- Faculty: Ananth Kalyanaraman (WSU), Kirti Rajagopalan (WSU), Alan Fern (OSU)
- Graduate students: Krishu Thappa (WSU), Bhupinderjeet Singh (WSU), Supriya Salvalkar (WSU)
- Snow Water-Equivalent (SWE) – the amount of water available if the snowpack is melted – is a key decision variable used by water management agencies to make irrigation, flood control, power generation, and drought management decisions. SWE values vary spatiotemporally – affected by weather, topography and other environmental factors.
- While daily SWE can be measured by Snow Telemetry (SNOTEL) stations with requisite instrumentation, such stations are spatially sparse requiring interpolation techniques to create spatiotemporally complete data. The objective of this project is to develop a machine learning model to predict SWE in a reliable manner, with the ultimate goal being to generate a spatially complete gridded product for SWE that can be used by various downstream studies relating to hydrology and geospatial seasonal water availability.
Our hypothesis is that the attention mechanism used in transformer architectures has a unique ability to capture and exploit correlations that may exist across locations or the temporal spectrum (or both). It also has the ability to expose spatial relationships between different locations, highlighting common principles and factors that potentially dictate snow-water equivalent observations.