Dylan Singh Sanghera
06th July – 31st July
In collaboration with:
Thousands of people go missing every year, however, there is little evidence based research on how forensic teams should use thermal imaging technologies in order to increase success in locating a corpse, whether in water or on land. The implications for this project are such that police departments will be provided with more information regarding the optimum time and day to use thermal imaging for detection of corpses to be most successful, depending on temperature variables and where the Post Mortem Interval (PMI) can be estimated (PMI referring to the number of days after death). To detect corpses it is recommended that the police use ground based thermal imaging, as it can detect temperature differences effectively, even after death. This paper’s aims are 1) To investigate the best time in the day to conduct a thermal imaging survey, 2) To find out the differences between Terrestrial and Aquatic temperature development, and 3) To find the optimum length of time that a cadaver is detectable after death using ground-based thermal imaging.
Data collected from previous 2017 terrestrial and 2019 aquatic projects at Keele University were used in this research project. In each project, a pig cadaver, simulating a human corpse, was left in an aquatic/ terrestrial environment of a controlled test site and imaged each day. FLUKE thermal imaging devices were used to take samples of cadavers in their aquatic and terrestrial environments. The images were downloaded onto a computer to be analysed using FLUKE SmartView software. Images were analysed to determine a] pig cadaver temperatures, b] background control site temperatures and c] relative temperature. The results were then recorded onto a spreadsheet table from which graphs were generated to help make conclusions on the decay of a pig cadaver. Other variables were also looked at such as weather data from Keele University’s weather station.
Results showed that both pigs were detectable for the PMI days observed (0-18), through the stages of decomposition. The optimum PMI day to carry out thermal imaging was found to be 10-14 for aquatic and 9 -13 for terrestrial. The optimum time of day for surveying to be conducted, based on terrestrial thermal anomaly data from PMI 0-18, was found to be in the evening. The weather variable (rainfall, direct sunlight hours, and dry bulb temperature) did affect optimum PMI, however more research needs to be done to assess their effect.
Last year, there were over 130,000 cases of missing individuals in the UK . In 2014, 9373 people in the federal state of Bavaria (Germany) were reported missing – but only 158 of them could be found dead . Searching for missing people is becoming an increasingly important task for the police. For the majority of families, locating the missing person is of utmost importance for closure, and in order for criminal investigations to proceed.
After death, body temperature cools down to ambient temperature within approximately 24 h, depending on climatic conditions. In this short window of time infrared radiation may be a tool for searching missed persons or estimating the post-mortem interval, but after 24–48 h such detection and use of body heat is no longer possible. However, a dead body usually will be colonized rapidly by necrophagous insects such as blow flies, often resulting in the formation of so-called maggot masses, i.e. thousands of densely aggregated fly larvae feeding on the body tissue. It is a well-known fact that in such body areas and larval aggregations, temperatures can exceed ambient air temperatures by up to 25 °C, depending e.g. on the season and the body mass of the carcass. 
Detection relies upon the differentiation of objects based on emitted infrared energy and depending on the thermal imaging technology used. It is possible to differentiate between two objects with as little as a 0.1ºC temperature difference. Living individuals emit significantly greater heat than their surroundings because of the exothermic process of metabolism from stored chemical energy (food), allowing detection by thermal imaging. After death, however, human body temperature declines rapidly over time, eventually thermally equilibrating with ambient temperature . It is very important for police to know the PMI (Post-mortem Interval), and the cause of death in order for them to investigate the events that led up to death, and potential locations. There is little known knowledge on the optimum time and day to locate a corpse using thermal imaging, the differences of this on land and in water, and the different variables that may affect detection of the corpse. This is what will be researched and investigated in this project.
1.2 The application of current technologies:
Thermal Imaging (TI) detects temperature by recognizing and capturing different levels of infrared light. This technology uses a sensor to convert the radiation into a visible light picture. In order to do so, the camera must first be fitted with a lens that allows Infrared radiation (IR) frequencies to pass through, focusing them onto a special sensor array which can detect and read them. The sensor array is constructed as a grid of pixels, each of which reacts to the infrared wavelengths hitting it by converting them into an electronic signal. Those signals are then sent to a processor within the main body of the camera, which converts them using algorithms into a colour map of different temperature values. This map can then be downloaded onto computer software such as Fluke SmartView. Not only does this picture help us identify objects in total darkness, but the sensor information can be used to measure temperature differences.
