CMCC OceanLab has been developing and maintaining the Adriatic-Ionian Forecasting System (AIFS) operational since November 2014.


The AIFS numerical core (called AIREG, Adriatic-Ionian REGional model) is based on the NEMO v3.4 ocean general circulation model (Nucleus for European Modelling of the Ocean, Madec et al. 2008). It solves the three-dimensional primitive equations on an Arakawa C-grid, assuming hydrostatic ad Boussinesq approximations. The AIREG turbulent mixing scheme uses parameterization and equations from Pacanowsky and Philander (1981) formulation with a non-linear bottom friction. A bi-laplacian operator is used for horizontal diffusion and viscosity and no-slip boundary conditions are used at rigid walls.

The model domain covers the Adriatic and the Ionian Seas, extending eastward until the Peloponnesus and the Libyan coasts; it includes also the Tyrrhenian Sea and extends westward, including the Ligurian Sea, the Sardinia Sea and part of the Algerian basin. The primitive equations are discretized on a horizontal grid at 1/45° resolution using 121 vertical levels and integrated in time using a time-splitting formulation. The bottom topography has been obtained from the U.S. Navy 1/60° database DBDB1, by interpolating depth data into the model grid using a bilinear interpolation method.

AIREG is forced by momentum, water and heat fluxes interactively computed by bulk formulae, using the 6h-0.25° horizontal-resolution operational atmospheric data provided by the European Centre for Medium-Range Weather Forecast (ECMWF) (Tonani et al. 2008, Oddo et al. 2009). The atmospheric pressure effect is included as surface forcing for the model hydrodynamics, too. The evaporation is derived from the latent heat flux, while the precipitation is provided by the Climate Prediction Centre Merged Analysis of Precipitation (CMAP) data. Concerning the runoff contribution, the model considers the estimate of the inflow discharge of 75 rivers that flow into the Adriatic-Ionian basin, collected by using monthly means datasets (Raicich 1996, Pasaric 2004, Malacic and Petelin 2009, Servizio Idrografico Italiano, Global Runoff Data Centre, ARPA Veneto, ARPA Emilia Romagna, ARPA Calabria, Autorita’ di Bacino della Basilicata, Autorita’ di bacino del fiume Serchio, CNR-IRPI, Piano Tutela della Acque Sicilia). The Po runoff contribution, instead, is provided by using daily average observations from ARPA Emilia Romagna observational dataset, because of its importance as freshwater input in the Adriatic basin.

The initial and lateral boundary conditions for temperature, salinity and velocity come from the Mediterranean Forecasting System MFS (Pinardi et al. 2003, Tonani et al. 2008). The lateral boundary conditions, in particular, are taken on a daily basis, imposing the interpolation constraint and correction (Pinardi et al. 2003) on the total velocity, which ensure that the total volume transport across boundaries is preserved after the interpolation procedures. In order to compute the lateral open boundary conditions, the model uses the Flow Relaxation Scheme (Engerdhal, 1995) for temperature, salinity and velocity and the Flather’s radiation condition (Flather, 1976) for the depth-mean transport.

Concerning the forecasting production cycle, AIFS produces 9-days forecast every day. In particular, the numerical model is integrated in forecast mode using as initial condition the corresponding fields produced by a 1-day hindcast except on Wednesday: the hindcast cycle rewind, done once a week on Wednesday and covering 7-days of simulation, occurs when also MFS analyses are available in order to use optimal lateral open boundary conditions and produce a new initial condition for the new 9-days forecast cycle. AIFS produces hourly and daily means of temperature, salinity, surface currents, heat flux, water flux and shortwave radiation fields.


Stefania Ciliberti

  • Engerdahl, H. (1995) Use of the flow relaxation scheme in a three-dimensional baroclinic ocean model with realistic topography. Tellus, 47A, 365–382.
  • Flather, R. A. (1976). A tidal model of the northwest European continental shelf. Mem. Soc. R. Sci. Liege, Ser. 6,10, 141-164.
  • Madec, G. (2008). NEMO Ocean General Circulation Model Reference Manual, Internal Report, LODYC/IPSL, Paris, 2008.
  • Malacic, V., and Petelin, B. (2009). Climatic circulation in the Gulf of Trieste (northern Adriatic), J. of Geophys. Res., 114: C07002.
  • Oddo, P., Adani, M., Pinardi, N., Fratianni, C., Tonani, and M., Pettenuzzo, D., (2009). A nested Atlantic-Mediterranean Sea general circulation model for operational forecasting. Ocean Sci., 5: 461-473.
  • Pacanowski, R. C. and Philander, S. G. H. (1981). Parameterization of Vertical Mixing in Numerical Models of Tropical Oceans. J. Phys. Oceanogr., 11: 1443-1451.
  • Pasaric, M. (2004). Annual cycle of river discharge along the Adriatic coast of Croatia. Rapports et Proces-Verbaux des Reunions CIESMM 37, 132.
  • Pinardi, N., Allen, I., Demirov, E., De Mey, P., Korres, G., Lascaratos, A., Le Traon, P. Y., Maillard, C., Manzella, G., and Tziavos, C. (2003). The Mediterranean ocean forecasting system: first phase implementation (1998-2001). Ann. Geophys., 2003, 21: 3-20.
  • Raicich, F. (1996). Note on the flow rates of the Adriatic rivers. CNR, Istituto Talassografico di Trieste Tech. Rep. RF 02/94, 8 pp. Available from CNR–Istituto Talassografico di Trieste, viale Romolo Gessi 2, I-34123 Trieste, Italy.
  • Tonani, M., Pinardi, N., Dobricic, S., Pujol, I., and Fratianni, C. (2008). A high resolution free surface model of the Mediterranean Sea. Ocean Sci. Discuss., 2008, 4: 213-244.
  • Ciliberti, S. A., Pinardi, N., Coppini, G., Oddo, P., Vukicevic, T., Lecci, R., Verri, G., Kumkar, Y., Creti’,. S. (2015). A high resolution Adriatic-Ionian Sea circulation model for operational forecasting. Geophysical Research Abstracts
, Vol. 17, EGU2015-10899, 2015
 - EGU General Assembly 2015