Combined Adaptation to Land Cover Change and Global Warming for the Optimization of Cascade Reservoir Operation in the Upper Sekampung Watershed

Abstract

Optimizing the operation of a cascade reservoir system requires a single adaptation-ready series of dependable discharge, yet land cover change and global warming are typically assessed separately and are rarely combined into one hydrological input ready for optimization. This study integrates two previously derived adaptation drivers for the Batutegi-Way Sekampung cascade system in the Upper Sekampung Watershed, Lampung Province, Indonesia: (i) the monthly runoff change factor due to land cover (2014-2024, SCS Curve Number method), and (ii) the delta-change-based percentage shift in dependable discharge due to temperature under four incremental temperature-rise scenarios (+0.5 to +2.0 degrees C), both applied multiplicatively to the same measurement-based historical Q80/Q50/Q20 baseline. Because two independent measures of warming actually observed in this watershed - a bi-period comparison (+0.425 degrees C) and a projected trend change to the present (+0.520 degrees C) - both identify the +0.5 degree C scenario as the one most representative of current conditions, the combined “2024 Land Cover + Temperature +0.5 degrees C” scenario is recommended as the primary adaptation-ready dependable discharge input for short-term cascade reservoir operation optimization under a non-linear programming (NLP) framework, while the combined +1.0 to +2.0 degree C scenarios are retained as medium- to long-term sensitivity cases. Under the recommended combined scenario, the mean annual Q50 is 19.93 m3/s at Batutegi and 23.63 m3/s at Way Sekampung, each within about 0.3% of the value due to land cover alone, indicating that the land cover change signal currently dominates over the temperature signal in this watershed, even though both drivers act in the same qualitative direction at Way Sekampung and in the same direction at Batutegi, where the temperature effect only partially offsets the land-cover-driven increase. The resulting monthly Q80/Q50/Q20 series are presented as a ready-to-use input table for NLP-based cascade reservoir operation optimization in the next stage of this research.