Our terms pretty specifically has language around the extraction of geographic data from the API. It’s an area where you can’t really go.
“You will not attempt or encourage others to: … use or access the Twitter API to aggregate, cache (except as part of a Tweet), or store place and other geographic location information contained in Twitter Content.”
So anything you do with geo-based data has to still be directly tied to tweets and respect the users wishes with that content.
You also cannot redistribute geo-based data or any Twitter content in your research. You can distribute IDs pointing to tweets or IDs or screen names pointing to users, but the onus is on the next researcher to obtain that data from Twitter on their own accord.