Short- and Long-Run Relationships Between Observed and Model 1 Output Rainfall Data in Majalengka Regency

This study examines the short-run and long-run relationships between observed monthly rainfall and CMIP6 climate model projections in Majalengka Regency, Indonesia. Monthly rainfall observations from the BMKG Kertajati Meteorological Station are analyzed using the Autoregressive Distributed Lag (ARDL) framework, which enables simultaneous assessment of short-term dynamics and long-term equilibrium relationships. Stationarity and cointegration are evaluated using the Augmented Dickey–Fuller test and ARDL bounds testing, respectively, while model performance is assessed through out-of-sample validation for the period 2015–2017 under three CMIP6 emission scenarios: SSP1-2.6, SSP2-4.5, and SSP5-8.5. The results indicate a positive and statistically significant short-run relationship between observed rainfall and CMIP6 projections across all scenarios, suggesting that climate models capture local scale monthly rainfall variability reasonably well. In contrast, the long-run relationship is weak and negative, highlighting limitations in representing long-term local rainfall dynamics. Model performance is highest under the low-emission SSP1-2.6 scenario and decreases under higher-emission scenarios. These findings suggest that CMIP6 outputs are more reliable for short-term rainfall analysis than for long-term local assessments without bias correction or downscaling.