Assessing the accuracy of satellite-derived precipitation products against ground observations and evaluating their suitability as forecasting model inputs are increasingly critical priorities in water resource management. This study comparatively evaluates CHIRPS v3.0 and PERSIANN-CDR against General Directorate of Meteorology (GDM) ground observations over Türkiye for the period 1983–2024, using ME, RMSE, BIAS Ratio, and MAPE metrics. Three time series models ETS (Holt's Damped Trend), ARIMA, and Naive were applied to each dataset across 1983–2017 training and 2018–2024 test periods, with 2025–2034 projections generated. ETS outperformed ARIMA and Naive on GDM and PERSIANN-CDR datasets; ARIMA(4,1,3) achieved a marginal advantage on CHIRPS v3.0. An anomalous CHIRPS estimate in 2018 inflated test-period error metrics across all models, confirming that input data quality directly conditions forecast performance. CHIRPS v3.0 is recommended for national hydroclimatic monitoring, and ETS for annual precipitation forecasting in Türkiye.