Watershed hydrological models are effective tools for simulating the hydrological processes in the watershed. Recent advancements for improved hydrological model simulations were focused on discrete calibration of the model for different wet and dry periods, and recombining the simulations. The efficacy of the recombined model in representing the physical processes of the watershed across the different periods of a year received little attention. The discretization of the simulation periods largely remains fixed throughout the year and the pattern is repeated for all the simulation years and ignore the climate forcing, variability of rainfall and soil moisture conditions along time. This may induce anomalies in the water balance of the basin. This research study is mainly focused on addressing these two limitations.Two different schemes of discrete calibration was assessed in this regard. The effectiveness of the rejoined discrete calibration in describing the process dynamics was examined by calibrating the Soil and Water Assessment Tool (SWAT) model for three different watersheds from different climate zones seasonally using an auto calibration approach.However, they failed to represent the physics of the processes in the transition periods of the season thus affecting the total water balance of the watershed. It was also apparent that the model parameters were sensitive in varying degree of magnitude in different seasons. The study proposed a discretization methodology using clustering methods that help eliminate the existing limitations of discrete calibrations. Gustafson- Kessel (GK) clustering algorithm was adopted to cluster the simulation period based on the current day precipitation and change in soil moisture as well as the five day antecedent conditions of the same. Subsequently, the clusters were calibrated independently, as well as simultaneous with multiple objectives.The procedure was illustrated through two models viz. SWAT and a grid based model set up for the US watersheds Cedar Creek, Indiana and Riesel Y2, Texas respectively.The cluster based calibration resulted in improved stream flow simulations with higher NSE (0.83 in multi-objective, 0.82 in simulations recombined from independent clusters for Cedar Creek) as compared to the traditional calibration. Alternatively, a procedure to dynamically update the parameter based on the antecedent moisture conditions was developed in the current study. Relationships connecting the change in soil moisture and the selected model parameter was developed for each cluster transition and incorporated into the hydrological model structure. The proposed methodology was tested using a grid based model set up for the Riesel watersheds W1 and Y2, Texas. The grid based model parameter ESCO was chosen for this analysis.The parameter value gets updated along the simulation period when the cluster transition changes and not on a daily basis.The results of the analysis indicated that the proposed method is effective in representing the temporal process dynamics of the catchment(NSE = 0.88 for modified model; 0.84 for conventional approach in Riesel Y2).