We call this kind of a model run "time-zero" model results. As much as possible you want results for year 0 (initial condition) and "time-zero" to match. This is your way of ensuring the conceptual model is working at your site initially. This comparison is especially important when working with high-vertical-resolution elevation data. An elevation analysis also assists in examining the conceptual model validity for your site. There will always be some discrepancy due to input data errors and simplifications within the conceptual model.
See
this thread and
this thread for more information about time-zero differences.
Good luck -- Jonathan