Democratizing 3D ecology: Mobile neural radiance field for scalable ecosystem mapping in change detection

This is a Preprint and has not been peer reviewed. This is version 3 of this Preprint. You must log in to post a comment. There are no comments or no comments have been made public for this article. This is a Preprint and has not been peer reviewed. This is version 3 of this Preprint. Add a Comment You must log in to post a comment. Comments There are no comments or no comments have been made public for this article. 1. High-resolution ecological data are fundamental to understanding the structure, function, and change of ecosystems. Yet, the ways we capture and represent these data have remained largely constrained by expensive instruments and narrowly defined measurement paradigms. Here, we propose that radiance fields, representations that encode the color and density of points in a system, offer a powerful and general additional to any existing ecological observation framework. Radiance field models preserve not only geometry but also texture, reflectance, and viewpoint-dependent properties of ecosystems, enabling reusable and reinterpretable datasets as analytical methods advance. 2. We illustrate the democratizing potential of this technology using mobile implementations of neural radiance fields (NeRFs) captured with consumer-grade smartphones across open-grown and forested environments. Despite operating without specialized sensors, thesereconstructions reproduce key structural metrics with high fidelity relative to terrestrial laser scanning (TLS), while providing photorealistic, fine-scale renderings of vegetation. Beyond structural measurement, the same radiance field parameterization enables new analyses, volumetric slicing, virtual flythroughs, and temporal change detection, derived entirely from image data. 3. Drawing on ecological and computational literature, we also outline areas in which radiance field methods should improve to better serve the ecological community. We highlight five key areas for increased collaboration and focus: addressing occlusions, pieces of the canopy obscured due to vegetation; scale ambiguity; interpretability; comutational efficiency; and benchmark alignment. 4. By reframing radiance fields as a novel way to represent ecological data rather than as simply a reconstruction tool, we outline a path toward more democratized, flexible, and enduring modes of environmental monitoring. We show that, not only do radiance field methods represent a standalone mode of monitoring an ecological system, but fill a need in existing monitoring methodologies. As radiance field methods continue to evolve, they promise to make ecological datasets both more accessible and more expressive, supporting a shift from static measurement to dynamic, light-based understanding of ecosystems. https://doi.org/10.32942/X2M93F Life Sciences Neural Radiance Fields (NeRF); D ecology • Ecosystem monitoring • Terrestrial laser scanning (TLS) • Structure-from-Motion (SfM) • Remote sensing • Citizen science • Vegetation structure • Hype Published: 2025-06-26 19:52 Last Updated: 2025-11-29 05:52 CC BY Attribution 4.0 International Conflict of interest statement: None Data and Code Availability Statement: NA Language: English