HydroMoE Data

This dataset is distributed in Apache Parquet format for efficient storage and analysis.During validation, the original Parquet file showed a reader-compatibility issue with the default pandas/pyarrow backend (error: “Repetition level histogram size mismatch”), while it remained readable with fastparquet.To ensure maximum portability and reproducibility, we provide a repaired Parquet version that was re-written using pyarrow after successful loading with fastparquet.Users are recommended to use the repaired Parquet file for cross-platform workflows and long-term archival reuse.The table includes station-level hydro-meteorological drivers and timestamps intended for model evaluation over the test period.