Regulation of the Alpine Lake Como under climate change. Bottom-Up vs Top-Down approach, and CMIP5 vs CMIP6

Climate change will largely change hydrological regimes in Alpine catchments, and water management thereby. Therefore, proper adaptation is crucial to reduce impacts upon water-related sectors. Here we assess potential shortfalls of lake management practice under climate change for the paradigmatic area of the Lake Como watershed, watered by the Adda river in Northern Italy. We set-up two hydrological models (Poli-Hydro, and Topkapi-ETH) to simulate inlet discharge to the lake, and an operational model of the lake, tuned to closely mimic observed operation patterns. We then compare two methods for climate projections, i.e. i) bottom-up B-U approach, i.e. a set of scenarios generated by independently re-sampling main climatic variables over chosen ranges of variability, and ii) a top-down T-D approach, i.e. a simulation-based analysis under CMIP5 and CMIP6 projections from IPCC, using multiple AR5/AR6 scenarios (RCPs/SSPs). Using objective scores, we project worsening of lake management performance under the top-down projections, with respect to the equivalent bottom-up scenarios, with CMIP6 ensemble giving worst conditions than CMIP5. T-D approach explicitly entail non-linear patterns of variability, such as seasonality/amount of precipitation, snow accumulation/melt, and changes in hydrological regime, challenging lake managers. ANOVA analysis is performed to highlight sources of uncertainty in future scenarios. Choice of a given GCM is a largest source of uncertainty until half century, while climate patterns (i.e. (RCPs/SSPs scenarios) are more relevant thereafter.