Multi-proxy, Multi-season Streamflow Reconstruction with Mass Balance Adjustment

Abstract:

While streamflow reconstructions have revealed important insights about past hydroclimatic variability, their use in water resources management is still limited, arguably because reconstructions are often annually-resolved, not suitable for making decisions at finer time scales. So far, attempts toward sub-annual reconstructions have relied on statistical disaggregation­, with none or limited use of proxy information when disaggregating annual reconstructions into sub-annual scales. This approach does not realize the full potential of paleoclimate proxies, as different proxies have different seasonal sensitivity. We hypothesize that different proxy combinations can be used to directly reconstruct streamflow for individual seasons. To test this idea, we pool together a multi-proxy network, and develop a novel multi-season reconstruction framework. Our proxy network consists of ring width from five tree species and cellulose δ¹⁸O from two species in Southeast Asia, which we use to reconstruct inflow for Thailand’s two largest reservoirs (Bhumibol and Sirikit). The framework selects the best proxy combinations and produces skillful dry season, wet season, and annual reconstructions simultaneously. More importantly, the framework accounts for mass balance: the sum of the seasonal flows matches the annual flow closely (the differences are within 10% of the mean flow). This mass balance criterion is necessary to avoid misguiding water management decisions, such as the allocation of water rights. This work is one step closer towards operational usability of streamflow reconstruction in water resources management.