FeatherFlame: An Arduino-Based Thermocouple Datalogging System to Record Wildland Fire Flame Temperatures in Agris

Abstract

Rangeland scientists have long relied on thermocouples for measuring temperature, especially in agris—in the field, under the extreme conditions of wildland fire. But the electronics required to sense and record thermocouple data remain expensive to both purchase and protect from exposure to heat and flames. Open-source, do-it-yourself (DIY) electronics platforms such as Arduino are increasingly popular among ecologists, and have been shown to perform as well as proprietary commercial systems when recording thermocouple data. The FeatherFlame system is a reliable, low-cost solution to sampling wildland fire, ranging from US$240–490 for 1–6 thermocouple sensors, including all fire protection equipment. The low cost and multi-sensor capacity facilitates spatial replication, which allows fire ecologists to measure and report meaningful data on rate of spread and calculate fire intensity.

Introduction

Rangeland scientists and managers have used thermocouples for decades (Fons, 1946, Vaartaja, 1949). Thermocouples, which are capable of operating at very high temperatures, facilitate the detection of temperature changes by placing wires of two different metals into contact; as the contact point that connects the wires absorbs or dissipates energy, differences in the electrical signals produced along each wire change in a manner that scales with temperature (e.g., the Seebeck effect). However, their utility in the field has historically been restricted by limitations in datalogger technology. The development of transistors improved power demands and mobility (Kenworthy, 1963). Magnetic tape storage allowed thermocouple data to be recorded at 2 s intervals (Engle et al., 1989), while electronic memory allowed a single datalogger to record data from multiple sensors simultaneously (Jacoby et al., 1992). But protecting expensive electronics from the heat and flame of wildland fire is itself costly: fire-proof boxes for Campbell Scientific dataloggers cost as much as US$500 (Butler, Jimenez, Forthofer, Shannon, Sopko, 2010, Jacoby, Ansley, Trevino, 1992).

Unfortunately the time-temperature data provided by thermocouple dataloggers are often misused. Flame temperature has been conflated with intensity for at least 60 years (e.g., Davies, Boyd, Bates, Hulet, 2015, Ryan, Williams, 2011, Whittaker, 1961). A majority of studies incorrectly associate the duration of heating above a certain threshold with physical or biological responses (McGranahan, 2020), despite the rate and magnitude of energy flux being the appropriate measurements (Bova, Dickinson, 2005, Dickinson, Johnson, 2004). Both the concept of the “heat dose” and the organismal response threshold have been harshly criticized (Pingree, Kobziar, 2019, Smith, Cowan, Fitzgerald, 2016). As such, a more appropriate application of thermocouple data is actually to use the timestamps to measure the movement of fire through the landscape (e.g., Davies et al., 2009). Unlike temperature, rate of spread is predicted by fire behavior models and can be used to calculate fire intensity in lieu of measuring it directly (Byram, 1959, Rothermel, 1972). However, deploying a sufficient number of thermocouples to sample rate of spread through heterogeneous wildland fuels is prohibitively expensive for many researchers given the high cost of purchasing and protecting multi-sensor dataloggers.

Many agricultural scientists seeking low-cost solutions to expensive electronic instrumentation and environmental sensors have discovered the open-source, do-it-yourself (DIY) Arduino platform (e.g., Barnard, Findley, Csavina, 2014, Greenspan, Morris, Warburton, Edwards, Duffy, Pike, Schwarzkopf, Alford, 2016, McGranahan, Geaumont, Spiess, 2018, Shipley, Kapoor, Dreelin, Winkler, 2017). Several published comparisons demonstrate that Arduino microprocessors are comparable in performance to commercial dataloggers. The computational accuracy of Arduino microprocessors are satisfactory for use in psychological and neurophysiology experiments (D’Ausilio, 2012), and Arduino systems have been used to monitor temperature via K-type thermocouples (Arnold et al., 2015). These Arduino-based K-type thermocouple dataloggers perform comparably to proprietary, commercial systems, such as the Campbell Scientific CR1000 (McGranahan and Poling, 2020). Here I describe the design, assembly, and research advantages of a compact and mobile solution for recording time-temperature data from thermocouples in agris—in the field, during wildland fires. This paper covers each step in the use of the DIY FeatherFlame system: How to construct it, program it, and protect it from heat and flame, and how to process, analyze, and interpret data using open-source software.