Wars and conflicts that were resolved even decades ago have the lingering danger of unexploded ordinances. Remediation of these explosive devices is particularly hazardous and time consuming when there are landmines or improvised explosive devices (IEDs) that are buried. A common detection method employs canines, which are able to sniff out devices based on the emission of volatile organic compounds (VOCs), even at minute levels. When an explosive device is buried, VOCs must travel through the soil to the surface. This process introduces sources of vapor loss, notably sorption to the soil, limiting vapor availability. This study considers the impact of various conditions on the simulated burial of two explosive-related compounds: 2,4-dinitrotolune (DNT) a degradant and contaminant from TNT and 2,3-dimethyl-2,3-dinitrobutane (DMNB) a common taggant in C-4. Headspace solid phase microextraction (HS-SPME) was used to measure free volatiles at varying heights along a simulated burial temporally, providing quantitative evidence of the gaseous diffusion process. A driving force of vapor transport was confirmed to be evapotranspiration, as conditions with moisture and sunlight resulted in higher concentrations of analyte present. Additionally, stronger adsorption to soil indicated that environments with a higher sand content are likely to have more successful detection of buried explosives. Complementary canine olfactory trials were conducted for comparison to the laboratory study. This work points to environments where detection of a buried explosive device is more likely to be successful or limited.
Explosive detection dogs are routinely deployed for parcel screening, yet realistic sealed-parcel searches may present challenges not captured by conventional odor-recognition training. In 2025, a black box study of ANSI/ASB Standard 092 identified substantially reduced detection of explosives in tape-sealed parcels, suggesting a potential training gap in this search scenario. This follow-on study evaluated whether brief, progressive training could improve dog performance in sealed-parcel searches. Thirteen operational dog teams participated in trials conducted in Lubbock, Texas, and Miami, Florida. Teams completed a baseline assessment using tape-sealed parcels containing 50 g of C4 or TNT, or distractors, followed by an odor recognition test (ORT) using controlled explosive odor presentations. Following the correction of identified odor-recognition deficiencies, teams completed a training intervention manipulating odor availability through variable parcel venting and requiring progression from odor-source recognition to trained final response. Performance was reassessed using the same parcel configuration. All teams detected C4 during the ORT, whereas only two teams detected TNT. Following intervention, overall sealed-parcel detection sensitivity increased by 35 percentage points (p = 0.0078). C4 detection increased from 38% to 69%, while TNT detection increased from 15% to 54%, although neither individual change reached statistical significance. The results demonstrate that odor recognition does not necessarily translate to effective source localization in sealed parcels and support progressive, operationally representative parcel-search training.