You may need longer period of data covering different climatic conditions for continuous modelling. If you have only short period of data, belongs to the wet season, you can go for event modelling. Model parameters, validated by event modelling, might not be applicable for dry periods. Anyway, I would like to follow this discussion. Don’t be discouraged. Begin with whatever the data available with you and see the results and get instructions of experts having experience. No one can find perfect data for this kind of work. If you like, you can go through my paper on 'Validation of HEC_HMS for the Kelani River'. I also had similar problems.
You can use SMA (soil moisture accounting method) for even a short duration like two months even. Unlike event-based loss methods (eg: SCS method) SMA method considers antecedent moisture conditions which considers for. So the duration of two months of wet years in different years fine.
Thank you Miyuru Bandara Gunathilake , Ismail Elhassnaoui , Prema Hettiarachchi all of you for your kind suggestion. I will try to do it and see the results