Satellite-derived plant cover maps vary in performance depending on version and product

Cara Applestein; Matthew J. Germino

doi:10.1016/j.ecolind.2023.110950

Home/Archive/USGS

USGS · 70248492

Satellite-derived plant cover maps vary in performance depending on version and product

Cara Applestein·Matthew J. Germino

View original source ↗Export metadata ↓

Abstract

Understanding the accuracy and appropriate application scale of satellite-derived maps of vegetation cover is essential for effective management of the vast, remote rangelands of the world. However, the underlying models are updated frequently and may combine with rapidly changing vegetation conditions to cause variations in accuracy and precision over time. We sought to assess how model performance changed between different versions of satellite-derived cover products (Rangeland Analysis Platform, RAP, and Rangeland Condition Monitoring and Assessment Protocol, RCMAP) and how the performance of LandCart compared to RAP and RCMAP. Additionally, we asked how variability in agreement between LandCart and field-based models varied with scale. We utilized an intensive dataset of grid-point intercept functional group cover data collected between 2016 and 2020 across the ∼113 kHA 2015 Soda Wildfire to 1) evaluate r 2 agreement between versions of each satellite-derived product and plot-level field data and 2) assess relative standard error of agreement in cover between LandCart and continuous field-based Empirical Bayesian Kriging (EBK) regression models. Agreement between satellite- compared to field-plot values of cover (r 2 ) increased for RCMAP Version 5.0 compared to Version 2.0, but there were negligible changes between versions of RAP. Despite this, r 2 values of RCMAP and LandCart were nearly always less than RAP. Variability in agreement between EBK regression model cover and LandCart-derived cover decreased with the scale of consideration. Variability in agreement between satellite-derived cover products and field-based metrics is lowest at larger scale (mega-fire or regional) and varies from year to year and across versions, which could complicate detection of temporal changes in plant cover.

Explore related subjects

Ecological Indicators Article

Source-reported bounding extent: 42.354755589015696° to 43.926296989230224° latitude; -117.56338308098071° to -115.93309004811645° longitude. This indicates report coverage, not an exact sampling location. View area on OpenStreetMap.

Keep this discovery

Explore connections, maps & timelines

BibTeX RIS

Cara Applestein, Matthew J. Germino. 2023. Satellite-derived plant cover maps vary in performance depending on version and product. https://doi.org/10.1016/j.ecolind.2023.110950

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related USGS reports

USGS2026

A tool to monitor hydrologic conditions on tree islands in the Everglades

Tree islands are patchy upland forested habitats in Florida's Everglades that face degradation and disappearance due to altered hydrologic patterns. The U.S. Geological Survey coordinated with the Miccosukee Tribe of Indians of Florida and the Seminole Tribe of Florida to co-develop a decision-support tool based on tree-island hydrologic conditions. Everglades managers can use this tool to help with restoration planning and water operations decisions that affect tree-island conditions. After a series of organized workshops and meetings, a list of hydrologic metrics was selected as indicators of tree-island health, including hydroperiod, number of days since last dry, and maximum water depth at the head of the island. As a result, a web application tool, called ETree, has been developed and is publicly available online. This web application provides data on daily metrics for the current Everglades water year and annual summaries for past years, beginning in 2000.

Florida↗

USGS2026

Vulnerability of different Colorado Plateau land types to drivers of change

Public lands are challenged by a range of pressures—changing climate, increasing visitation, resource extraction—and their effects can span spatial scales, often crossing land management jurisdictional boundaries. Research approaches which explicitly span jurisdictions can support strategies to contend with regional pressures. We assess management-relevant drivers of change—aridification, livestock grazing, invasive species, surface disturbance, and fire—across a patchwork of land management units and agencies on the Colorado Plateau, focusing on southeastern Utah, USA. We use vulnerability analysis, first evaluating exposure to drivers across the landscape, then quantifying sensitivity to each driver across different land types, defined by mapped Ecological Site Groups, a system for classifying landscapes according to physical factors including climate, soils, and topographic setting. We address the questions: 1) how are drivers spatially distributed across the study region; and, 2) based on exposure and sensitivity, are certain land types more vulnerable to these drivers? We find that the study region has high exposure and sensitivity—and thus high vulnerability—to aridification and grazing, but low exposure and vulnerability to other drivers. Although more sensitive land types were not generally more exposed, identifying which areas are most sensitive can guide adaptive measures, like where new uses or disturbances would be least harmful and which areas could be prioritized for restoration. The method we demonstrate is a flexible tool for assessing landscape-scale impacts, is built on nationally available datasets, and can be tailored to different datasets and sensitivity metrics.

Utah↗

USGS2026

Satellite time series analysis to quantify changing climax ciénegas using a state and transition model approach

Ciénegas are rare wetlands in arid landscapes of the North American Southwest, historically providing critical ecological and hydrological functions but increasingly threatened by changing climate and land use pressures. This study quantifies changes in ciénega condition and floodplain dynamics using a state-and-transition model (STM) informed by expert knowledge and remote sensing. Key factors include woody plant encroachment, water availability, and soil aggradation. We mapped 31 ciénegas with high-resolution imagery and analyzed Landsat data (1985–2023) to assess vegetation health and moisture using the Normalized Difference Vegetation Index (NDVI) and Normalized Difference Infrared Index (NDII). Results show substantial interannual variability in phenology, water stress, and soil moisture, with regional drying and elevation strongly influencing ciénega resilience. We classified ciénegas into three functional states—healthy, desiccated, and dormant—and mapped their 2023 condition. Trend analyses indicate most ciénegas exhibit greening despite drought, though localized variability underscores the need for site-specific management. None are in a stable climax (reference) state; rather, they transition among states in response to external drivers. Increasing woody plant cover and surface drying, likely linked to declining regional water tables, favor deep-rooted species over wetland grasses—a pattern mirrored in adjacent control plots. Spatially explicit analysis revealed intra-ciénega variability often masked by aggregated data, highlighting the importance of high-resolution monitoring. Seasonal and long-term trends provide context for understanding ciénega dynamics, including degradation and restoration pathways. This study emphasizes the importance of groundwater conservation and demonstrates how remote sensing supports long-term monitoring. The STM framework offers a practical tool for adaptive management to sustain freshwater resources in arid environments.

Arizona, New Mexico↗