A Simple Algorithm for Identifying Periods of Snow Accumulation on a Radiometer

Downwelling solar, Q_si, and longwave, Q_li, irradiances at the earth’s surface are the primary energy inputs for many hydrologic processes, and uncertainties in measurements of these two terms confound evaluations of estimated irradiances and negatively impact hydrologic modeling. Observations of Q_si and Q_li in cold environments are subject to conditions that create additional uncertainties not encountered in other climates, specifically the accumulation of snow on uplooking radiometers. To address this issue, we present an automated method for estimating these periods of snow accumulation. Our method is based on forest interception of snow and uses common meteorological observations. In this algorithm, snow accumulation must exceed a threshold to obscure the sensor and is only removed through scouring by wind or melting. The algorithm is evaluated at two sites representing different mountain climates: 1) Snoqualmie Pass, Washington (maritime) and 2) the Senator Beck Basin Study Area, Colorado (continental). The algorithm agrees well with time-lapse camera observations at the Washington site and with multiple measurements at the Colorado site, with 70 to 80% of observed snow accumulation events correctly identified. We suggest using the method for quality controlling irradiance observations in snow-dominated climates where regular, daily maintenance is not possible. Code for running the quality control algorithm can be found at https://github.com/klapo/moq.