Extreme values of the maximum depth of soil freezing are of interest for engineering design specifications. For instance, building codes must consider the maximum depth of frost penetration to assure that footings and utilities are buried at the appropriate depths. If these specifications are too lax, freezing conditions are likely to result in structural damage during the design lifetime of the structure. Alternatively, codes that are too stringent inflate building costs unnecessarily due to increases in labor and material costs. Unfortunately, the only direct practical analysis of maximum soil freezing depths in the U.S. is based on unofficial, undocumented and antiquated (1899-1938) measurements (USDA 1941). More recently, Crandell et al. (1994) present a map of 100-year return period air freezing indices which can be used to derive empirical frost depth values. However, these values neglect the effects of a changing winter snow cover.
This paucity of measured frost depth data has led us to develop a one-dimensional heat flow model capable of estimating frost depths using only meteorological variables measured at cooperative network weather stations. Given that approximately 900 cooperative network stations are in the Northeast, model-derived frost depths can be developed for a relatively dense network of sites across this region. Using these modeled frost depth estimates, an extreme-value climatology for the maximum depth of soil freezing is produced. This atlas presents extreme annual maximum soil freezing depth statistics in the form of average return periods for the 12-state northeastern region of the United States.
32 pp.