Understanding the efficacy and efficiency of thermal infrared UAV for wildlife monitoring
Authors:
- 11 January 2026
- journal article
- research article
- Published by Springer Nature in Environmental Monitoring and Assessment
- Vol. 198 (2) , 118
Abstract
Biodiversity conservation requires rigorous wildlife assessments, and uncrewed aerial vehicles (UAVs) equipped with thermal infrared (TIR) cameras offer a promising tool for surveying large mammals. Despite growing use, formal comparisons of UAV survey methodologies remain limited. We evaluated manual versus programmed flight methods and conducted an orthomosaic trial in Chitwan National Park, Nepal, in May 2022, performing six flights per method across four Terai grassland sites. We compared wildlife counts, survey effort (flight length, duration, number of images, and post-processing requirements), and issues regarding image overlap. Manual flights required 35 min on average, covered 6370 m, and generated 86 images per flight, whereas programmed flights averaged 66 min, 9360 m, and 205 images, representing increases of 57% in flight time, 47% in distance, and 138% in image volume for programmed surveys. There was no significant difference in total mammal counts (P = 0.781) or for specific groups such as deer (P = 0.181) and rhinos (P = 0.515) between the manual and programmed flights. However, manual flights yielded imagery that was better suited for species identification. Both approaches were influenced by observer bias, either in real-time species identification during manual flights or post-processing for programmed flights. Our results highlight that for our study area and species of interest, manual UAV flights were able to reduce survey effort while maintaining comparable detection rates and improving species identification. Orthomosaic processing, using both direct georeferencing and Structure-from-Motion, proved largely ineffective for thermal imagery of mobile mammals, as moving animals were often excluded due to image overlap requirements. The study also offers guidance for designing UAV-based wildlife monitoring programs, highlighting the potential of AI, video, and advanced sensors, as well as important limitations to consider before conducting surveys.Keywords
Funding Information
- National Science Foundation (BCS 1826839)
This publication has 94 references indexed in Scilit:
- Imperfect detection and wildlife density estimation using aerial surveys with infrared and visible sensorsRemote Sensing in Ecology and Conservation, 2022
- Methodology matters when estimating deer abundance: a global systematic review and recommendations for improvementsThe Journal of Wildlife Management, 2022
- Multispecies detection and identification of African mammals in aerial imagery using convolutional neural networksRemote Sensing in Ecology and Conservation, 2021
- Drone‐based thermal remote sensing provides an effective new tool for monitoring the abundance of roosting fruit batsRemote Sensing in Ecology and Conservation, 2021
- Enumerating White‐Tailed Deer Using Unmanned Aerial VehiclesWildlife Society Bulletin, 2021
- New technologies in the mix: Assessing N‐mixture models for abundance estimation using automated detection data from drone surveysEcology and Evolution, 2020
- Spatially explicit power analysis for detecting occupancy trends for multiple speciesEcological Applications, 2019
- Assessing the impacts of imperfect detection on estimates of diversity and community structure through multispecies occupancy modelingEcology and Evolution, 2018
- Unmanned aircraft systems in wildlife research: current and future applications of a transformative technologyFrontiers in Ecology and the Environment, 2016
- Large scale wildlife monitoring studies: statistical methods for design and analysisEnvironmetrics, 2002