Search

Improving intensity simulation and forecast of tropical cyclones has always been a challenge, although in recent years the track forecasts have been remarkably improved. In this study, we explore the sensitivity of typhoon simulation to three physical processes using a fully coupled atmosphere-ocean-wave model. Two storms, a strong and a weak one, have been chosen. The effects of wave breaking induced sea spray, ocean vertical mixing associated with nonbreaking surface waves, and sea surface cooling due to intense rainfall are assessed by means of a set of numerical experiments. The results show and confirm that sea spray leads to an increase of typhoon intensity by enhancing the air-sea heat flux, while nonbreaking wave-induced vertical mixing and rainfall lead to a decrease. Each process can be relevant, depending on wind and wave conditions. These can vary dramatically when typhoons interact with not sufficiently well-defined coastal areas, typically an archipelago. Compared with the control runs, when all the three physical processes are considered, the (absolute) difference between the modeled sea level pressure and best track data is reduced from 26.05 to 0.70 hPa for typhoon Haiyan, and from -9.42 to -8.67 hPa for typhoon Jebi. We have found a steady overestimate of the dimensions of the typhoons. We have verified an extreme sensitivity to the initial conditions, especially when small differences in the typhoon track may imply different relevance of the physical processes, like the ones we have considered, governing the evolution of the storm.

The wave climatology of the Red Sea is described based on a 30-year hindcast generated using WAVEWATCH III configured on a 5-km resolution grid and forced by Red Sea reanalysis surface winds from the advanced Weather Research and Forecasting model. The wave simulations have been validated using buoy and altimeter data. The four main wind systems in the Red Sea characterize the corresponding wave climatology. The dominant ones are the two opposite wave systems with different genesis, propagating along the axis of the basin. The highest waves are generated at the centre of the Red Sea as a consequence of the strong seasonal winds blowing from the Tokar Gap on the African side. There is a general long-term trend toward lowering the values of the significant wave height over the whole basin, with a decreasing rate depending on the genesis of the individual systems.

Within the progressive improvement in wave modeling we focus on the attenuation of swell waves by rain. Till now ignored, the process is shown to be relevant, especially for the correct estimate of swell. Following the practical impossibility of devoted field experiments, we exploit the global model results over a period of more than four years to extract the tiny signal associated to the attenuation by rain. A direct comparison of the ratio model by altimeter significant wave heights versus the encountered rain amount hints to a marked related dependence. A proper quantification of the related physical effect requires a multiple step procedure that we describe in detail. We check the reliability of the results, and we provide the related source function ready for implementation in operational wave models.

The wind climatology of the Red Sea is described based on a 30-year high-resolution regional reanalysis generated using the Advanced Weather Research Forecasting model. The model was reinitialized on a daily basis with ERA-Interim global data and regional observations were assimilated using a cyclic three-dimensional variational approach. The reanalysis products were validated against buoy and scatterometers data. We describe the wind climatology and identify four major systems that determine the wind patterns in the Red Sea. Each system has a well-defined origin, and consequently different characteristics along the year. After analysing the relevant features of the basin in terms of their climatology, we investigate possible long-term trends in each system. It is found that there is a definite tendency towards lowering the strength of the wind speed, but at a different rate for different systems and periods of the year.

Using the new high-resolution operational model of ECMWF, we revisit the storm during which the Draupner freak wave of 1 January 1995 was recorded. The modeling system gives a realistic evolution of the storm highlighting the crucial role played by the southward propagating polar low in creating the extreme wave conditions present at the time the freak wave was recorded. We also discuss the predictability of the meteorological event. The hindcast wave spectra allow a new analysis of the probability of occurrence of the Draupner wave that we analyze not only in time at a specific position, but also in space. This leads us to discuss how exceptional the so-called freak waves really are. For a given sea state, as characterized by the significant wave height, they are namely part of the reality of the ocean, the key point being the probability of encountering them. In this respect, the often considered record at a specific location can be misleading because the probability of detecting a freak wave must be considered both in space and time.

The performance of state-of-the-art meteorological and oceanographic numerical systems in predicting the sea state in the Adriatic Sea during intense storms is assessed. Two major storms that affected Venice are discussed. The first storm occurred on 4 November 1966, when Venice suffered its most dramatic flood event. The damage and loss of life caused by the storm and the associated flood were extremely high also because the event was poorly forecast. The 1966 event is reanalysed using state-of-the-art meteorological and oceanographic numerical systems to investigate whether the poor forecast quality was due to a lack of data or of suitable numerical modelling. The second severe storm took place on 22 December 1979, when Venice experienced the second-worst 'acqua-alta' conditions in recorded history. Results show that with the present numerical systems both storms and associated wave and surge conditions could have been forecast several days in advance. Potential implications for the prediction of more frequent less intense storms are discussed, and a suitably enhanced system based on a global meteorological model and a limited area one is outlined.

Not available

We have explored the sensitivity of the wave modelled results in the Mediterranean Sea to the grid and integration time step resolutions. The results show that in areas with a complicated coastal shape an improved grid resolution provides often substantial local differences, mostly associate to the effect of, but not necessarily close to, the coasts. Some smaller differences are associated also to the accuracy of the numerical procedure. The use of different time steps did not provide appreciable differences. We have also explored how the quality of the wind fields in the Mediterranean Sea is going to improve with the shift of the resolution of the ECMWF meteorological model from T511 (40 km) to T799 (25 km).

A number of meteorological situations are studied using the operational models at the European Centre for Medium-Range Weather forecasts, and a series of short-term forecast experiments is made for four different resolutions of meteorological model. A comparison with the measurements and between the corresponding results allow an evaluation of the relative benefits of increased. The focus is on the surface wind speeds and the wave heights. Good quality results are found over the oceans, especially in areas with strong spatial gradients. Consistent with the already good quality of the present the present operational results,the improvememets with icreased resolution are small,but significant.In enclosed basins, as in the case of the Mediterranean Sea, it is found that both the surface wind speed and wave height are strongly underestimated. the improvements with increased resolution are greater than over the oceans, but still not sufficient to give agreement with measured data.