Reliability-based design snow loads: I. Site-specific probability models for ground snow loads

David J. DeBock, Abbie B. Liel, James R. Harris, Bruce R. Ellingwood, & Jeannette M. Torrents

Published November 1, 2016, SCEC Contribution #7138

This paper describes a new method for fitting probability distributions for modeling annual maximum ground snow loads for use in structural design. These probability models are intended for use in reliability assessments to determine reliability-targeted ground snow loads, as described in the companion paper. The proposed method emphasizes the upper tail of the distributions, because the upper tail is most critical for the reliability assessment and for determination of design loads. A combination of site-specific and region-of-influence approaches results in annual maximum snow load distributions whose magnitudes are consistent with historical data observed at the site of interest, but with upper tail shapes that are informed by historical snow records at a number of similar sites. Clusters of sites with similar snow accumulation patterns are utilized to improve the definition of the snow load distribution at sites in the cluster.

Citation
DeBock, D. J., Liel, A. B., Harris, J. R., Ellingwood, B. R., & Torrents, J. M. (2016). Reliability-based design snow loads: I. Site-specific probability models for ground snow loads. Journal of Structural Engineering, 143(7).