diff --git a/java_based_implementation/tests/test_data_types.py b/java_based_implementation/tests/test_data_types.py
index 8f09fae..4919a2d 100644
--- a/java_based_implementation/tests/test_data_types.py
+++ b/java_based_implementation/tests/test_data_types.py
@@ -25,7 +25,7 @@
 import unittest
 
 from java_based_implementation.api_impl import Catalog
-from java_based_implementation.util import setup_utils
+from java_based_implementation.util import setup_utils, java_utils
 from paimon_python_api.table import Schema
 
 
@@ -80,6 +80,19 @@ def test_bool(self):
         expected_types = ['BOOLEAN']
         self._test_impl(pa_schema, expected_types)
 
+    def test_null(self):
+        pa_schema = pa.schema([('_null', pa.null())])
+        expected_types = ['STRING']
+        self._test_impl(pa_schema, expected_types)
+
+    def test_unsupported_type(self):
+        pa_schema = pa.schema([('_array', pa.list_(pa.int32()))])
+        schema = Schema(pa_schema)
+        with self.assertRaises(ValueError) as e:
+            java_utils.to_paimon_schema(schema)
+        self.assertEqual(
+            str(e.exception), 'Found unsupported data type list<item: int32> for field _array.')
+
     def _test_impl(self, pa_schema, expected_types):
         scheme = Schema(pa_schema)
         letters = string.ascii_letters
diff --git a/java_based_implementation/util/java_utils.py b/java_based_implementation/util/java_utils.py
index adad49d..052a537 100644
--- a/java_based_implementation/util/java_utils.py
+++ b/java_based_implementation/util/java_utils.py
@@ -48,7 +48,7 @@ def to_paimon_schema(schema: Schema):
 
     for field in schema.pa_schema:
         column_name = field.name
-        column_type = _to_j_type(field.type)
+        column_type = _to_j_type(column_name, field.type)
         j_schema_builder.column(column_name, column_type)
     return j_schema_builder.build()
 
@@ -61,7 +61,7 @@ def check_batch_write(j_table):
         raise TypeError("Doesn't support writing dynamic bucket or cross partition table.")
 
 
-def _to_j_type(pa_type):
+def _to_j_type(name, pa_type):
     jvm = get_gateway().jvm
     # int
     if pa.types.is_int8(pa_type):
@@ -83,5 +83,11 @@ def _to_j_type(pa_type):
     # bool
     elif pa.types.is_boolean(pa_type):
         return jvm.DataTypes.BOOLEAN()
+    elif pa.types.is_null(pa_type):
+        print(f"WARN: The type of column '{name}' is null, "
+              "and it will be converted to string type by default. "
+              "Please check if the original type is string. "
+              f"If not, please manually specify the type of '{name}'.")
+        return jvm.DataTypes.STRING()
     else:
-        raise ValueError('Unsupported pyarrow DataType:' + pa_type)
+        raise ValueError(f'Found unsupported data type {str(pa_type)} for field {name}.')