- Thermal imaging cannot be used in applications where the materials absorb long wavelength radiation.
- Thermal imaging cannot look through common materials such as glass.
The applications for thermal and infrared imaging are broad and range from aerial surveillance and perimeter security, astronomy, military imaging and night vision, to automotive, law enforcement (locating corpses), medical and laboratory imaging, machine vision, inspection and other industrial tasks, unmanned systems (e.g. drones) and more . In forensics thermal imaging is commonly used to detect fleeing fugitives and heat by-products of illegal cannabis operations. TI is also utilised by the military to act as night vision to detect heat signatures . In the case of the police’s duty in locating corpses, the application of this technology is detecting the body’s temperature changes after death.
1.3 Similar reports/ Literature Reviews:
Helicopter thermal imaging for detecting insect infested cadavers
Article author/ year: Jens Amendt, Sandra Rodner, Claus-Peter Schuch, Heinz Sprenger, Lars Weidlich, Frank Reckel /2017
Amendt et al. is the most similar report relating to this research project. This report focused on ‘Helicopter thermal imaging for detecting insect infested cadavers. The report examines the thermal history of two pig cadavers in May and June 2014 in Germany. Thermal imaging from a helicopter using the FLIR system was performed at three different altitudes up to 1500 ft. during seven day-flights and one night-flight.  In the study, temperatures were collected from internal bacterial digestion, and external larval masses.
Conclusion: the study proved that even 3 weeks post-mortem aerial thermal imaging was possible, however the limiting factor of this investigation was that it was conducted in a warm season. The timeline was nearly the same for both carcasses, but the loss of tissue during the stage of active decay was much more in the first pig. The period of larvae presence was longer by 3 days on pig 2. Future studies should analyse the temperature profiles of decomposition during winter so that it can be proven that thermal imaging has the same efficiency in the colder periods of the year.
Preliminary investigation of Aircraft Mounted Thermal Imaging to locate decomposing remains via the heat produced by larval aggregations
Article author/year: Michael J. Lee, Sasha C. Voss, Daniel Franklin, Ian R. Dadour/ 2018
Key points and findings:
- This report investigated the potential of aircraft thermal imaging (ATMI) in locating decomposing remains (clothed and unclothed) through detection of heat, generated by larval aggregations of carrion feeding insects.
- Investigation: Two trials were carried out, both using four pigs. They were exposed to insect activity in autumn (trial 1) and winter (trial 2) on the Swan Coastal Plain, Western Australia.
- Activity of blow fly larval aggregations and corresponding heat generation was greatest during the active decay stage, where remains were strongly detectable by AMTI at distances up to one kilometre away.
- Implications and Conclusion: As the timing of larval aggregation activity, and therefore visibility, varied significantly in different seasons. Climatic conditions must be considered when assessing the window of opportunity for AMTI as a viable search technique. Further studies are needed to improve the tool’s predictive accuracy under varying circumstances.
- AMTI was determined to be most effective between 9 pm and 5 am (pre-dawn hours). It is a good tool for the successful detection of larval aggregations under Western Australian environmental conditions.
- ATMI can be used successfully where remains are strongly visible at a distance greater than 1000 m, either clothed or unclothed. It was also possible to detect carcasses into the advanced decay stage using the AMTI as a detection tool.
Figure 3: Photographic images for carcass decomposition on indicated days during Trial 1 and Trial 2 and corresponding thermal image (right) indicating detectability to AMTI 
The Forensic Development of Aquatic Targets
Article author/year: Rayhaan Perager/ 2019
This unpublished report was written by a previous Nuffield student in which they attempt to find out the optimum diurnal time and decomposition stage at which the cadaver is most detectable via thermal imaging, and explore the factors affecting thermal visibility on a corpse. The report concludes that overall temperature of the environment and the temperature difference are directly linked. It also found out the variables that affected the amount of infrared radiation emitted from the body post-mortem, as well as how they affect the temperature difference between the body and the background.
