Sampling Strategies for Accurate Hazard Mapping of Noise and Other Hazards Using Short-Duration Measurements.

Abstract

Hazard mapping is an effective way to depict spatial variability in hazard intensity obtained with direct-reading instruments on a facility floor plan. However, the extent to which temporal variability affects map accuracy is unknown, and guidance on sampling strategies to minimize map bias is lacking. In this study, we evaluated the accuracy of hazard maps produced for simulated sources and sampling strategies in a hypothetical facility. Hazard maps were produced from sampled data at high, mid, and low spatial resolution and with and without replicates and compared to a reference time-weighted average hazard map using several map comparison metrics. In agreement with 'real-world' mapping datasets, the simulation showed that increasing the number of replicates improved the overall comparability of the hazard map produced from the sampled data with the time-weighted average hazard map more efficiently than increasing the sampling spatial resolution. However, if accurately capturing peak exposures near sources is of interest, increasing the spatial resolution of the measurements, particularly near sources, is needed. From these results, we formulated guidelines to use the preliminary assessment of the temporal variability of large and intermittent sources to inform the spatial resolution and need for replicate measures to minimize the bias in hazard maps.