Partitioned transtension: an alternative to basin inversion models