There are times when a cube already has the data loaded in it, but you would like to reprocess aggregations.
One example I can think of is if you have re-designed aggregations using usage based optimization.
How do you just re-calc the aggs?
Instead of choosing Process Full, choose Process Index. Process Index only reprocesses the indexes and aggregates.
For a more detailed article on this topic, there is an excellent article by T. K. Anand.