Caution! All data are not created equal: The hazards of using National Weather Service data for calculating accumulated degree days

Abstract

An increasing number of anthropological decomposition studies are utilizing accumulated degree days (ADD) to quantify and estimate the post-mortem interval (PMI) at given decompositional stages, or the number of ADD required for certain events, such as tooth exfoliation, to occur. This study addresses the utility of retroactively applying temperature data from the closest National Weather Service (NWS) station to these calculations as prescribed in the past. Hourly temperature readings were collected for 154 days at a research site in Farmington, AR between June 30 and December 25, 2008. These were converted to average daily temperatures by calculating the mean of the 24 hourly values, following the NWS reporting procedure. These data were compared to comparable data from the Owl Creek and Drake Field NWS stations, the two closest to the research site, located 5.7 and 9.9km away, respectively. Paired samples t-tests between the research site and each of the NWS stations show significant differences between the average daily temperature data collected at the research station, and both Owl Creek (2.0°C, p<0.001) and Drake Field (0.6°C, p<0.001). When applied to a simulated recovery effort, the further NWS station also proved to represent the better model for the recovery site. Using a published equation for estimating post-mortem interval using ADD and total body decomposition scores (Megyesi et al., 2005 [1]), the Drake Field data produced estimates of PMI more closely mirroring those of the research site than did Owl Creek. This demonstrates that instead of automatically choosing the nearest NWS station, care must be taken when choosing an NWS station for retroactively gathering temperature data for application of PMI estimation techniques using accumulated degree days to ensure the station adequately reflects temperature conditions at the recovery site.