You may try to do the following assuming you can store data coming from CSV on a UNICODE varchar data type like fiedl1 on below example:
DECLARE @Data TABLE (Field1 NVARCHAR(100))
INSERT @Data VALUES ('ABC')
INSERT @Data VALUES (N'Value with 化ける unicode chars in')
SELECT * FROM @Data WHERE Field1 <> CAST(Field1 AS VARCHAR(100))
The above T-SQL code should return all rows with Japanese rows, then you can decide what to do with them.
About how to skip special characters or escape characters coming from CSV files please read this thread for a solution.