Spot size, distance and emissivity errors in field applications of infrared thermography

Abstract

Infrared thermography is increasingly emerging as an analytical approach within the thermal ecology research community, providing unique and rapid temperature information crucial to understanding how plants and animals exchange heat with their environment. What is difficult to appreciate are the numerous ways in which thermography may still yield inaccurate (i.e. deviation from the ‘correct’ value) information if certain tenets are not followed. In this paper, we examine, demonstrate and discuss these tenets with an aim to provide methodological advice to ecologists interested in employing thermography. We found that spot size and distance strongly influenced the surface temperature estimates of known, calibrated temperature sources, with similar results observed in maximum eye temperature measurements in wild birds. We also report on how the angle of incidence affects the apparent emissivity of various biological surfaces (fur, feather, skin, leaves), another source of uncertainty in thermography. The variation in temperature caused by variation in distance and uncertainty in emissivity are large enough to raise flags for field applications of thermography where accuracy is necessary but control over study subjects is limited. Since accurate emissivity and distance parameters are crucial to thermography calculations, our results should serve as a framework to assist ecologists in better experimental design with respect to the use of thermography.

Infrared thermography as a technique to measure physiological stress in birds: Body region and image angle matter

“…Recent studies have shown that the emissivity and perceived surface temperature of an object can vary according to angle of incidence in an infrared thermographic image (Playà‐Montmany & Tattersall, 2021; Winder et al, 2020). As such, changes in the relative orientation of an object during infrared thermographic imaging may conceal or distort true changes in surface temperature that are driven by biological processes (e.g., vasomotion or contraction of skeletal muscle).…”

“…Effects of acute stress exposure on surface temperature trends across time differed little from those of our original models at an α of 0.05 (Table S1) despite a reduction in sample size. Next, the yaw of individuals at the time of thermographic image capture was included at a fixed and nonlinear covariate in our models using a thin‐plate regression split with five knots (nonlinear effects reported in Playà‐Montmany & Tattersall, 2021), and both models were subsequently re‐run. In our adjusted models, significant correlations between yaw and surface temperature values were expected to indicate systematic effects of angle of incidence on mean surface temperature values of a specific facial region.…”

“…More practically, the relative efficacy of approximating ANS‐mediated stress responsiveness from measurements of eye region or bill temperature also depends upon how susceptible measurements from each anatomical region are to systematic biases. Recently, variation in the spatial orientation of individuals within infrared thermographic images has been raised as a possible source of error in surface temperature measurement (Playà‐Montmany & Tattersall, 2021; Winder et al, 2020). Specifically, changes in the angle of incidence of a focal object have been shown to alter its perceived emissivity, thus resulting in consistent under‐estimation of surface temperatures at certain angles (Playà‐Montmany & Tattersall, 2021) that may conceal or distort true changes in surface temperatures that are driven by biological functions (e.g., vasconstriction or vasodilation).…”

“…Recently, variation in the spatial orientation of individuals within infrared thermographic images has been raised as a possible source of error in surface temperature measurement (Playà‐Montmany & Tattersall, 2021; Winder et al, 2020). Specifically, changes in the angle of incidence of a focal object have been shown to alter its perceived emissivity, thus resulting in consistent under‐estimation of surface temperatures at certain angles (Playà‐Montmany & Tattersall, 2021) that may conceal or distort true changes in surface temperatures that are driven by biological functions (e.g., vasconstriction or vasodilation). Critically, the degree to which angle of incidence influences emissivity and surface temperature estimates appears to vary across biological tissues (Playà‐Montmany & Tattersall, 2021).…”

“…Specifically, changes in the angle of incidence of a focal object have been shown to alter its perceived emissivity, thus resulting in consistent under‐estimation of surface temperatures at certain angles (Playà‐Montmany & Tattersall, 2021) that may conceal or distort true changes in surface temperatures that are driven by biological functions (e.g., vasconstriction or vasodilation). Critically, the degree to which angle of incidence influences emissivity and surface temperature estimates appears to vary across biological tissues (Playà‐Montmany & Tattersall, 2021). As such, surface temperatures of some body regions may suffer from greater error due to variations in angle of incidence than others, with possible implications on the capacity to detect or quantify stress‐induced changes in body surface temperature across varying anatomical regions.…”