Basically, your requirement is for Replication, so you should look at SQLServer's Replication features in detail. Sometimes log shipping (that is, copying transaction log backups) is used to achieve similar results. This article explains the difference:

qa.sqlservercentral.com/columnists/pibison/logshippingvsreplication.asp

Other points to consider:

- You are creating databases that shadow the branch databases, so you can replicate from the branch to the central location; then you must update from the shadow database to the real central database, which is an extra step that could be avoided with real replication

- on the other hand, sometimes an intermediate step allows greater control, such as a place to validate or resolve conflicts before data hits the main database.

- if your application could datestamp all data updates then a fully custom process to regularly extract new data is another option; its more work but sometimes is more flexible, eg the central database might aggregate the imports rather than holding the full detail.