.. _ifs_nemo_model_control:

Control Simulation
##################

The IFS-NEMO control simulation covers 11 years in total (1990-2000) and helps to address whether the historical simulation starts close to equilibrium.
The atmospheric component IFS used a horizontal resolution of 5km (TCo2559 grid) while the ocean and sea ice components NEMO and SI3 used a resolution of 10 km (eORCA12 grid).

The simulation was conducted in two stages in MN5, as indicated in the table.

+----------+------------------------------------+--------------+-------------+-----------------------------------------------+-----------------------------------+
| Model    | Spatial Resolution                 | Time         | Realization | Experiment ID                                 | DestinE Data Lake (bridge site)   |
+==========+====================================+==============+=============+===============================================+===================================+
| IFS-NEMO | 5km atmosphere; 10km ocean/sea-ice |   1990-2000  | 1           |  climatedt-gen2-ifs-nemo-baseline-cont-5km-r1 | MN5  data bridge                  |
+----------+------------------------------------+--------------+-------------+-----------------------------------------------+-----------------------------------+

Forcing data
------------
In this control simulations all external radiative forcings, including greenhouse gases, ozone, and aerosol concentrations were kept constant at year 1990 levels.

.. Assessment of the Control Simulation
.. ------------------------------------

.. Gregory Plot
.. ^^^^^^^^^^^^

.. The Gregory plot (:numref:`ifs-nemo_control_gregory`) summarizes the co-evolution of global mean surface air temperature (TAS) and net top-of-atmosphere (TOA) radiation over the 10-year control period. The simulation starts in year 0 near the ERA5 1990 reference value of 14.4 °C with a TOA imbalance of approximately -0.2 W m-2, indicating that the system is initially absorbing more energy than it emits. Over the following years, TAS cools down by roughly 0.5 °C, reaching approximately 13.9 °C by year 10, while the TOA flux increases steadily and crosses zero after year 2, ultimately reaching approximately 0.8 W m-2 by the end of the simulation. The negative Gregory slope of -2.02 W m-2 K-1 indicates a strong net radiative feedback that acts to restore the energy balance as the surface warms. The transition from negative to postive TOA flux, combined with the ongoing surface cooling, suggests that the coupled system has not yet reached equilibrium and is adjusting from its initialization state. The large interannual scatter in the TOA flux reflects the natural internal variability of the climate system over a short 10-year window.


.. .. figure:: ../../../../evaluation/mn5/figures/IFS-NEMO-Tco2559_Control_Gregory_IFS-NEMO.png
..     :name: ifs-nemo_control_gregory

..     Gregory plot for the IFS-NEMO control simulation. Each dot represents one simulation year, colored by year number (0–10). The x-axis shows globally averaged annual surface air temperature and the y-axis the net TOA radiative flux. The green dashed line marks the ERA5 1990 reference temperature (14.24 °C), the orange dashed line indicates the CERES 2001-2021 observed mean TOA flux (0.88 W m-2), and the black line shows the Gregory regression with a slope of -2.02 W m-2 K-1.

.. Ocean Temperature Drift
.. ^^^^^^^^^^^^^^^^^^^^^^^
.. The Hovmoeller diagram of global mean ocean temperature anomalies (:numref:`ifs-nemo_control_ocean_drift_temperature`) reveals a vertically differentiated drift structure over the 10-year control period. The uppermost layers (0–200 m) exhibit a pronounced seasonal cycle, a much stronger signal that the long-term cooling trend revealed by the Gregory plot. Between 200-500m the cooling trend is more evident, and is in stark contrast with the warming trend experienced from 500 to 1000m, likely reflecting a redistribution of heat from the upper ocean to the intermediated waters. The deep ocean (below 2000 m) remains largely stable with near-zero anomalies, indicating that the model drift has not yet propagated to abyssal depths within this 10-year window.

.. .. figure:: ../../../../evaluation/ifs_nemo_eval/hovmoller-control-ifsnemo-temperature.png
..     :name: ifs-nemo_control_ocean_drift_temperature

..     Time-depth Hovmoeller diagram of global mean ocean conservative temperature anomalies (°C) relative to the initial state in the IFS-NEMO control simulation (1990–1999), as derived from the TEOS10 equation of state. 

.. Ocean Salinity Drift
.. ^^^^^^^^^^^^^^^^^^^^^^^
.. The Hovmoeller diagram of global mean ocean salinity anomalies (:numref:`ifs-nemo_control_ocean_drift_salinity`) shows a complementary vertical structure to the temperature drift. The near-surface layer (0-200 m) displays again a strong seasonal cycle with a superimposed salinification trend. Immediately below, there is a thin layer where the ocean develops a slight freshening over time. This pattern of upper-ocean salinification and subsurface freshening is consistent with an adjustment of the haline stratification during model spin-up, potentially driven by excessive surface evaporation or insufficient freshwater input in the early years of the coupled simulation. At intermediate depths  (roughly 300–1000 m) a salinitifaction signal develops that penetrates deeper with time. As with temperature, the deep ocean (below 1000 m) remains virtually unaffected, confirming that the drift is largely confined to the upper and intermediate water masses within the 10-year integration period.

.. .. figure:: ../../../../evaluation/ifs_nemo_eval/hovmoller-control-ifsnemo-salinity.png
..     :name: ifs-nemo_control_ocean_drift_salinity

..     Time-depth Hovmoeller diagram of global mean ocean absolute salinity anomalies (g/kg) relative to the initial state in the IFS-NEMO control simulation (1990–1999), as derived from the TEOS10 equation of state.