| Description |
Salt Lake City, located at the base of the Wasatch mountain range in Utah, receives a majority of its potable water from a system of mountain creeks. Snowmelt runoff from mountain watersheds provides the city a clean and relatively inexpensive water supply, and has been a key driver in the city's growth and prosperity. There has been keen interest recently on the possible impact of the deposition of darkening matter, such as dust and black carbon (BC) on the snow, which might lead to a decrease in its `albedo' or reflective capacity. Such a decrease is expected to result in faster melting of the snow, shifting springtime streamflows to winter. This study aimed to develop a modeling framework to estimate the impact on snowmelt-driven runoff due to various BC deposition scenarios. An albedo simulation model, Snow, Ice, and Aerosol Radiation (SNICAR) model, was used to understand the evolution of albedo under different BC loadings. An Albedo-Snow Water Equivalent (A-SWE) model was developed using a machine learning technique, `Random Forests', to quantify the effect on the state of snowpack under various albedo-change scenarios. An Albedo-Snow Water Equivalent-Streamflow (A-SWE-S) model was designed using an advanced statistical modeling technique, `Generalized Additive Models (GAMs)', to extend the analysis to streamflow variations. All models were tested and validated using robust k-fold cross-validation. Albedo data were obtained from NASA's MODIS satellite platform. The key results found the snowpack to be depleted 2-3 weeks later with an albedo increase between 5-10% above current conditions, and 1-2 weeks earlier under albedo decrease of 5-10% below current conditions. Future work will involve improving the A-SWE-S model by better accounting for lagged effects, and the use of results from both models in a city-wide systems model to understand water supply reliability under combined deposition and climate change scenarios. |
