Stay cool or get hot? An applied primer for using temperature in forensic entomological case work

Abstract

The reliable reconstruction of the temperature conditions at a crime scene is still a great challenge in forensic-entomological case work. Despite many published standards and guidelines for reconstructing temperature, and studies analysing the influence of various factors on the accuracy on such reconstructions, there are astonishingly many cases in the literature in which the temperature at the place of discovery is not reconstructed at all, i.e. the most common method is using the data of the nearest meteorological weather station without any correlation with on-site data. This study summarizes the state of the art in temperature reconstruction from an entomological point of view and compares the application of generalized additive models (GAMs) and linear regression on the basis of hypothetical death scenarios with various post mortem intervals (PMI) and measurement periods for the correlation between crime scene and weather station. We show that GAMs i.e. analysing the potential delay effect of temperature within a day, are the tools of choice because they give better, i.e. more accurate estimations than linear regression in 95,6% of all analysed cases regardless of the PMI, body discovery site and correlation period. Nevertheless, each case and crime scene is unique and therefore each entomological expertise should discuss the possible strengths and weaknesses of its temperature reconstruction. Even if temperature is not or cannot be reconstructed for various reasons, a comparison of on-site data with those of a meteorological weather station is the minimum forensic experts should do.

Keywords: Correlation period; GAM; Linear regression; Post mortem interval; Scene of death; Temperature model.