Reconstruction and analysis of negatively buoyant jets with interpretable machine learning

Such phenomena mostly occur during the disposal of wastewater from desalination plants, power plants, industrial factories, and cooling water discharge from liquefied natural gas (LNG) plants. Nowadays, numerous arid and semiarid coastal regions are encountering freshwater scarcity due to population growth and deficiency of portable water, hence the number of desalination plants is significantly increased. During the desalination process, using brackish or seawater, drinkable water is obtained, while the by-product is high-salinity concentrated effluents, socalled desalination brine, which is discharged back into the coastal waters using submerged outfalls. Besides elevated salt concentrations, desalination brine contains traces of chemicals, such as antiscalants, flocculants and coagulants ([43], which can lead to environmental degradation ([4])). To minimize harmful environmental effects and maximize dilution, the brine is predominantly discharged from diffusers directed upwards at an angle ([41]), producing negatively inclined buoyant jets. The design of the discharge systems and condition parameters are based on detailed experimental, mathematical, and numerical investigation in order to determine the characteristics of the jet. Experimental investigation of inclined buoyant jets was done by numerous researchers: [15], [31], [18], [45], [46], [2], [6]. Mathematical models such as VISJET ([14]) and CORJET ([16]) were implemented and compared with experimental data ([27], [32], [42]).