negative temperature

[mashinde]

Dear ROMS users,

I am running model with ERA40 daily data(00hrs)for 1995-2000. I am getting negative temperature. I don't know why it is giving such -ve temperature.
I attached here one figure from 249 day and also log file.

any suggestions/comments ?


Thanking you

[julian.kuhlmann]

Since the most extreme negative temperatures are located at the coasts: Have you checked if maybe the land-sea-mask of the forcing does not exactly match the one of your ocean model? You can turn on surface net heat flux output to see if you lose the heat toward the atmosphere.

[mashinde]

Dear julian,

I fix the problem. The problem was in input radiation flux. Model runs fine and no more negative temperatures.

Thanks

[tony1230]

hi mashinde

I'd like to hear what is the problem with the input radiation flux and why you did that.

thank you

-shou

[mashinde]

Hi Shou,

I replace "Surface thermal radiation" with "Surface thermal radiation downwards" from ERA40 data set.

regards

[tony1230]

Thank you mashinde, for your reply.

Is that mean that the "Surface thermal radiation" is the Net radiation from surface atmosphere into the water and the later one stands for just the short wave radiation downwards and without net longwave?

If that is the case, how it cause such a negative temperature. It's a interesting topic and i want to know more.

thank.

-shou

[arango]

Notice that the FAQ for the ECMWF ERA dataset has the following information:

What are the definitions of the radiation fields?

• Surface solar radiation downwards is the incident solar (shortwave) radiation and surface solar radiation is the net solar radiation. Surface net solar radiation, clear sky is a hypothetical field assuming no clouds.
• Surface thermal radiation downwards is the incident thermal (longwave) radiation and surface thermal radiation is the net thermal radiation. Surface net thermal radiation, clear sky is a hypothetical field assuming no clouds.
• Top solar radiation is the net solar (shortwave) radiation at the top of the atmosphere. Top net solar radiation, clear sky is a hypothetical field assuming no clouds.
• Top thermal radiation is the net thermal (longwave) radiation at the top of the atmosphere. Top net thermal radiation, clear sky is a hypothetical field assuming no clouds.

Notice that according to this documentation the Surface thermal radiation downwards is the ROMS longwave radiation downward (lwrad_down) and not the shortwave radiation (swrad).

It is always a good idea to read the documentation about any dataset carefully.

[xiaozhu557]

Hello Arango,
I have got the same problem when I simulated Bohai Sea. But I also found the other problem is that sea water was over heated along the coast of Bohai Sea in summer, such as

[xiaozhu557]

I have used the surface forcing data from NCEP Climate Forecast System Reanalysis (CFSR). The data base has provided some variables such as:

Code: Select all

'UFLX_surface'  ...   %var1——momentum flux,U-Component [N/m^2]
        'VFLX_surface'  ...   %var2——momentum flux,V-Component [N/m^2]
        'DSWRF_surface' ...   %var3——downward short-wave Rad. flux [w/m^2]
        'USWRF_surface' ...   %var4——upward short-wave Rad. flux [w/m^2]
        'LHTFL_surface' ...   %var5——Latent heat net flux [w/m^2]
        'PRATE_surface' ...   %var6——Precipitation rate [kg/m^2/s]        
        'DLWRF_surface' ...   %var7——downward long-wave Rad. flux [w/m^2] 
        'ULWRF_surface' ...   %var8——upward long-wave Rad. flux [w/m^2]          
        'TMP_surface'   ...   %var9——0 m temp. [k]
        'UGRD_10maboveground' ...   %var10——10 m u wind [m/s]
        'VGRD_10maboveground' ...   %var11——10 m v wind [m/s]
	'SPFH_2maboveground' ...    %var12——2 m specific humidity [kg/kg]
        'SHTFL_surface' ...    %var13——surface sensible heat flux [w/m^2]
        'TMP_2maboveground'};  %var14——2 m temp. [k]

and I caculated the input variables for ROMS as follow:

Code: Select all

swrad= var3 - var4
        swflux=8640*(var5/10^6/2.5-var6)
        shflux=var7-var8+var3-var4-var5-var13;

Could you help me to check them? Is anything wrong for caculating the surface heat flux?

Many thanks.

[LiuZHQ]

Hi there,
I don't think that over-heated temperature in Bohai Sea during the summertime is because of any possible problem in ROMS itself, and the wrong thing is your configuration of the model and forcing you adapted in this domain.

Please re-review the related papers and you'll find the solution.

Happy modelling...
Zhiqiang

[tony1230]

Hi Zhiqiang,

You can tell explicitly what you think that may cause those kind of problems, or give some clear directions/guidance.

I would also like to know more about that.

Regards.

-shou

[LiuZHQ]

Hi Shou,
According to the figure and question "xiaozhu557" posted here. The so called "over-heated" coastal waters are his/her major concern.

Yes, the waters are warmer here. But, the temperature is still in reasonable range, and this implies that the wrong thing is not that strongly related to the coding or numerical algorithm of ROMS. You may want to refer to the figure given by "mashinde" at the very top and please note the differences. Then, I suggested "xiaozhu557" to check the forcing file and the flow field.

I'm not an expert in Bohai Sea studies and my study area is in the ECS.

But, in general, in the coastal areas, where the waters are very shallow (<10m in Bohai I believe), the waters are easily mixed well especially when we also take tides into consideration. The warmer waters heated by the heat flux in the surface layers occupied the coastal areas in the Bohai Sea, and possibly in the entire water column. This is a possible reason for that warm coastal water. This warm water may also change the local pressure field, and then the geostrophic currents and Ekman processes (related to upwelling/downwelling processes). It is not that easy to find out the complicated interactions among these processes, I think.

Secondly, when there is river plume (Yellow River), the warmer and fresher river discharges flow along the Bohai coast, and provides another source of warm waters. Buoyancy effects of the plume waters to coastal current are complicated also, and of course, the remote forcing from the ambient current and interaction between Bohai and Yellow Sea is important.

My point is we can not take model as a hundred percent reliable "world" that mirrors the reality. Complexity of the real ocean is far from our imagination.

For your reference and happy mid-autumn.

Zhiqiang

[tony1230]

Hi Zhiqiang,

Thank you so much.

Happy mid-autumn.

Diaoyu islands are belong to China!!!

-shou

[anthena]

Hi everyone!
i am new to ROMS.Now i run a case in Bohai, and met the same question with negative temperature appeared near the coast,which seems unreasonable,but salinity seems right,i have checked the land-sea-mask, also the foring data,there are no problems. i have changed the horizontal diffusion and viscosity coefficient smaller, the negtative temperature is still there but delayed some time, is it implicate that i should change the coefficients to get the result correct?
any suggestions/comments?
Thanks

[arango]

Then, something is wrong with the surface heat flux forcing. You need to check the surface forcing fields.