JRA-55 vs. CORE2 Forcing

Comparison of HYCOM forced by JRA-55 and CORE2 forcing.

JRA-55: 1958-2015

CORE2:1958-2007

HYCOM 2.2.98 within CESM 1_3, SSS relaxation of 4 years/50m + 6month in Southern Ocean, starts from Levitus PHC2 (9 terms for 012 and 17 terms for 101), salinity normalization.

Evolution of SST, SSS, SSH, global temperature and global salinity for JRA-55 in black and CORE2 in blue.

CORE2 starting in 1948 in green; JRA-55 with CORE2 climatological rivers is in red (no normalization after year 1967, this simulation is actually re-integrated ).

JRA-55

CORE2

SST Bias
SSS Bias

2007

JRA-55

CORE2

Barotropic StreamFunction

Vertical Streamfunctions

Evolution of the Maximum Streamfunction at 26ºN

Winter Ice Cover for 2007

SSMI/NSIDC Climatology

JRA-55

CORE2

Winter Ice Thickness for 2007

IceSat/NSIDC Climatology

JRA-55

CORE2

Wind-stress for year 2007 for JRA-55 on the left and CORE2 on the right.

Evolution of Surface heat flux and freshwater flux for JRA-55 in black and CORE2 in blue (Freshwater flux should corresponds to a 0. saltflux ... ).

Surface Heat Flux (W/m2) for year 2007 for JRA-55 on the left and CORE2 on the right

Surface Freshwater Flux (mm/d) for year 2007 for JRA-55 on the left and CORE2 on the right

Sensitivity of AMOC to JRA-55 rivers

Experimental set-up :

REF: JRA-55 forcing and JRA-55 rivers  with 4y+6mo SSS relaxation

SSS6mo: JRA-55 forcing and JRA-55 rivers  with 6mo SSS relaxation

JRAriver: JRA-55 forcing and CORE2 climatological rivers with 4y+6mo SSS relaxation

CORE2: CORE2 forcing and rivers with  4y+6mo SSS relaxation

Comparison of the rivers flux between JRA-55 and CORE2 : here for year 1958

(No interannual variability for at least the first 10 years from what I could see)

  1. ✓ JRA-55 rivers  flux higher than in CORE2 for the Arctic

  1. ✓ The increase comes from the Atlantic side of the Arctic

  1. ✓ Main differences in Greenland+Iceland  and Hudson region

  2. ✓ JRA-55 Greenland+Iceland rivers are 10 times higher than CORE2 rivers

  3. ✓ JRA-55 Hudson Bay rivers 25% high than CORE2 rivers

  4. ✓ JRA-55 Greenland+Iceland rivers flux closer to the observations than CORE2 (Dukhovskoy et al., 2016)

Comparison of the evolution of the maximum AMOC at 26N and Arctic freshwater content :

  1. ✓ Collapse of AMOC with JRA-55 with SSS relaxation of 6 months or 4 years

  2. ✓ CORE2 AMOC variability similar after 1980 starting in 1948 or 1958

  3. ✓ JRA-55 with CORE2 rivers does better than JRA-55 REF but remains lower than CORE2.

  1. ✓ Freshwater content for the JRA-55 REF not stabilizing after 1 cycle.

  2. ✓ A 6 month SSS relaxation inhibits the freshwater content increase after 30 years, but still larger than in CORE2 simulation

  3. ✓ JRA-55 with CORE2 rivers  freshwater content behaves like the CORE2 simulations until the 1990s but fails to stabilize after that.

Year 1982

Year 2007

JRA-55 REF

CORE2 START1958

JRA-55 CORE2 rivers

JRA-55 REF

CORE2 START1958

JRA-55 CORE2 rivers

SSH

Ano SSH

SSH

Ano SSH

Comparison JRA-55 REF vs. JRA-55 CORE2 rivers

  1. ✓ CORE2 rivers helped a little bit but not enough to keep a reasonable AMOC strength

  1. ✓ The freshwater content in the Arctic continues to increase without stabilizing in both JRA55-REF and JRA55-Rivers.

  1. ✓Similar evolution of temperature and salinity at the surface and globally in JRA55-REF and JRA55-Rivers

  1. ✓ Similar Biases at the surface with a North Atlantic slightly fresher in JRA55-REF than in JRA55-Rivers

  2. ✓ Slightly salty bias over the Arctic at the surface for both simulations.

  1. ✓ Similar Biases at 110m with a strong fresh bias over the Arctic and North Atlantic in both simulations

SSMI/NSIDC Climatology

IceSat/NSIDC Climatology

JRA55-REF

JRA55-Rivers

JRA55-REF

JRA55-Rivers

Sensitivity to SSS relaxation