Resumen de Implementation of a high resolution regional ocean model for investigating air-sea interaction in the Mediterranean Area
Mahesh A. Shinde
The investigation of climate variability in different timescales such as daily, monthly, seasonal and inter-annual has utmost importance for managing the socio-economic processes on regional to global scale. Indeed, the variation in the climate has a crucial impact on agriculture, water, health, tourism, economy and transportation. Therefore the development of climate forecasting tools is necessary which helps to manage these sectors more efficiently. However, there are limitations on producing accurate climate forecast for more than two weeks in advance due to the chaotic nature of the climate system, especially for the region like the Mediterranean, which is characterized by high interannual variability. Due to its importance and challenging nature, a collective effort is being done to improve the skill of models and climate forecasting in the Mediterranean. The contribution of this thesis is a overall effort, which consists of developing a high resolution model application in the Mediterranean, which can provide reliable estimate of Sea Surface Temperature (SST) and Mixed Layer Depth (MLD). This approach is based on the fact that the atmospheric predictability in seasonal to interannual time scale is significantly dependent on slowly varying lower boundary conditions (e.g. Charney and Shukla 1981) i.e. Mediterranean SSTs. The spatial resolution of model is increased for taking into account the mesoscale processes in the Mediterranean. Since the first internal Rossby radius of deformation in the Mediterranean sea is of the order of 10-15 kms, the spatial resolution of an eddy resolving model should have at least a resolution one half of the Rossby radius. Based on this assumption, the spatial resolution is explored to the order of ~5 km (1/16o). The regional ocean modeling system (ROMS) adopted from the Rutgers University is used in the current study. The objective is to validate certain fields (such as SST and MLD) obtained from model simulations and study air-sea interactions. The validation is done by performing two experiments namely, climatological and interannual simulations. The model simulated results are validated with observations as well as intercompared to evaluate the skill of model. The monthly mean SST climatology is obtained from ten years of model run forced with climatological air-sea fluxes is well captured by the model configuration and follows the annual cycle. Model simulated summer SST climatology shows biases of the order of 0.8-1.0 oC with observation (MedAtlas) and 1.0-1.2 oC with other datasets (intercomparison). The vertical structure of temperature climatology is found to be well simulated by model in which upper layer shows a difference of 1.0 oC and it further decreased at intermediate layers. The simulated sea surface height and surface currents is validated with Aviso altimetry data. On the large scales the surface currents generated by model captures general structures of surface circulation. The monthly mean mixed layer depth (MLD) climatology computed from model is validated with observed monthly MLD climatology and found that the winter MLD is overestimated by model. In second experiment, model is forced with six hourly air-sea interaction fluxes from ERA-Interim and interannual simulations are obtained for the period 1998-2007. The monthly mean SST climatology obtained from above interannual simulation follows climatological annual cycle with cold biases in summer season. The weak SSTs (bias of the order of 1.0 to 1.5 oC) are observed in the summer for the period 2002-2007 in the model simulations. The monthly mean SST anomalies are well simulated by model except for the year 2006. The time evolution of monthly mean SST anomalies area averaged over different subbasins are exhibits interannual variability. The comparison with satellite derived SSTs reveals that our model is able to capture both, the seasonal and inter-annual variability, although it still has a bias of the order of 1 to 1.2 oC. The model is able to reproduce the temperature at subsurface layer having the signatures of existence of intermediate water masses. The monthly mean mixed layer climatology derived from interannual simulations is quite well reproduced by model. In the Gulf of Lions, MLD values are reached upto 1500 meter deep in winter whereas it shows 50 meter in summer season. The time evolution of monthly mean mixed layer climatology derived from the model is able to reproduce annual variability. The interannual variability of monthly mean mixed layer depth is simulated quite well by model for the year 2004-2007. The timeseries of climatological, monthly and daily mixed layer depth which is area averaged over various sub-basins follows seasonal cycle. The high resolution regional model application developed in the current study is thus able to reproduce certain fields. The surface currents and eddy kinetic energy in the model shows small scale structures and strong variability. The model is also capable to generate mesoscale eddies in the western Mediterranean although model overestimated surface fields.
