The area that I am working on falls under an area whose data are classified, that is, we cannot get the data from the respective body easily. In that case without the actual discharge data how will I calibrate my HEC-HMS model?
You can't. From what I understand, "Calibration" in models refer to the act of tuning the parameters, so the output matches the observed values as closely as possible. If you don't have observed values, you cannot calibrate the model.
You can however compare the results of different models, like use different formula for calculation, or follow the methods used for hydrologically similar basins. And do some sensitivity analysis to see which parameters are sensitive, and be extra careful with how you obtained those.
Best is to calibrate your model with observed data.
I suggest that you begin to calibrate the model on similar watershed where discharge data are available and exchange data against model results with the data owner.
The U.S. National Weather Service (NWS) has developed techniques for a-priori estimation of parameters of its distributed river forecasting models based on soils data. See:
Yu Zhang, et al. "An enhanced and automated approach for deriving a priori SAC-SMA parameters from the soil survey geographic database"
You might search to see if such a-priori parameter estimation techniques have been developed for HEC-HMS.
NWS and other government and private, weather and streamflow forecasting organizations use coupled and uncoupled atmospheric/land surface models where the land surface models are parameterized based on similar a-priori techniques. You might try to find out if such global or hemispheric models exist covering your area of interest.