New Pub: Reliability of Manual Counting for Thermal Census of Bats Within Human-Made Structures
New research evaluates the accuracy and consistency of thermal imaging methods for monitoring wild bat populations
Monitoring wild bat populations can be challenging, particularly when colonies roost in large or complex human-made structures. Thermal infrared cameras offer a non-invasive approach to monitoring bat emergence and estimating population sizes, but the reliability of manually counting bats in thermal footage can vary. In our new study published in Remote Sensing in Ecology and Conservation, researchers evaluated the consistency of manual bat counts across thermal videos from three Northern California roosts inhabited by Mexican free-tailed bats (Tadarida brasiliensis). The team found that differences in observer counting practices and the presence of non-bat flying objects, including insects and birds, contributed to variation in population estimates. At the largest roost, differences between observers could translate to discrepancies of more than 20,000 bats across an entire emergence.
These findings highlight the importance of standardized counting methods, observer training, and statistical validation when using thermal imagery to monitor wildlife populations. Establishing reliable manual counting methods is also essential for developing and validating automated computer vision tools that could improve the efficiency and scalability of wildlife monitoring.
Lagattuta SK, Freeman JME, Gardner NR, Johnson CK. Reliability of Manual Counting for Thermal Census of Bats Within Human-Made Structures. Remote Sensing in Ecology and Conservation. 2026 Sep 6. https://doi.org/10.1002/rse2.70102