ENH: changelog and some comments

CHANGELOG.md

@@ -70,6 +70,8 @@ New features and improvements:
 - `explore` now supports `GeoDataFrame`s with additional columns containing datetimes, uuids and 
  other non JSON serializable objects (#3261).
 - The `GeoSeries.fillna` method now supports the `limit` keyword (#3290).
+- Added support for `bbox` covering encoding in geoparquet. Can filter reading of parquet
+files based on a bounding box, and write out a bounding box column to parquet files (#3282)
 
 Backwards incompatible API changes:
 

geopandas/io/arrow.py

@@ -674,9 +674,13 @@ def _read_parquet(
 
  if_bbox_column_exists = _check_if_bbox_column_in_parquet(metadata)
 
+ # by default, bbox column is not read in, so must specify which
+ # columns are read in if it exists. Not enforced if bbox is
+ # listed in columns argument.
  if not (read_bbox_column or columns) and if_bbox_column_exists:
  columns = _get_non_bbox_columns(schema, metadata)
 
+ # if both bbox and filters kwargs are used, must splice together.
  kwargs["use_pandas_metadata"] = True
  if "filters" in kwargs:
  filters_kwarg = kwargs["filters"]