An entropy approach for the optimization of river cross-sectional spacing

Accurate modeling of rivers geometry is a relevant task in one-dimensional hydraulic models. In this framework, the optimal spacing of river cross sections plays a key role in the description of river behavior. In this work it is presented an entropy approach that allows to evaluate the optimum number of river cross sections and their location on the river geometry. Through a MultiObjective Optimization Problem (MOOP) it is possible to select the points that give the least redundant information and the most informative scenario. The first condition is reached by minimizing the total correlation of the set of points taken into account and the second by maximizing the joint entropy of the same set. By comparing the couple of values of joint entropy and total correlation, the best set of cross sections is chosen, and so their spatial position. The proposed method is applied to a reach for the Po river between Cremona and Borgoforte, and results are compared with a methodology already presented in literature.