Mathematical models based on physical laws and incorporating data on ecosystem behavior are especially powerful at predicting future climate scenarios. The models allow temporal progression through mathematical integration. However, the models can only be validated through reproduction of past climate change scenarios—hence the importance of the combined tasks in GRACCIE of climate modeling and paleoclimatic reconstructions. The interaction between marine air masses, land surface (heat and moisture exchanges), and atmospheric pollution (aerosols affecting nucleation), within the sea-breeze circulations can work as a threshold-dominated system. Thus, in summer meteorological processes at different scales, such as sea-breezes with a diurnal cycle, become self-organized at the meso-α scale (to ≈ 2200 km). During this development vertical recirculations are generated that extend to the whole Western Mediterranean Basin. These processes and their feedbacks can thus propagate perturbations up and down the climatic ladder (local-regional perturbations ↔ global climate changes). Modest changes in land use or air pollution emissions along the coastal areas may change the properties of the air masses and modify the summer storm regime inland.

The critical threshold is the height of the cloud condensation level of the air mass within the breeze with respect to the height of the coastal mountain ranges (Millan et al., 2005). If exceeded, the probability of storm development and maturing greatly decreases. Loss of storms results in drier soil inland, increased surface heating, and even higher cloud condensation levels, all of which reinforce a feedback loop towards desertification in these areas. On the other hand, the loss of the moisture accumulated over the sea (non-precipitated in the local storms) alters the evaporation-precipitation balance within the Western Mediterranean Basin.

The modelling work will be aimed to determine the interactions of the atmosphere-land-ocean system in the Mediterranean Basin and the effects of its feedbacks on the meteorology and the four-dimensional distribution of the air pollutants (e.g. Palau et al., 2005; Perez-Landa et al., 2006a,b). Those feedbacks are key aspects in the changes of the precipitation regime in the Mediterranean. Specific attention will be devoted to characterize the critical thresholds along the mountains surrounding the Mediterranean, in order to develop land-use management procedures aimed at restoring (and/or maintaining) the summer storms, and minimising erosion from increased torrential rains. Mesoscale meteorological modelling using remote sensing and ancillary data will allow exploring the interactions and feedback processes relating human activities, including air pollutants, and climate change at the local and regional scales.