The Stewart River watershed in the Traditional Territory of the First Nation of Na-Cho Nyäk Dun (FNNND) in Yukon, Canada, has experienced extensive surface disturbance from anthropogenic land-use and wildfires. These disturbances can impact chemical and physical water quality parameters that influence aquatic ecosystems and watershed health. Of particular concern are high levels of suspended sediment and turbidity, due to their potential impacts to fish populations and benthic communities. To support FNNND in land-use and conservation planning, this research explores the relationship between disturbance and water quality in the Stewart River watershed. The objective is to develop spatially-explicit empirical models relating observed water quality to mapped surface disturbance, while accounting for watershed characteristics and spatial distribution of disturbance relative to the stream network. These models provide insight into how watershed composition and configuration influence water quality, potentially amplifying or mitigating disturbance effects. We collected water quality data from 40 sites during summer 2024, measuring suspended sediment, turbidity, and other parameters. Surface disturbances, including mining, transportation networks, and wildfire, were compiled from existing spatial datasets and refined using spectral indices (NDVI and NBR). We used multiple regression with multimodel inference to relate water quality to disturbance metrics while simultaneously fitting optimized distance-weighting functions and assessing predictor support and uncertainty. Mining and wildfire emerged as strong predictors of TSS and turbidity. Ensemble-averaged distance-decay parameters reflected each disturbance type's spatial distribution and influence, with mining weighted most heavily near streams and wildfire influence extending across the entire catchment.