Experimental data of Ca-induced activated sludge flocculation under different conditions of temperature and dissolved oxygen are investigated in order to model the influence of changing physical and chemical factors. However, current kernel structures for collision frequency and efficiency are unable to describe activated sludge flocculation data. Therefore, an earlier developed methodology based on an inverse problem is applied, yielding empirical models, to find out how flocculation is affected by these different environmental conditions. This contribution shows the useful application of inverse problems to improve the understanding of complex aggregation mechanisms.