Figure 4: Thermographs of 2019 Project Aquatic Targets and the edited images from Fluxe SmartView Version 4.3 Software
Evaluation of Previous Research
Most of the previous research analyses why temperatures change after death, what the best methods and equipment tools (e.g. ATMI, Infrared Thermal Imaging) are to detect and locate corpses after death, stages of larval decomposition, and the processes that take place during decomposition. The results from these investigations have been largely variable, depending on a number of factors such as weather conditions (climate), time of year, temperature, soil and surrounding environment (geographic), diurnal time, and PMI (Post-Mortem Interval). The lack of research into the efficiency of TI in the detection of cadavers and their remains leads to a certain amount of uncertainty, how long after death the application of TI can be recommended. This was highlighted by a case in Germany where a 10-year old boy was kidnapped and killed. The investigators were concerned about the operation of a TI system and its appropriate timing when it took almost 2 weeks before the first (eventually unsuccessful) search .
There has been little research conducted on the best PMI for ground based thermal imaging to be the most successful, the optimum time of day to use thermal imaging technologies for detection, and the differences in optimum PMI/ Diurnal time in Aquatic and Terrestrial conditions. The previous unpublished Nuffield report analyses some of these factors, however not in great detail, and with a lack of Terrestrial and Aquatic data.
1.5 Predicted results and theory:
One factor which is discussed to be responsible for the rise in temperature after death is the metabolic heat generated by maggot masses, since the feeding larvae release proteolytic enzymes for tissue breakdown and external digestion. Bacterial metabolism could contribute considerably to the heating of carcasses as well . The prediction was that due to the body cooling down, PMI days 1-2 would generate little heat and cool down to a similar temperature to its environment. This is known as the ‘fresh’ stage of decomposition. The human body has a higher specific heat capacity (~3,500 J/Kg ºC) , when compared to typical wet (~1480 J/Kg ºC) and dry (~800 J/Kg ºC) soil , so it will take longer for a human body to cool down and warm up.
The principal component of decomposition, putrefaction, begins during the Bloated Stage (stage 2). ‘The combined processes of putrefaction and the metabolic activities of the maggots begin to cause an increase in the internal temperatures of the body. These temperatures can be signiﬁcantly above ambient temperature (36.1-37.2°C) and the body becomes a distinct habitat’ . Unless affected by external factors, it is predicted that the temperature will rise above the background temperature, and therefore be detectable by thermal imaging.
The temperature difference will be highest during the active decay stage (stage 3), where the body will reach a peak temperature, and temperature difference will be the greatest, dependent on external factors. This is due to the voracious feeding of maggots and the purging of decomposition fluids into the surrounding environment . Masses, often composed of thousands of larvae, generate significant levels of heat as a result of feeding and fast metabolisms, particularly during this stage. It is believed that the optimum time for the police to use thermal imaging technology will be during this stage, PMI 10 – 20.
Factors such as when the image was taken, where the body is located, and exposure to sunlight will affect thermal visibility. Weather conditions will be a significant variable in determining how long each stage of decompositions lasts. PMI may be shorter, or longer as expected as a result. This needs to be taken into consideration when analysing data by considering rainfall, sun hours, and air temperature data.
Warmer temperatures increase visibility and accelerate the process of decomposition whereas colder temperatures will slow the process down and, if cold enough, stop it altogether. The temperature will equally affect insect succession, which will ultimately affect how quickly the body is broken down. Therefore the prediction was made that the best time of day for thermal imaging will be mid-day, 12am-4pm, where temperatures are generally higher, and visibility is high, especially considering the season the investigation took place in (autumn).
2. Aims and Objective:
- To investigate the best time of day to conduct a thermal imaging survey
- To find out the differences between terrestrial and aquatic development
- To find the optimum length of time that a cadaver is detectable after death using ground-based thermal imaging
- Analyse thermal imaging data taken for aquatic pigs to get quantitative temperatures for pig and background
- Compare and analyse results with terrestrial data from previous data gathered at Keele University
- Plot graphs based on pig and weather data, analysing, and comparing results for patterns and key findings.
