Integration of the Global Water and Lake Sectors within the ISIMIP framework through scaling of streamflow inputs to lakes

Climate change impacts both lakes and their surrounding catchments, leading to altered discharge and nutrient loading patterns from catchments to lakes, as well as modified thermal stratification and mixing dynamics within lakes. These alterations affect biogeochemical processes and water quality in lakes. Coupled catchment-lake modeling provides both a holistic evaluation of the effects of climate change on lakes and a framework for explicitly assessing the importance of how catchments effect lakes. The Inter-Sectoral Impact Model Intercomparison Project (ISIMIP) provides a framework for projecting the impacts of climate change across multiple sectors (e.g. water, lakes, energy, health) of the Earth System consistently, enabling integrated cross-sectoral assessments. However, climate impacts on lake dynamics are modeled in ISIMIP without consideration of the links between lakes and the surrounding catchments. This is a significant limitation, as it restricts assessments to only the direct impacts of climate change on lakes, overlooking the critical interactions between lakes and their catchment areas. In this study, we establish the first dynamic connection between the Global Water and Lake Sectors in ISIMIP, achieved by scaling the gridded modeled outputs of water fluxes from the Global Water Sector to the catchments of the representative lakes of the Lake Sector. The streamflow to the representative lake of each grid cell, as defined by the ISIMIP Global Lake Sector, was calculated based on runoff proportional to the catchment area of each representative lake. If the lake surface area was larger than the grid cell area, water from upstream grid cells was included as the corresponding proportion of river discharge. The methodology was applied to 71 lakes of widely different size across Sweden, and the estimated streamflow was validated against both the streamflow outputs from the hydrological model HYPE and observed data. Our procedure showed good performance in terms of long-term streamflow mean and seasonality, with a yearly average Kling-Gupta efficiency, KGE, of 0.54±0.23 and a monthly average KGE of 0.59±0.18 when compared to HYPE outputs, and with yearly and monthly average KGEs of 0.73±0.16 and 0.50±0.19, respectively, when compared to observations. This estimated streamflow, representing water flow into lakes, will provide a valuable dataset for the scientific community within the ISIMIP Lake Sector supporting hydrological and water quality modeling efforts aimed at understanding the impacts of climate change on lakes.