Timestamp gets modified when written in parquet

I have a Spark application that loads CSV files, converts it to a Parquet file, store Parquet file in my Data Lake storage, and then load data into a BigQuery table.

The problem is when CSV has too much old timestamp value, the conversion occurs but the timestamp column can't be shown in a BigQuery table.

When I set the config spark.sql.parquet.outputTimestampType to TIMESTAMP_MICROS, I got this error on BigQuery:

Cannot return an invalid timestamp value of -62135607600000000 microseconds relative to the Unix epoch. The range of valid timestamp values is [0001-01-1 00:00:00, 9999-12-31 23:59:59.999999]; error in writing field reference_date

When I set the config spark.sql.parquet.outputTimestampType to TIMESTAMP_MILLIS, I got this error on Airflow:

Error while reading data, error message: Invalid timestamp value -62135607600000 for field 'reference_date' of type 'INT64' (logical type 'TIMESTAMP_MILLIS'): generic::out_of_range: Invalid timestamp value: -62135607600000

• CSV file:

id,reference_date
"6829baef-bcd9-412a-a2f3-abdfed02jsd","0001-01-02 21:00:00"

• Read the CSV file (and convert reference_date to a Timestamp column):

def castDFColumn(
  df: DataFrame,
  column: String,
  dataType: DataType
): DataFrame = df.withColumn(column, df(column).cast(dataType))

...
var df = spark
  .read
  .format("csv")
  .option("header", true)
  .load("myfile.csv")

df = castDFColumn(df, "reference_date", TimestampType)

• Convert to a file Parquet file:

df
  .write
  .mode("overwrite")
  .parquet("path/to/save")

• Spark application runtime configurations:

val conf = new SparkConf().setAppName("Load CSV")
conf.set("spark.serializer", "org.apache.spark.serializer.KryoSerializer")
conf.set("spark.sql.parquet.outputTimestampType", "TIMESTAMP_MILLIS/TIMESTAMP_MICROS")
conf.set("spark.sql.session.timeZone", "UTC")

It looks like the timestamp is changed to 0000-12-31 21:00:00, or something like that, getting out of the acceptable range of INT64 timestamp.

Has anyone been through this?