3. Methods and Materials:
3.1 Using Pigs as Human substitutes:
The use of pig cadavers as human analogues is well established in forensic science studies as they have similar chemical compositions, body sizes, tissue-to-body fat ratios, gut fauna, skin ⁄ hair types, and emissivity value to humans . Using a human analogue would raise ethical issues.
3.2 Taking images of the pig:
Due to COVID 19 restrictions, all images and data used were from previous years. Last year (2019), domestic pig carcasses were emplaced in 2000 litre tanks on the 1st July 2019 by Nuffield Research students. This was also the date of death. Images were taken at 7 post-mortems, and were also taken daily over a period of 2 weeks, excluding weekends.
Figure 5: A) Image and B) schematic of survey site, survey point denotes location where all thermal image data were acquired (2.5 m from cadaver). 
The terrestrial images were taken in 2017. The test site was in Keele University, Staffordshire, situated in a restricted area of grassed semi-rural ground surrounded by deciduous woodland and hedges. Images were taken from the 10th of April to the 9th June 2017, when warmer temperatures would facilitate insect cadaver activity. Surveying continued until active decomposition was complete. The pig cadaver had a metal cage placed over it to avoid scavenging activities.
3.3 Thermal imaging equipment and software:
To take the thermal images of the pig cadaver, a Fluke Ti100 Thermal Imaging Camera was used in both the 2019 and 2017 projects. This camera has an accuracy of 0.1 °C. This allows quantitative comparisons of repeat thermal image data to be calculated.
In 2017 the following data collection description was made – ‘Immediately after the pig cadaver had been placed on the cleared ground surface on day 1, an initial thermal imaging survey was undertaken, with readings taken at 2 h – 3 h intervals over a 24 h period, in order to determine when optimal survey time(s) for the subsequent thermal imaging surveys should be undertaken.’ 
The images taken in 2019 had the pig and the water surroundings in them, to compare the heat difference. These images were uploaded onto a free software program called Fluke SMARTVIEW. This software allows the user to analyse images to find temperatures in different image regions. On each image the polygon marker was used to draw two polygons, one around the pig and the other around the perimeter of the background. The software then generated the maximum, average and minimum temperatures. These numbers were then inputted into a spreadsheet in Microsoft Excel (see figure below). Quick reports were carried out on each image, to find out the date and time each image was taken. This was essential in order to organise spreadsheet data in date order, and to help determine PMI. This background and pig data were then plotted against post-mortem interval and ADD (Accumulated Degree Days) – a way to measure the passage of time and temperature simultaneously, used to determine whether the total heat requirement for a stage of development has been met.
Figure 6: Fluke SmartView Software (Version 4.3) being used to analyse the thermal images of the cadavers
Figure 7: Polygons drawn around the pig, in order to generate average, minimum, and maximum temperature values
3.4 Analysing the data:
The PMI was matched up for the aquatic and terrestrial data. There was a limited amount of aquatic data provided, whereas the terrestrial data was quite extensive, as the pig had been observed for over a month. Therefore, only day/ PMI 0-18 of the terrestrial data was used, which is the same PMI range as the aquatic data. This was done to clearly compare the aquatic and terrestrial pig data. This also resulted in graphs being more specific, so it was clearer to identify patterns, trends, anomalies, and differences on a micro and macro scale.
Images for the 2019 project were only taken during the first two weeks, as this was when the first three stages of decomposition occurred. After two weeks, the pig would be in advanced decay, and therefore little heat produced. The results would show a small temperature difference and extending the project to the skeletal stage is not necessary considering this project aims and objectives.
3.5 The spreadsheet:
Using a Microsoft Excel workbook, temperature, PMI, date, time, and stage of decay were recorded in a table. Potential variables that many effects the output temperature emitted from the pig on land and in water were also recorded, for example – weather data.
