This post gathers some information and diagnostics about the historical run (1850-2010 CE) performed with EC-EARTH3.3.
General information
The simulation was performed using EC-EARTH 3.3 (develop-fmasson branch). It starts in 1850 CE from a ~100 year-long control run, and is driven through time by CMIP6 forcing until 2010 CE.
Diagnostics
Global and hemispheric near-surface air temperatures
The simulated temperatures (HIS4 on the figures below) are too cold during the whole period, by about 1.5°C at the global scale compared to the CMIP6 EC-EARTH simulation. The recent warming is not strong enough.
Ice
Consistently with the cold bias in temperature, there is too much ice both in the North and South Poles, during the whole period 1850-2010 CE.
This ice overestimation holds during the last 20 years of the simulation. Apart from that, the recent (1990-2010 CE) spatial patterns of sea ice look realistic.
AMOC
The absolute values of the AMOC are realistic. Its relative changes do not correlate with the ice volume in the northern hemisphere.
Mixed layer depth over the Labrador Sea
Large box over the Labrador Sea (50-60°N, 65-45°W)
Single grid point over the Labrador Sea (57°N, 54°W)
Despite the fact that the simulation starts from conditions at the equilibrium in 1850 CE (after 100 years of control run), there is a strong drop in temperature and increase in sea ice over the first 40 years of simulation. No idea why, but this is an important issue.
The simulation is too cold.
This post gathers some information and diagnostics about the historical run (1850-2010 CE) performed with EC-EARTH3.3.
## General information
The simulation was performed using EC-EARTH 3.3 (develop-fmasson branch). It starts in 1850 CE from a ~100 year-long control run, and is driven through time by CMIP6 forcing until 2010 CE.
## Diagnostics
### Global and hemispheric near-surface air temperatures
The simulated temperatures (HIS4 on the figures below) are too cold during the whole period, by about 1.5°C at the global scale compared to the CMIP6 EC-EARTH simulation. The recent warming is not strong enough.
<img src="https://nextcloud.cism.ucl.ac.be/s/eLmTNnjW3cy77ef/preview">
### Ice
Consistently with the cold bias in temperature, there is too much ice both in the North and South Poles, during the whole period 1850-2010 CE.
<img src="https://nextcloud.cism.ucl.ac.be/s/q949qw42HgwKG3r/preview">
This ice overestimation holds during the last 20 years of the simulation. Apart from that, the recent (1990-2010 CE) spatial patterns of sea ice look realistic.
<img src="https://nextcloud.cism.ucl.ac.be/s/J3PDSTYgooEyqsq/preview">
<img src="https://nextcloud.cism.ucl.ac.be/s/KzkbnAB62rE6KwS/preview">
### AMOC
The absolute values of the AMOC are realistic. Its relative changes do not correlate with the ice volume in the northern hemisphere.
<img src="https://nextcloud.cism.ucl.ac.be/s/5weZMxH43zCkNSA/preview">
<img src="https://nextcloud.cism.ucl.ac.be/s/eSFzSfsYsDmDs6s/preview">
### Mixed layer depth over the Labrador Sea
<img src="https://nextcloud.cism.ucl.ac.be/s/NP2AbHrCnL7b2bA/preview">
#### Large box over the Labrador Sea (50-60°N, 65-45°W)
<img src="https://nextcloud.cism.ucl.ac.be/s/otXcE3Ycmdax8WG/preview">
#### Single grid point over the Labrador Sea (57°N, 54°W)
<img src="https://nextcloud.cism.ucl.ac.be/s/SCyZ7522FqFpaiw/preview">
For comparison: https://www.sciencedirect.com/science/article/pii/S146350031730152X
## Caveats
- Despite the fact that the simulation starts from conditions at the equilibrium in 1850 CE (after 100 years of control run), there is a strong drop in temperature and increase in sea ice over the first 40 years of simulation. No idea why, but this is an important issue.
- The simulation is too cold.
francois klein
This post gathers some information and diagnostics about the historical run (1850-2010 CE) performed with EC-EARTH3.3.
General information
The simulation was performed using EC-EARTH 3.3 (develop-fmasson branch). It starts in 1850 CE from a ~100 year-long control run, and is driven through time by CMIP6 forcing until 2010 CE.
Diagnostics
Global and hemispheric near-surface air temperatures
The simulated temperatures (HIS4 on the figures below) are too cold during the whole period, by about 1.5°C at the global scale compared to the CMIP6 EC-EARTH simulation. The recent warming is not strong enough.
Ice
Consistently with the cold bias in temperature, there is too much ice both in the North and South Poles, during the whole period 1850-2010 CE.
This ice overestimation holds during the last 20 years of the simulation. Apart from that, the recent (1990-2010 CE) spatial patterns of sea ice look realistic.
AMOC
The absolute values of the AMOC are realistic. Its relative changes do not correlate with the ice volume in the northern hemisphere.
Mixed layer depth over the Labrador Sea
Large box over the Labrador Sea (50-60°N, 65-45°W)
Single grid point over the Labrador Sea (57°N, 54°W)
For comparison: https://www.sciencedirect.com/science/article/pii/S146350031730152X
Caveats