Comparison of Extreme Rainfall Characteristics between CHIRPS and Observational Data in North Sumatra in Relation to IOD and ENSO Events

Abstract

This research assesses the accuracy of CHIRPS satellite-based rainfall data and investigates the spatial distribution of extreme precipitation indices in North Sumatra under the influence of ENSO and IOD phases. Validation against 61 rain gauge stations revealed that the Linear Mean (LM) correction method outperformed other approaches, yielding the lowest RMSE and MAE, along with the highest correlation and R2 values.Therefore, the LM-corrected CHIRPS data were used for subsequent spatial analysis.The findings indicate that El Nino phases are associated with suppressed rainfall intensity and prolonged dry spells (Rx1day < 100 mm, CDD > 60 days), especially across eastern and southern regions. Conversely, La Nina phases enhance extreme rainfall events (Rx5day > 150 mm, R95p > 200 mm) and increase wet spell durations (CWD > 100 days), particularly in the western highlands. Similarly, Negative IOD phases amplify rainfall extremes in central and southern zones, whereas Positive IOD phases intensify dry conditions in northeastern areas. overall, the results underscore the critical role of large-scale climate variability in modulating rainfall extremes. Integrating validated satellite data with ENSO and IOD signals improves understanding of hydrometeorological patterns and supports more effective disaster risk reduction and water resource planning in Sumatra's climate-sensitive regions.