The data was organised into tables, to avoid any confusion. This raw data was used to calculate negative and positive error bars for pig and background temperature, and weather temperature. Positive and negative error values specify the upper and lower limits of data (the range from all the data that was collected). Error bars were then calculated because when standard deviation error bars overlap even less, it’s a clue that the difference is probably not statistically significant. Error bars are also a good graphical representation of the variability of data indicating the error or uncertainty in a reported measurement to give a general idea of how precise a measurement is. Relative temperature (pig temperature – background temperature) was also calculated for both terrestrial and aquatic tables.
After collecting and inputting all the data into the table, calculating error bars and relative temperature, graphs were made (see results section) based on weather, relative temperature, and pig and background temperatures on the day each image was taken. This was plotted against PMI and ADD. Each graph was a scatter graph, with a polynomial trend line (Order = 4) inserted. The R squared value ((a statistical measure of how close the data are to the fitted regression line) was also added to show the goodness of fit of the trend line. A benchmark R squared value of 0.5 was set to indicate that the model/ graph explains most of the variability of the response data around its mean. An R squared value of less than 0.3 is considered weak and having a low effect size.
4.1 The Temperature Differences:
Figure 8: Aquatic TI scatter graph showing pig and background temperature values, with linear and polynomial trend lines, against PMI (X- axis)
Figure 9: Aquatic TI scatter graph showing relative temperature (pig – background), with linear and polynomial trend lines, against PMI (X- axis)
The above graphs show the pig temperature compared to the background temperature, and the temperature difference between them. Figure 8 shows the temperature initially decreasing from PMI day 7 to 9. On day 8 and 9 pig temperature drops below the background temperature. The background temperature remains the same throughout at around 20 degrees Celsius, represented above by the dashed linear line. From PMI day 10-14, the temperature increases at a steeper rate, reaching a peak of 32 degrees Celsius on day 14. At this point, the temperature difference between the background and the pig temperature is greatest. The polynomial trend line, order = 4, produced an R squared value of 0.65 in figure 8 and 0.57 in figure 9 indicating a strong data response (See Spreadsheet section). Figure 9 shows the same result, where the most significant increases in temperature occur from PMI days 10 (20.2 degrees Celsius) to 14 (33.4 degrees Celsius), a large temperature difference of 13.2 degrees Celsius on average.
Figure 10: Terrestrial TI scatter graph showing pig and background temperature values, with linear and polynomial trend lines, against PMI (X- axis)
Figure 11: Terrestrial TI scatter graph showing relative temperature (pig – background), with linear and polynomial trend lines, against PMI (X- axis)
The terrestrial graphs above show the pig temperature compared to the background temperature, and the temperature difference between them. Figure 10 shows the temperature initially decreasing at a rapid and steep rate from 3.6 degrees Celsius on PMI day 0 to a trough of -1.2 degrees Celsius on PMI day 4. From PMI days 1 – 6, the pig temperature drops below the background temperature. After PMI day 4, the temperature begins to rise, reaching a peak of 15 degrees Celsius on PMI day 10. The background temperature linear trend line shows a general increase and is much more variant compared to aquatic data.
Figure 11 shows that from day 6 to 12, temperature begins to increase, with the largest temperature change between PMI day 9 – 13. The temperature increases from 0.2 degrees Celsius to a peak temperature of 1.3 degrees Celsius, a 1.1 degree Celsius increase. The polynomial trend line, order = 4, produced an R squared value of 0.72 in figure 11 indicating a strong data response (See Spreadsheet section). After PMI day 12, temperature starts to decrease and eventually cools below the background temperature after PMI day 15.
4.2 The Weather Data:
Figure 12: Aquatic weather data combo chart, displaying rainfall, direct sunlight and relative temperature data over the PMI days observed
Figure 13: Aquatic temperature scatter graph to observe to effect dry bulb temperature has on relative temperature. Polynomial trend lines used.
