<p>Climate-induced changes in snow cover, river flow, and freshwater ecosystems will greatly affect the indigenous groups in the Alaska and Yukon River Basin. To support policy-making on climate adaptation and mitigation for these underrepresented groups, an ongoing interdisciplinary effort is being made to combine Indigenous Knowledge with western science (https://www.colorado.edu/research/arctic-rivers/).</p><p>A foundational component of this project is a high fidelity representation of the aforementioned land surface processes. To this end, we aim to obtain a set of reliable high-resolution parameters for the Community Territory System Model (CTSM) for the continental scale domain of Alaska and the entire Yukon River Basin, which will be used in climate change simulations. CTSM is a complex, physically based state-of-the-science land surface model that includes complex vegetation and canopy representation, a multi-layer snow model, as well as hydrology and frozen soil physics necessary for the representation of streamflow and permafrost. Two modifications to the default CTSM configuration were made. First, we used CTSM that is implemented with hillslope hydrology to better capture the fine-scale hydrologic spatial heterogeneity in complex terrain. Second, we updated the input soil textures and organic carbon in CTSM using the high-resolution SoilGrid dataset.</p><p>In this study, we performed a multi-objective optimization on snow and streamflow metrics using an adaptive surrogate-based modeling optimization (ASMO). ASMO permits optimization of complex land-surface models over large domains through the use of surrogate models to minimize the computational cost of running the full model for every parameter combination. We ran CTSM at a spatial resolution of 1/24<sup>th</sup> degree and a temporal resolution of one hour using the ERA5 reanalysis data as the meteorological forcings. The ERA5 reanalysis data were bias-corrected to account for the orographic effects. We will discuss the ASMO-CTSM coupling workflow, performance characteristics of the optimization (e.g., computational cost, iterations), and comparisons of the default configuration and optimized model performance.</p>
