Far future climate (2060–2100) of the northern Adriatic air–sea heat transfers associated with extreme bora events

Abstract

The northernmost part of the Mediterranean Sea, the northern Adriatic shelf, is a complex area where the intensity of dense water formation and the consequent Adriatic-Ionian thermohaline circulation are shaped by a combination of extreme wintertime bora winds and substantial freshwater loads. To better understand the impact of global warming on extreme bora dynamics and the associated sea surface cooling, this study applies the Adriatic Sea and Coast (AdriSC) kilometer-scale modelling suite to the far future climate (2060–2100) period. Under both Representative Concentration Pathway (RCP) 4.5 and RCP 8.5 greenhouse emission scenarios, the AdriSC simulations are carried out via the combination of a statistical approach—consisting of an ensemble of 3-day simulations for 22 extreme bora events, and a pseudo-global warning (PGW) methodology—imposing a climatological change to the forcing used to produce the evaluation (present climate) runs. Despite a noteworthy decrease in intensity of the bora winds (by up to 3 m/s), the latent heat losses are simulated to increase (by up to 150 W/m²) due to the reduction in relative humidity in the northern Adriatic (by up to 3%). Consequently, the sea surface cooling associated with severe bora events and preconditioning the dense shelf water formation in the northern Adriatic is projected to not significantly change compared to present climate. Although these results need to be further confirmed, this study thus provides a new view on the future of processes driven by sea surface cooling, such as the dense shelf water formation or the Adriatic-Ionian thermohaline circulation, that were projected to decrease in the future climate by regional climate models an order of magnitude coarser than the AdriSC simulations.

Appendix

Physical quantities used for the description of the bora dynamics:

$\theta$: Potential temperature (K).

$r$: Mixing ratio (kg/kg).

$g = 9.81$: Gravitational acceleration (m/s²).

$\alpha$: Specific volume (m³/kg).

${{\varvec{\Omega}}}$: Angular velocity vector of the earth's rotation (rad/s).

${\mathbf{u}}$: Three-dimensional wind velocity vector (m/s).

$U$: Horizontal wind speed (m/s).

$\theta {}_{va} = \theta \left( {1 + 0.61r} \right)$: Virtual potential temperature (K).

$N = \left( {\frac{g}{{\theta {}_{va}}}\frac{{\partial \theta {}_{va}}}{\partial z}} \right)^{\frac{1}{2}}$: Brunt-Väisälä frequency (1/s or Hz).

$PV = \alpha \left( {2{{\varvec{\Omega}}} + \nabla \times {\mathbf{u}}} \right).\nabla \theta$: Potential vorticity (PVU = 10^–6 m²K/s/kg).

$F_{r} = \frac{U}{hN}$: Froude number, with h the height of the mountain.

Physical quantities used for the calculation of the surface heat fluxes:

$U_{a}$: Horizontal wind speed at 2 m (m/s).

$T_{a}$: Air temperature at 2 m (°C).

$T_{s}$: Sea surface temperature (°C).

$r_{h}$: Relative humidity at 2 m (%).

$\rho_{a}$: Density of moist air at 2 m (kg/m³).

$P_{a}$: Mean sea level pressure (hPa).

$e_{sat} \left( T \right)$: Saturation vapor pressure (hPa).

$L\left( T \right) = 2501000 - 2370T$: Latent heat of vaporization (J/kg).

$C_{H} ,C_{E} = 0.00115$: Turbulent transfer coefficients.

$C_{p} = 1004.67$: Specific heat capacity (J/K/kg).

$q_{a} \approx \frac{{0.62197\left( {0.01 \, r_{h} \, e_{sat} \left( {T_{a} } \right)} \right)}}{{p_{a} }}$: Air saturation specific humidity at 2 m (kg/kg).

$q_{s} \approx \frac{{0.62197\left( {0.98 \, e_{sat} \left( {T_{s} } \right)} \right)}}{{p_{a} }}$: Sea surface saturation specific humidity (kg/kg).

$Q_{H} = \rho_{a} C{}_{H}C_{p} U_{a} \left( {T_{a} - T_{s} } \right)$: Sensible heat flux (W/m²).

$Q_{E} = \rho_{a} C{}_{E}U_{a} L\left( {T_{s} } \right)\left( {q_{a} - q_{s} } \right)$: Latent heat flux (W/m²).