Figure 14: Aquatic temperature scatter graph over total study period (PMI 0-18), with linear trend line to identify general trend
The three graphs above show weather data for the Aquatic project. The first graph shows the rainfall levels (mm) in bar format and direct sunlight (hours) in line format compared to PMI on the days observed, and relative temperature (dashed line) on each PMI day. There is some relation between the two variables analysed. Where sunlight hours were high, and rainfall low/ none, relative temperature is generally high. For example, on PMI day 14 there was no rainfall, and direct sunlight occurred for 3.5 hours. Relative temperature on this day reached a peak point. Generally, higher rainfall levels correlative with less direct sunlight hours. From day 15 to 16, rainfall increased massively from 0mm to 4.6mm, where we can also notice a sudden decrease in sunlight hours from 4.5hrs to 1.5hrs.
The second and third graphs show the possible effect of the dry bulb temperature variable on relative temperatures. Generally, there is little to no correlation. Over the study period, dry bulb temperature generally increased by a small amount (peak =18.2, trough = 13.2). The polynomial trend line, order = 4, produced an R squared value of 0.23 in figure 14 and 0.57 in figure 13 indicating a strong data response (See Spreadsheet section).
Figure 15: Terrestrial weather data combo chart, displaying rainfall, direct sunlight and relative temperature data over the PMI days observed
Figure 16: Terrestrial temperature scatter graph to observe to effect dry bulb temperature has on relative temperature. Polynomial trend lines used.
Figure 17: Terrestrial temperature scatter graph over total study period (PMI 0-18), with linear trend line to identify general trend
The three graphs above show weather data for the terrestrial project. The first graph again shows the rainfall levels (mm) in bar format and direct sunlight (hours) in line format compared to PMI on the days observed, and relative temperature (dashed line) on each PMI day. The data shows a similar trend to the aquatic graph, where direct sunlight hours are higher where rainfall is low. For example, on PMI day 7 where rainfall is 1.1mm, and direct sunlight hours is near 0hrs. There are some anomalies, such as PMI day 11 where sunlight hours are low despite no rainfall.
The second and third graphs show the possible effect of the dry bulb temperature variable on relative temperatures. Generally, over the study period temperature decreases (peak = 10.7, trough = 4.4), with significant increases only present during PMI day 9-13. The polynomial trend line, order = 4, produced an R squared value of 0.76 in figure 16 and 0.01 in figure 17 indicating a weak data response (See Spreadsheet section). Generally, there is little to no correlation between dry bulb temperature and relative temperature.
4.3 24hr Terrestrial Survey:
Figure 18: Average relative thermal anomaly pig temperatures over the pig cadaver with respect to – average background temperature of the two 24 h surveys (Day 1 and Day 30).
Figure 19: Morning relative thermal anomaly pig temperatures displayed on scatter graph with linear trend line, over the total study period observed.
Figure 20: Afternoon relative thermal anomaly pig temperatures displayed on scatter graph with linear trend line, over the total study period observed.
Figure 21: Evening relative thermal anomaly pig temperatures displayed on scatter graph with linear trend line, over the total study period observed.
The data images were quantified by 2017 project researchers, and relative temperatures were calculated. These graphs show a graphical summary of thermal imaging data quantitative analysis, showing the relative thermal difference for morning, midday, and evening surveys.
The graphs show there are different times of day determined to be optimal for surveying, which depend upon the PMI. For the period observed, evening surveys are the most optimal. Morning surveys generally had -1 ℃ to 0 ℃ anomalies, midday surveys had -1 ℃ to +1 ℃ anomalies, and evening surveys had +1 ℃ to +2 ℃ anomalies over the surface pig cadaver, when compared to background values.
Morning and afternoon surveys had lower temperatures between PMI 1-18, and showed smaller temperature increase when analysed further, meaning the success of a survey at this time is likely to be lower than an evening survey.
This project’s first aim was ‘To investigate the best time in the day to conduct a thermal imaging survey’. Data collected from the 2017 Terrestrial project was used in order to create graphs based on data collected in the morning, afternoon, and evening. The optimal time of day to survey pig cadavers within PMI days 1-18 was during the evening. This result was surprising, considering the temperature and weather data at this time. The reason for this may be due to increased maggot activity in the evening during the first 18 days, however more research will need to be done to further investigate this. Searching at the optimum time will increase the chances of detecting a missing individual in a shorter time period, and therefore provide a significant milestone in the criminal investigation.
