Future projections of extreme precipitation over Indonesia's new capital under climate change scenario using CORDEX-SEA regional climate models

Abstract

As Indonesia's new capital (hereafter called Nusantara City or NC) aspires to become a leading sustainable city, it must be developed with resilience to climate change. This study investigates the future seasonal extreme precipitation conditions in NC using a set of climate projections from the CORDEX-SEA Regional Climate Models (RCMs). The analysis integrates four Global Climate Models (GCMs) downscaled using two RCMs: RegCM4 and RCA4. The model results were analyzed for the periods 1970–2005 (historical), 2011–2040 (early century), 2041–2070 (mid-century), and 2071–2099 (late century) under both the low-emission (RCP 2.6) and high-emission (RCP 8.5) scenarios. Model evaluation using Integrated Multi-satellitE Retrievals for GPM (IMERG) Version 07B data shows that RegCM4-based downscaling more accurately captures mean rainfall, seasonal patterns, and extreme precipitation indices in NC than RCA4. The ensemble average of three RegCM4 runs (HadGEM2-ES, MPI-ESM-MR, and NorESM1) indicates significant increases in extreme precipitation indices under RCP 8.5 from the mid to late century. By the late century, indices such as Maximum 5-Day Precipitation (RX5days), Consecutive Dry Days (CDD), and Number of Very Heavy Precipitation Days (R50mm) are projected to increase by 5.88 %, 41.12 %, and 29.30 %, respectively, while Total Precipitation (PRPTOT), Consecutive Wet Days (CWD), and Number of Heavy Precipitation Days (R20mm) decrease by −15.69 %, −21.93 %, and − 13.78 %. Rainfall during the dry season (June–September) is projected to decline by approximately 17 % under RCP 2.6 and 25 % under RCP 8.5. Historically, El Niño has triggered droughts during the dry season in Kalimantan, and under future warming, severe droughts resembling those driven by El Niño may occur even in its absence. Furthermore, the divergent projections resulting from the various models employed here suggest the critical role of model selection in hydrological impact analysis, as different models can produce strongly divergent estimates of future extreme rainfall.