From Fragments to Footprints: Harmonizing Southeast Asian Climate Assessment & Dataset (SACA&D)for the Next-Generation Regional Grid Data

Accurate characterization of climate variability and hydrometeorological risk in the Southeast Asia relies
on robust observational baselines. The Southeast Asian Climate Assessment & Dataset (SACA&D) serves
as the primary regional repository, providing not only an extensive archive of daily station records but also
the derived climate extreme indices essential for monitoring climate change impacts. However, as is
common in the long-term stewardship of large-scale climate archives, integrating evolving data streams-
historical records alongside centralized daily observation systems-inherently introduces metadata
complexities.
Within the SACA&D database, the ingestion of these various data streams has yielded over 10,703 station
metadata entries covering the period 1980-2025. These entries largely consist of fragmented time series,
coordinate misalignments, and reused station identifiers corresponding to a core network of
approximately 200 active physical stations in Indonesia. Spatial interpolation of such unconsolidated
records generates potentially spurious trends and spatial density, compromising the reliability of
downstream gridded products.
To maximize the climatological utility of these observations, this study implements a data stewardship and
harmonization framework to prepare the in-situ baseline for the next-generation gridded dataset (SA-
OBSv3e). Following the protocols outlined in the European Climate Assessment & Dataset (ECA&D)
Algorithm Theoretical Basis Document (ATBD), extensive data rescue, deduplication, and quality control
efforts were applied. By resolving overlapping station IDs, removing invalid date entries, and merging
fragmented records belonging to identical physical locations, we reconstructed continuous, long-term
precipitation time series.
This study extended the archive’s temporal coverage to December 2025, yielding over 153 million quality-
controlled daily precipitation values across the network. At the upcoming conference, we will present the
fully harmonized SACA&D dataset. We will demonstrate that resolving these inherent metadata
anomalies and consolidating the fragmented network is a mandatory prerequisite for developing the
upcoming high-resolution SA-OBS v3e dataset.