The second aim of this project was ‘To find out the differences between terrestrial and aquatic development’. Graphs were made based on all the data collected. Scatter graphs with a trend line were chosen because there were large variations from data, which can be seen by the use of error bars on each of the graphs. This could have been due to the bad weather or/and unreliable aquatic data (see Limitations Section), or because multiple readings were taken over the same day. A trend line presents a general trend, and therefore minimised the mistake of me misinterpreting the graph data in the results section. As predicted, the aquatic project had a bigger temperature difference between background and pig due to a higher specific heat capacity of water. Soil has more variations in its temperature and is more exposed to weather, so as predicted, produced smaller temperature differences. Depending on which stage the corpse is in, the cadaver emits heat due to decomposition from maggots and internal enzyme activity. It was predicted that it would be during active decay. This was true for both the aquatic and terrestrial data, and both had the largest temperatures between PMI 10-20.
The final aim of this project was ‘To find the optimum length of time that a cadaver is detectable after death using ground-based thermal imaging’. The cadavers were detectable for both aquatic and terrestrial pigs from the PMI range analysed (PMI days 1-18). It was found that the optimal day of surveying was PMI day 10-14 for aquatic pigs and 9-13 for terrestrial pigs. This was when the pigs were in active decay, which is what was predicted in the hypothesis. As predicted, the aquatic targets also produced larger temperature differences, and higher temperatures.
Thermal imaging did detect the carcass over the active decomposition period. The best time of the day to conduct thermal imaging surveys was found to be in the evening, depending on the PMI variable. The best time of day was found to be PMI days 10-14 for aquatic cadavers, and PMI 9-13 for terrestrial cadavers. Finally, aquatic data was compared to the terrestrial data and the optimum location – on land or in water- was found to analyse the differences in PMI, and how variable factors such as weather (rain, sunlight, temperature) affect each one. New information such as optimal time of day has been found out as a result of this research. Unlike previously published papers, the data analysed was during the spring season, where weather conditions are different. This research will therefore also be useful to compare the similarities and differences compared to the other seasons.
One of the biggest limitations of this project is that the data used is secondary. Due to COVID-19 restrictions, data could not be collected in person, or be carried out using a different method. This would have been more systematic. The previous project data used also had some data missing. For example, the 2019 Aquatic Pig data did not have recorded values for different times of the day, and there were only a limited number of images available to me. Therefore, some PMI days were missing from the table. Also, the times extracted from the images looked unusual and were most likely wrong, affecting the reliability when analysing this data.
Other research limitations included using one and not multiple pig cadavers for replication purposes in the 2019 project, one naked deposition style and one study site. Replicating the study using replicate pig cadavers in different depositional styles (i.e. wrapped in carpet/plastic sheeting, clothed) will undoubtedly have different thermal signatures and should be done in future research.
6.3 Future work:
Current research has demonstrated that thermal imaging is a viable tool in searching for corpses. Further investigation should be carried out in order to test its capabilities in a range of scenarios. For example, using pig cadavers with clothes on, or pig cadavers floating on the water surface (to simulate dead bodies floating).
To further investigate the weather variable, this method could be replicated at different times of the year – summer, winter, autumn, and spring (repeated again). Different locations should also be looked into, looking at different scenarios a dead body would be located in such as woodland areas, cars, houses, lakes, and ponds (slower decomposition rate), distance offshore, and underground. This may require collaboration with the police force to share research ideas and look at forensic needs.
Research has been done on investigated the use of aircraft mounted thermal imaging on locating decomposing remains, but with current policing budget cuts it may be worthwhile to look at drone-mounted thermal imaging as a cheaper alternative for police forces, focusing on their ability to detect remains at different heights, speeds and over different types of terrain. There are also many other thermal imaging platforms that could be investigated such as satellite data, fixed-winged aircraft, and UAV drones. An example research question could be ‘what is the optimum height of UAV drone thermal survey?’. New technologies such as time lapses from ground-based cameras could also be looked into.
