Abstract
Reservoirs collectively contribute 1-2% of global anthropogenic greenhouse gas emissions, although individual emissions can vary widely. While emission models have considerably advanced our understanding of the lifetime carbon impacts of reservoirs globally and offer means to inform judicious planning, their widespread adoption is hindered by high manual processing requirements, uncertainties, and linkages to geospatial drivers that can be obscure for planners. Meanwhile, simpler Tier 1 methods fail to capture variability across individual reservoirs and can overestimate national emissions by 50% compared to model-based estimates. Here we introduce an automated and transparent framework for large scale reservoir emission assessments and planning with spatially-explicit emission models to address key limitations in current approaches. By applying our framework to strategic hydropower expansion in Myanmar, we show how emission models can support low-carbon reservoir development at large scales. Our results show that the proposed methodology can yield a hydropower strategy for Myanmar that eliminates 0.94 MtCO(2e) in emissions (1% of national total), conserves 239 km(2) of forest and arable land, and reduces the number of barriers in lower river reaches from 28 to 7.