Sea Surface Temperature and Air-Sea Interaction in the Mediterranean Region: Direct measurements, satellite estimates and model based assessments

Salvatore Marullo1,2, Alcide di Sarra3, Vincenzo Artale4, Dr. Marco Bellacicco, PhD5, Jaime Pitarch6, Damiano Sferlazzo7, Chunxue Yang8, Giandomenico Pace3 and Rosalia Santoleri8, (1)ENEA, Centro Ricerca Frascati, Frascati, Italy, (2)CNR, Institute of Marine Sciences, Rome, Italy, (3)Agenzia nazionale per le nuove tecnologie, l'energia e lo sviluppo economico sostenibile (ENEA), Rome, Italy, (4)Enea Cr Casaccia, S. Maria Di Galeria, Italy, (5)ENEA National Agency for New Technologies, Energy and Sustainable Economic Development, Frascati, Italy, (6)Royal Netherlands Institute for Sea Research and Utrecht University, Department of Coastal Systems, Texel, Netherlands, (7)ENEA, Lampedusa, Italy, (8)CNR Institute for Marine Science, Rome, Italy
Abstract:
The air-sea interactions largely affect regional and global climate, as well as weather evolution at both local and global scales. These processes are essentially governed by heat, momentum, freshwater and gas exchange at the air-sea interface. Thanks to the data acquired at the climatic station of Lampedusa, in the central Mediterranean Sea, a research effort aimed at assessing the capability of reproducing the air-sea interaction processes was started from the comparison of directly measured radiative budget components with satellite and model derived estimates. Lampedusa is an integrated atmospheric/oceanic observatory composed by two sections: a ground-based laboratory operating since 1997, and an oceanographic buoy operating since 2015, dedicated to the investigation of air-sea interactions and to ground-truth of satellite observations. The data collected at Lampedusa constitute a unique dataset to continuously evaluate the accuracy of turbulent and radiative heat fluxes derived from satellites, models, and bulk formulae, and to develop of new fluxes parameterizations.

Our analysis reveals that SEVIRI MSG4 observations overestimate the short-wave irradiance by 5 W/m2 with a RMS deviation of 22 W/m2 and the downward atmospheric longwave irradiance by 7 W/m2 with a RMS deviation of 12 W/m2.

Similarly, based on ECMWF ERA5 re-analysis, the shortwave incoming radiation shows a very small bias of 1.2 W/m2 and a RMS deviation of 32 W/m2, while the longwave irradiance is underestimated by 17 W/m2 with a an RMS deviation of 12 W/m2. The impact of these determinations was also investigated by using a 1D numerical model, the General Ocean Turbulence Model, for the evolution of the upper ocean temperature profile.

The model simulations have been compared with water temperatures recorded at 1 and 2 meters of depth. The mean bias is 0.02 °C and -0.01 °C respectively, and the RMSE is 0.5 °C when the model is forced using measured radiative fluxes at the mooring and turbulent fluxes estimated from in situ observations. When the model, instead, is forced by satellite based radiative fluxes the bias increases to 1.5 °C with a RMSE of about 1°C. These results confirm the importance of reducing uncertainties in satellite estimates of radiative fluxes including effects of atmospheric aerosols.