GLOBAL CALCULATION OF EQUILIBRIUM RESPONSE TIMES FOR CLOSED LAKES

Lakes are commonly heralded as “sentinels of climate change”— natural systems which record the interplay of biological, geological, and chemical processes that occur in their catchments. Hydrologically closed lakes, or those without drainage outlets, are subject to larger magnitude and higher frequency lake-level fluctuations compared to hydrologically open lakes, as variations in lake inflow can only be compensated by a change in in lake surface area. As a result, closed lakes are highly sensitive to their catchment-level climate regimes as their equilibrium water levels are a direct result of only: 1) the shape of the lake basin (hypsometry), 2) basin-level precipitation (influx), and 3) basin-level evaporation (outflux). The rate at which a closed lake attains a new hydrologic equilibrium is governed by an e-folding timescale, or a physical speed limit on how fast the lake can respond to a change in climate. For large, deep, closed lakes, values can be 100s of years (Lake Titicaca = ~ 522.2 yr), for shallow, wide lakes, values can be as small as ten years (Great Salt Lake = ~ 9.1 yr) or even less than five years (Lake Eyre = 3.2 yr). This means that two lakes with the same volume and climate regime can have equilibrium response times that differ by an order of magnitude, obscuring climate attribution. Determining empirical values of for the world’s hydrologically closed lakes is therefore critical for predicting changes to the water storage budget into the next century. To that end, we calculate a global dataset of equilibrium response time values for 27,288 closed lakes by harmonizing five separate global lake datasets. We find values correlate positively with the variance of water volumes for a suite of lakes from 1990 to 2020 as well as the slope of the de-trended volume hydrograph. Additionally, values roughly correlate to the timescale of the descending limb of area-normalized hydrographs, indicating their predictive utility. We further present an example of forward-modelling two lakes based on hindcast climate data as well as future CMIP6 climate projections.