The length scale and the spatio-temporal variation of turbulence intensity has relevant implications on phytoplankton dynamics. Thus, it is important to estimate the relevant parameters that characterize the turbulence in the water column, such as epsilon (kinetic energy dissipation rates). One of the main objectives in this work is the characterization of the physical dynamics at scales relevant to the biology. Here we show different approaches to estimate the epsilon in the Alfacs Bay (Ebre Delta), where recurrent harmful algal bloom events occur. First, we applied the solid boundary layer theory wind velocities obtained by a nearby meteorological station. Secondly, the gradient temperature microstructure method, based on the Batchelor spectrum adjustment was applied on temperature data obtained by a Self-Contained Autonomous MicroProfiler (SCAMP). These two approaches have methodological restrictions, i.e. isotropic turbulent or the process applied to do the Batchelor spectrum fitting. A new method to characterize the turbulence is proposed. The velocity fields measured by a deployed high resolution 2 MHz acoustic Doppler current profiler were processed using the Reynolds decomposition to obtain an empirical parameter which provides us the information about the turbulent kinetic energy in the water column.