Hive comprises of an RDBMS(Relational-Database-Management-Systems) on its Master-Node which is used to keep track of the status of the data table.
consider you create a certain table with a certain schema or meta-data, then this particular schema is stored in the RDBMS. In case, if the table is partitioned, then these partitions are stored into a database and hive is allowed to use these lists of partitions to access the data with having the need to go the original file-system to fetch the data. this data which helps the Hive to fetch data is known as Metadata.
This is why when the internal table is dropped, the data along with its schema(metadata) is also dropped.
And when an external table is dropped, only the metadata is dropped making the Hive to ignore the data present in the table.