Quality-controlled and gap-filled daily rainfall dataset for Borneo, 1995–2024

This dataset provides complete daily rainfall (mm/day) for 363 quality-controlled stations in Borneo from 1 January 1995 to 31 December 2024 (10,958 days). The stations were selected from a network of 577 stations through multi-stage quality control and homogeneity screening. Missing values were filled with multiple linear regression using neighboring stations (within 50 km) and the bias-corrected gridded products MSWEP and CHIRPS, with a fixed fallback order. Where no neighboring station had data, values were filled from the gridded products alone. Of all 3,977,754 station-days, 60.7% are observed values and 39.3% are filled values. Of the filled values, 37.1% were filled using neighboring stations and 62.9% using gridded products alone. In an independent test with hidden observations, the Kling-Gupta Efficiency was 0.64 where neighboring stations were used and 0.35 where gridded products alone were used; bias was small in both cases. Every value carries a source code (0 = observed; 1–3 = filled using neighboring stations; 4–8 = filled using gridded products alone). We recommend the 310 stations at which gridded products alone filled no more than half of the days for analyses that depend on the timing of individual days, such as clustering of drought and extreme rainfall. Filled values on heavy-rainfall days (≥50 mm/day) are underestimated, so the dataset should not be treated as an unbiased substitute for direct observations. File is an Excel workbook with the daily rainfall, source codes, station metadata, homogeneity results, and method documentation.