Grid-Free Modeling of Large Scale Ocean Dynamics


issam lakkis, principal investigator

Samah El-Mohtar, Reseach Assistant (Masters)

Ocean dynamics on the planetary scale are well described by the shallow-water quasi-geostrophic equations. Vortex methods offer an interesting tool to simulate these dynamics, however, their use in this field was quite limited.

Here, a grid-free vortex method is applied to study the evolution of the planetary waves. Potential vorticity carried by water columns advected over the planet evolves accordingly to keep in harmony the latitude, relative vorticity and height of each water column. By making use of the Rossby-Haurwitz solution as a background wave propagation, implementation of the method is further simplified.

Rossby-Haurwitz waves. Colors show stream-function in m2/s

Forward and Inverse Lagrangian Particle Tracking in Stochastic Flow-Fields: Oil Spill in the Red Sea


issam lakkis, principal investigator

Ali Ayoub, Ali Saab, Nabil Ramlawi and Wael Hajj Ali Mechanical Engineering Undergraduate Students

Samah El-Mohtar, Reseach Assistant (Masters)

Lagrangian particle tracking technique that implements a combined Eulerian-Lagrangian approach is a tool that tracks Lagrangian particles advected using a Eulerian velocity field. This versatile modeling tool finds many applications in oceanography and has become a reliable technique that decision makers rely on in order to establish appropriate plans. However, the quest to develop accurate and efficient tracking techniques is often faced by challenges that need to be overcome. Eulerian uncertainties arising from the many velocities a particle can have at a given time step are one of those challenges. For instance, the velocity field resulting from data assimilation is an ensemble of several members generated at each assimilation cycle. This results in an exponentially increasing number of trajectories at each time step if a particle is said to follow the combinations of velocities of the previous time steps, a fact that makes unpractical tracking each individual particle.

To overcome this difficulty, binning the final positions of particles is proposed. Whenever many trajectories lead a particle to the same bin, they are regarded as a single trajectory that moves the particle to the center of that bin which will have a probability related to the number of trajectories leading that particle to it. This allows to generate probability maps that describe efficiently the dispersion of particles while managing wisely the computational cost by reducing significantly the sheer amount of growing data.

This method was applied to an oil spill problem in the Red Sea. Forward Lagrangian particle tracking was employed to simulate the propagation of the oil spill from a fixed as well as a moving source, while inverse (backward) Lagrangian particle tracking was used to specify probable sources of the spill assuming that the age of particles is provided. The model also accounts for beaching which is an important feature that allows to predict the location of the coasts with high probability of beached oil.

Dispersion of an oil spill from a fixed source in the Red Sea

PM10 Pollution Caused by the Zouk Mikael Power Plant


issam lakkis and Najat Saliba

Samer Salloum, Ph.D. Candidate

The electrical power plant in Zouk Mikael, built in 1956, was first operated in 1983.  The plant is equipped with four turbines (steam engines) and four stacks. The height and diameter of each stack are respectively 145 m and 9 m. The plant operates continuously at an average of 2.5 units (turbines).Each steam engine burns heavy fuel oil number 6 (density of 3.67 kg/gal) at the rate of 40 tons per hour.  

The model tracks the PM10 concentrations profile induced by the mass flow rate of pollutant emitted by the Zouk power plant and thus neglecting all other sources of pollutant and assuming a zero background level of PM10. 


The simulation was carried out using TAPM for the entire year of 2014. The results showed that the power plant contributed to a large degree to the pollution in the surrounding area. According to TAPM, the average concentration of PM10 for the year of 2014 that was induced by the power plant alone reached a maximum of 145 μg/m3 (average over the grid area surrounding the power plant) in the region adherent to the power plant at ground level.


The simulation also showed that the polluted zones are more likely situated to the east and south east of the power plant due to the wind direction. 

Consequently, Beirut is not much affected by the power plant and the average concentration of PM10 in Beirut due to the power plant alone is around 6 μg/m3. However it is clear that the region situated east Beirut (Baabda - Fayyadeyye - Jamhour) is polluted by the power plant and the average concentration in this region experienced an added 25 μg/m3in PM10 concentrations. This added value decreases as we head east and south east when keeping same distance from the power plant as shown in figure below.

The Zouk Mikael Power Plant

TAPM also identified 2 distinct zones that have the highest concentration of PM10. The first zone is situated to the east of the power plant at a distance that varies between 2 and 4.7 km (Zouk Mikael) in which the concentration of PM10 was around 45μg/m3. The second zone is situated to the south east of the power plant at a distance of around 4.8 km in which the concentration of PM10 was also around 45μg/m3. Hence, these two zones are the most affected regions by the power plant emission and the concentration induced by the plant is about 90% of the estimated background of 50μg/m3 based on measurements conducted in Beirut and extrapolated to this area

The pollutant transport simulation has been carried out for the year 2014 and the results are recorded using the GIS visualization tool offered by TAPM.

The video of the hourly pollutant concentration in the nearby region can be seen by pressing the following link TAPM video.

* We thank Dr. Marwan Katurji and Prof. Peyman Zawar-Reza (Center of Atmospheric Research, University of Canterbury, New Zealand) for giving access to TAPM and providing help and support.