Thank you to the Nuffield research project organisers Dr Farzana Aslam and Dr Clare Everson for coordinating this project, and project leader Dr Jamie Pringle, who provided the project task and relevant materials and sources to complete research. Thank you for also holding Microsoft Team meetings three times a week, and for helping to answer questions/queries. Thank you Kris Wisniewski who provided the land thermal data and graphs.
- Amendt, Jens, Sandra Rodner, Claus-Peter Schuch, Heinz Sprenger, Lars Weidlich, and Frank Reckel. “Helicopter thermal imaging for detecting insect infested cadavers.” Science & Justice 57, no. 5 (2017): 366-372.
- Lee, Michael J., Sasha C. Voss, Daniel Franklin, and Ian R. Dadour. “Preliminary investigation of aircraft mounted thermal imaging to locate decomposing remains via the heat produced by larval aggregations.” Forensic science international 289 (2018): 175-185.
- Stowell, Charles A. “Thermal imaging by law enforcement 2001.” In Thermosense XXIII, vol. 4360, pp. 192-202. International Society for Optics and Photonics, 2001.
- Richards, Austin A. “Alien vision: exploring the electromagnetic spectrum with imaging technology.” SPIE, 2011.
- Faber, P., and L. Garby. “Fat content affects heat capacity: a study in mice.” Acta physiologica scandinavica 153, no. 2 (1995): 185-187.
- Bristow, Keith L., Gerard J. Kluitenberg, Chris J. Goding, and Terry S. Fitzgerald. “A small multi-needle probe for measuring soil thermal properties, water content and electrical conductivity.” Computers and electronics in agriculture 31, no. 3 (2001): 265-280.
- Tibbett, Mark, and David O. Carter, eds. Soil analysis in forensic taphonomy: chemical and biological effects of buried human remains. CRC Press, 2008.
- Stokes, Kathryn L., Shari L. Forbes, and Mark Tibbett. “Human versus animal: contrasting decomposition dynamics of mammalian analogues in experimental taphonomy.” Journal of forensic sciences 58, no. 3 (2013): 583-591.
- Pringle, Jamie, Kristopher Wisniewski, Vivienne Heaton, Georgina Handley, Murray Haynes, and Andrew Mitten. 2020. “Preliminary Investigation on Thermal Imaging Of Decomposing Pig Remains On Land”. Accessed July 15th.
- “Specific Heat Capacity – Energy And Heating – AQA – GCSE Physics (Single Science) Revision – AQA – BBC Bitesize”. 2020. BBC Bitesize. https://www.bbc.co.uk/bitesize/guides/z2gjtv4/revision/5.
- Heaton, Vivienne, Abigail Lagden, Colin Moffatt, and Tal Simmons. “Predicting the postmortem submersion interval for human remains recovered from UK waterways.” Journal of forensic sciences 55, no. 2 (2010): 302-307.
- Heaton, Vivienne, Colin Moffatt, and Tal Simmons. “Quantifying the temperature of maggot masses and its relationship to decomposition.” Journal of forensic sciences 59, no. 3 (2014): 676-682.
- Goff, M. Lee. “Early post-mortem changes and stages of decomposition in exposed cadavers.” Experimental and Applied Acarology 49, no. 1-2 (2009): 21-36.
- Edelman, Gerda J., Richelle JM Hoveling, Martin Roos, Ton G. van Leeuwen, and Maurice CG Aalders. “Infrared imaging of the crime scene: possibilities and pitfalls.” Journal of forensic sciences 58, no. 5 (2013): 1156-1162.
- Evers, Rykker, and Peter Masters. “The application of low-altitude near-infrared aerial photography for detecting clandestine burials using a UAV and low-cost unmodified digital camera.” Forensic science international 289 (2018): 408-418.
- Megyesi, Mary S., Stephen P. Nawrocki, and Neal H. Haskell. “Using accumulated degree-days to estimate the postmortem interval from decomposed human remains.” Journal of Forensic Science 50, no. 3 (2005): 1-9.
About the Author
Dylan Sanghera is an A-level Student, currently studying Chemistry, Geography, and Economics at Queen Mary’s Grammar School. After Sixth Form, he is passionate about pursuing a degree in Environmental Science and Geography at university so he can help tackle the ever present environment challenges we face on a global scale.