The detection of blood by HRD canines is very crucial to both SAR and crime scene investigations. Not only can they be used to find missing persons (SAR), but they can be used to detect blood at crime scenes where a suspected homicide or an assault occurred. These canines can detect trace amounts of blood that cannot be seen by the naked eye, as well as locate murder weapons. The difficulty with blood detection is the complexity of the odor profile and its change over time. Previous studies have shown that the volatile organic compounds (VOCs) of blood change significantly as it ages, which has been demonstrated with blood up to 336 days (about 11 months) old. Previous research has focused on a longer timescale with almost no data on the first 48 hours or the changes that may occur in this time. The focus of this present study is within the first 48 hours, as this is the most crucial time in the transition from fresh to decomposed blood.

This investigation started with data collected in the laboratory using headspace-solid phase microextraction-gas chromatography-mass spectrometry (HS-SPME-GC-MS), which was then analyzed using chemometrics and confirmed with canine trials. Blood odor profiles were analyzed from 1 hour after withdrawal to 2 weeks, with a focus on the first 48 hours, in order to determine when the earliest change is between fresh and aged blood. A combination of principal component analysis (PCA) and multivariate curve resolution using constrained alternating least squares algorithm (MCR-ALS) was applied to the data to determine how the blood odor profile changes over time and how many distinct profiles occur during the sampling timeframe. This data was then confirmed with canine trials with HRD canines trained on either only aged blood (older than 1 week), or a mixture of aged and fresh blood.

The results of the laboratory analysis indicated that there were multiple, distinct odor profiles between 1 hour and 2 weeks old. Overall, there appeared to be a fresh odor profile, an aged odor profile, and then intermediates in between these. The PCA data showed this based on clustering in the graph, which was then taken a step further with MCR-ALS. The latter helped show how these different odor profiles changed with respect to time. The differences in these odor profiles were also probed using trained HRD canines. Based on this data, there does appear to be a change in the odor profile within the first 48 hours. These results will assist both SAR workers in determining approximately when there might be a change in odor profile during their searches, and HRD canine training to make sure that they are training on all the different profiles during this aging process.