qualitatively different profiles. One way around this would be to nudge to a more realistic nitrate profile at depth.
Hope this helps.
Cheers, Katja
It looks like your fall mixing is not bringing enough nutrients into the euphotic zone to trigger a fall bloom. It may help to plot the mixed layer depth on top of your time series plots (at least to illustrate the reasons for the lack of a fall bloom). If this is meant to be just an idealized model case, you may want to modify your atmospheric forcing to drive deeper mixing earlier in the fall.
Also note, BIO_TOY is not very realistic in terms of the vertical nutrient profiles it produces. In the TOY model the nitrate profile (especially below the winter maximum depth of the mixed layer) is dominated by diffusion, while in a proper 3D model and in the real ocean lateral nutrient supply processes are really important producing qualitatively different profiles. One way around this would be to nudge to a more realistic nitrate profile at depth.
Hope this helps.
Cheers, Katja