Toward resolving the model-measurement discrepancy of radon entry into houses: a study of the scale dependence of soil permeability to air

We studied natural soil at a site used to study radon entry into basement structures, in order to determine why models of radon entry into houses significantly and consistently under predict measured radon and soil-gas entry rates. We found soil air-permeability to be strongly scale dependent, increasing monotonically with length scale. Therefore, soil-gas flows occurring at scales of ∼3 - 5 m around houses can be seriously underestimated by traditional techniques that assess permeability by averaging over multiple small scale measurements (∼0.1 - 0.5 m). Scale-dependence appears to be a result of networks of relatively homogeneously distributed fast flow paths caused by roots, burrows, and weathering channels. When permeability, assessed at the appropriate scale, is used as input to the model, the model under prediction is reduced from a factor of ∼8 to a factor of ∼2. (orig.). (15 refs., 2 figs., 1 tab.)