Koalas are in crisis, and their picky eating habits aren’t making it any easier. But what if we could use cutting-edge technology to find their perfect meal from the sky? That’s exactly what researchers are doing with airborne hyperspectral imaging, a game-changing tool for spotting the exact eucalyptus trees koalas love. Dubbed Project Airbear, this initiative uses a high-tech hyperspectral imager mounted on a light aircraft to scan fields and forests near Gunnedah, NSW, in search of the ideal koala habitat.
Here’s the fascinating part: this technology doesn’t just look at trees—it analyzes them down to the molecular level. By using narrowband visible and infrared light, it detects subtle characteristics like leaf pigment, water content, and, most crucially, nitrogen levels. And this is the part most people miss: koalas aren’t just fussy; they rely on specific nitrogen levels in eucalyptus leaves to survive. With habitat loss threatening their populations, finding and preserving these nutrient-rich trees is more critical than ever.
Professor Mathew Crowther from the University of Sydney calls it a ‘Goldilocks’ situation. It’s not enough to find the right tree species—it must also have the perfect nutritional quality to support koalas long-term. Even within the same species, these qualities can vary wildly. But here’s where it gets controversial: while other studies have used remote sensing to predict tree chemistry, they’ve often sacrificed accuracy for coverage. UAVs offer precision but limited range, while satellites cover vast areas but lack detail. Airborne hyperspectral imaging, however, strikes a balance, making it ideal for regional ecological studies.
Published in Science of the Total Environment (https://doi.org/10.1016/j.scitotenv.2025.181104), this research highlights how airborne hyperspectral data detects subtle spectral differences critical for predicting foliar chemistry and distinguishing between closely related tree species. Its finer spatial resolution also minimizes errors, making it perfect for assessing habitat quality. The study, a collaboration between the University of Sydney, the Sydney Institute of Agriculture, the University of New England, and HyVista Corporation, used the HyMap hyperspectral scanner—a powerhouse with 128 bands across the solar wavelength region, gyro stabilization, and NIST-traceable calibration for low-noise imagery.
The findings are groundbreaking. By using ‘pixel-based’ datasets, researchers can train models to predict tree features despite noise from factors like branch differences or shading. This approach improves predictions and works even with limited data—a common challenge in ecology. But is ‘pixel-based’ data the future of habitat mapping, or are we overlooking simpler methods? Let us know what you think in the comments.
Looking ahead, Professor Bradley Evans from the University of New England emphasizes the potential of this technology for Australian ecology. With faster coverage than drones, routine flights could soon provide detailed assessments of biodiversity and koala habitats. The next phase? Partnering with NASA JPL to use their latest hyperspectral imager and establishing a ‘National Collaborative Facility’ to maximize this tech for Australia and beyond.
What do you think? Is airborne hyperspectral imaging the key to saving koalas, or are there better ways to protect their habitats? Share your thoughts below—we’d love to hear your perspective!
