Probabilistic prediction of snow loads on sloped roofs

Snow loads on sloped roofs are influenced by roof slope, roof surface, snow properties, and meteorological conditions. A statistical model of the snow accumulation and sliding processes for slippery roofs is developed using field data for the winters of 1983-84 and 1984-85 from six prototype shed roofs with varying slopes. The predictive model uses temperature as the primary indicator of sliding and incorporates sliding both at the roof surface and within the roof snowpack. Roof snow loads at four different geographical locations are simulated using the model and 25 years of meteorological data. The ground-to-roof conversion factors for sloped roofs Cs, as proposed by the American National Standards Institute (ANSI) are compared with those predicted by the simulation. The results indicate that Cs is affected by climatological differences, and the values of Cs as given by ANSI are condiderably higher than the simulated coefficients.