GC7: testing impact CaMaFlood rivers in GOSI

[ukmo-cguiavarch]

Fitsum Woldemeskel fitsum.woldemeskel@bom.gov.au has prepared a climatology of river discharges derived from CaMaFlood simulations.
The climatology has been copied here: /data/users/harley.kelly/Discharge_data_for_NEMO_test

• it is a daily climatology
• the period used for calculating the climatology is 2019-2023
• Fitsum has remapped the discharged on the eORCA025 grid follwing Dan Copsey's approach
• the river mouth "spreading" is similar to what is currently used in GOSI (Bourdalle-Badie&Treguier)

It has been agreed to test the climatology in GOSI10 in parallel with the work on JRA55-DO rivers see #21.

[ukmo-cguiavarch]

Preparing CaMaFlood ancil for GOSI10

The file provided does not include socoef, the mask used to apply local mixing at the river mouth, and need to be calculated.

The mask socoefr is calculated using the average climatology as follow:

• ading time_counter to file: ncks -o --mk_rec_dmn time_counter --cnk_plc=uck eORCA025_camaflood.nc eORCA025_camaflood.nc
• averaging: ncra eORCA025_camaflood.nc eORCA025_camaflood_rec_Avg.nc
• calculating the mask: ncap2 -O -s socoefr=(sorunoff>0) eORCA025_camaflood_rec_Avg.nc eORCA025_camaflood_rec_Avg.nc

socoefr has been added in the file located in exab:/data/users/catherine.guiavarch/GOSI10/rivers/camaflood/eORCA025_camaflood.nc

[ukmo-cguiavarch]

Testing CaMaFlood runoff in GOSI10

• CaMaFlood runoffs have been implemented in u-eb102.
• u-eb102. is based on the GOSI10.beta.6@5.0.2 u-ea039
• Changeset for CaMaFlood implementation 362143
• 65 years integration starting from 1958 forced by JRA55 and initialised from WOA

Results:

• Marine_assess y2011-2021
• implementation of CaMaFlood rivers has a very significant impact on the salinity ( unrealistically high salinity)
• increase in salinity triggers deep convection in Labrador Sea and GIN seas
• deep convection results in an unrealistic AMOC (28 Sv) with significant impact of temperature in the subpolar gyre.

WARNING: the provided ancil is not suitable , more investigation is required

[ukmo-cguiavarch]

Investigating issue with CaMaFlood

• CaMaFlood (CMT) ancil: exab:/data/users/catherine.guiavarch/GOSI10/rivers/camaflood/eORCA025_camaflood.nc
• Bourdalle-Badie and Treguier (BBT) ancil: exab:/data/users/nemo/GOSIdev/GOSI10p0/eORCA025/eORCA025_runoff_GO6_icb_r42.nc

1. Values of fluxes associated with rivers in CMT are O(100) smaller than those in BBT. See below the value for the Amazon river mouth. It results in a salinification and large scale changes described GC7: testing impact CaMaFlood rivers in GOSI #47 (comment)
2. CMT also displays (very small) negative values (especially in Arctic), there are very small values but negative freshwater flux is translated by the model as salt flux so it will also contribute to the salinification of the model.

CaMaFlood - Amazon river mouth (x=958,y=684) : maximum value 1.475 e-5

Bourdalle-Badie&Treguier - Amazon river mouth (x=958,y=684) : maximum value 0.002