Integrating unoccupied aerial systems and satellite data to map the patchiness of bare ground at a landscape scale

Authors: PONCE-CAMPOS, GUILLERMO - University Of Arizona; Heilman, Philip; NORTON, CYNTHIA - University Of Arizona; GAO, SHANG - University Of Arizona; CRIMMINS, MICHAEL - University Of Arizona; MCCLARAN, MITCHEL - University Of Arizona

Submitted to: Landscape Ecology
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 9/16/2025
Publication Date: 11/18/2025
Citation: Ponce-Campos, G.E., Heilman, P., Norton, C.L., Gao, S., Crimmins, M., McClaran, M. 2025. Integrating unoccupied aerial systems and satellite data to map the patchiness of bare ground at a landscape scale. Landscape Ecology. 40. Article 217. https://doi.org/10.1007/s10980-025-02226-6.
DOI: https://doi.org/10.1007/s10980-025-02226-6

Interpretive Summary: Semiarid rangelands face the threat of accelerated erosion, but understanding the threat is difficult at the landscape scale. Remote sensing is appropriate for monitoring at the landscape scale, but typically lacks substantial enough monitoring on the ground to have confidence in the landscape scale results. USDA researchers in Tucson, AZ in collaboration with University of Arizona researchers linked intermediate scale Unoccupied Aerial System (UAS) observations to satellite-based landscape assessments. The variable of interest, the Largest Patch Index (LPI) of bare ground patches (percent of the the total area in the largest patch), is preferred over bare ground because of the importance of connected bare patches allowing for more erosion by water. We found that LPI results were robust through both very dry and wet years and that LPI was spatially distributed as expected (more at lower elevations). This work will help detect and monitor ecological states at the landscape scale.

Technical Abstract: Context Integrating fine-scale measurements with broad-scale monitoring presents a persistent challenge in rangeland ecology, particularly when scaling detailed Unoccupied Aerial System (UAS) observations to satellite-based landscape assessments. This challenge is critical as rangelands face increasing climate variability, requiring reliable methods to detect and monitor ecological changes across landscapes. Objectives We investigated how the Largest Patch Index (LPI) of bare ground patches, derived from 3-dimensional UAS observations, can be scaled to landscape levels for mapping bare ground patchiness. Our study aimed to develop and validate methods for integrating UAS and satellite data to support landscape- scale ecological monitoring. Methods We conducted our study across a 100 km2 semi-arid rangeland in southern Arizona during 2019–2023, a period of extraordinary climate variability. We used Random Forest modeling to scale UAS-derived LPI measurements to satellite platforms (Landsat 8 and PlanetScope) with systematic comparison of spatial resolution and sensor data density effects. Our methodology maintained consistency across different sensor platforms while capturing finescale ecological processes. Results LPI effectively captured vegetation responses to extreme climate events, showing clear sensitivity to severe drought (SPEI -2.47) and wet periods (SPEI + 1.95). LPI values were consistently 30–60% higher in lower elevations, validating detection of known ecological gradients. LPI values increased with larger grid cell sizes in satellitederived estimates, with the magnitude varying by sensor data density. This data density effect represents a previously unrecognized mechanism that modifies scaling relationships independently of spatial resolution. The approach successfully integrated UAS training data with satellite observations for landscapescale pattern mapping. Conclusions This research provides a practical framework for integrating UAS and satellite observations to support ecological monitoring under increasing climate uncertainty. Our findings challenge fundamental assumptions about scale effects in landscape pattern analysis by revealing the role of sensor data density in scaling relationships.