Ensemble Machine Learning for Mapping Tree Species Alpha-Diversity Using Multi-Source Satellite Data in an Ecuadorian Seasonally Dry Forest DOI Creative Commons
Steven E. Sesnie, Carlos I. Espinosa, Andrea Jara‐Guerrero

et al.

Remote Sensing, Journal Year: 2023, Volume and Issue: 15(3), P. 583 - 583

Published: Jan. 18, 2023

The increased variety of satellite remote sensing platforms creates opportunities for estimating tropical forest diversity needed environmental decision-making. As little as 10% the original seasonally dry (SDTF) remains Ecuador, Peru, and Colombia. Remnant forests show high rates species endemism, but experience degradation from climate change, wood-cutting, livestock-grazing. Forest census data provide a vital resource examining methods to estimate levels. We used spatially referenced trees ≥5 cm in diameter simulated 0.10 ha plots measured 9 SDTF southwestern Ecuador compare machine learning (ML) models six α-diversity indices. developed 1 m tree canopy height elevation stem mapped trees, at scale conventionally derived light detection ranging (LiDAR). then an ensemble ML approach comparing single- combined-sensor RapidEye, Sentinel-2 interpolated topography surfaces. Validation showed that combined often outperformed single-sensor approaches. Combined sensor model ensembles richness, Shannon’s H, inverse Simpson’s, unbiased Fisher’s alpha indices typically lower root mean squared error (RMSE) goodness fit (R2). Piélou’s J, measure evenness, was poorly predicted. Mapped richness (R2 = 0.54, F 27.3, p <0.001) H′ 26.9, most favorable agreement with field validation observations (n 25). Small-scale experiments revealed essential relationships between multiple sensors repeated global coverage can help guide larger-scale biodiversity mapping efforts.

Language: Английский

Human fingerprint on structural density of forests globally DOI
Wang Li, Wen‐Yong Guo, Maya Pasgaard

et al.

Nature Sustainability, Journal Year: 2023, Volume and Issue: 6(4), P. 368 - 379

Published: Jan. 19, 2023

Language: Английский

Citations

48
LiDAR GEDI derived tree canopy height heterogeneity reveals patterns of biodiversity in forest ecosystems DOI Creative Commons
Michele Torresani, Duccio Rocchini,

Alessandro Alberti

et al.

Ecological Informatics, Journal Year: 2023, Volume and Issue: 76, P. 102082 - 102082

Published: March 30, 2023

The "Height Variation Hypothesis" is an indirect approach used to estimate forest biodiversity through remote sensing data, stating that greater tree height heterogeneity (HH) measured by CHM LiDAR data indicates higher structure complexity and species diversity. This has traditionally been analyzed using only airborne which limits its application the availability of dedicated flight campaigns. In this study we relationship between diversity HH, calculated with four different indices two freely available CHMs derived from new space-borne GEDI data. first, a spatial resolution 30 m, was produced regression machine learning algorithm integrating Landsat optical information. second, 10 created Sentinel-2 images deep convolutional neural network. We tested separately in plots situated northern Italian Alps, 100 forested area Traunstein (Germany) successively all 130 cross-validation analysis. Forest density information also included as influencing factor multiple Our results show can be assess patterns ecosystems estimation HH correlated However, indicate method influenced factors including dataset choice their related resolution, calculate density. finding suggest LIDAR valuable tool ecosystems, aid global estimation.

Language: Английский

Citations

45
Integrating GEDI, Sentinel-2, and Sentinel-1 imagery for tree crops mapping DOI Creative Commons
Esmaeel Adrah, Jesse P. Wong, He Yin

et al.

Remote Sensing of Environment, Journal Year: 2025, Volume and Issue: 319, P. 114644 - 114644

Published: Feb. 11, 2025

Language: Английский

Citations

2
Vegetation structure derived from airborne laser scanning to assess species distribution and habitat suitability: The way forward DOI
Vítězslav Moudrý, Anna F. Cord, Lukáš Gábor

et al.

Diversity and Distributions, Journal Year: 2022, Volume and Issue: 29(1), P. 39 - 50

Published: Oct. 30, 2022

Abstract Ecosystem structure, especially vertical vegetation is one of the six essential biodiversity variable classes and an important aspect habitat heterogeneity, affecting species distributions diversity by providing shelter, foraging, nesting sites. Point clouds from airborne laser scanning (ALS) can be used to derive such detailed information on structure. However, public agencies usually only provide digital elevation models, which do not Calculating structure variables ALS point requires extensive data processing remote sensing skills that most ecologists have. extremely valuable for many analyses use distribution. We here propose 10 should easily accessible researchers stakeholders through national portals. In addition, we argue a consistent selection their systematic testing, would allow continuous improvement list keep it up‐to‐date with latest evidence. This initiative particularly needed advance ecological research open datasets but also guide potential users in face increasing availability global products.

Language: Английский

Citations

62
Structural and species diversity explain aboveground carbon storage in forests across the United States: Evidence from GEDI and forest inventory data DOI Creative Commons
Erin T.H. Crockett, Jeff W. Atkins, Qinfeng Guo

et al.

Remote Sensing of Environment, Journal Year: 2023, Volume and Issue: 295, P. 113703 - 113703

Published: July 6, 2023

Language: Английский

Citations

36
Monitoring Earth’s climate variables with satellite laser altimetry DOI
Lori A. Magruder, S. L. Farrell, Amy Neuenschwander

et al.

Nature Reviews Earth & Environment, Journal Year: 2024, Volume and Issue: 5(2), P. 120 - 136

Published: Jan. 30, 2024

Language: Английский

Citations

15
Integrating remote sensing and field inventories to understand determinants of urban forest diversity and structure DOI Creative Commons
Vinícius Marcilio‐Silva,

Sally Donovan,

Sarah E. Hobbie

et al.

Ecology, Journal Year: 2025, Volume and Issue: 106(2)

Published: Feb. 1, 2025

Abstract Understanding the determinants of urban forest diversity and structure is important for preserving biodiversity sustaining ecosystem services in cities. However, comprehensive field assessments are resource‐intensive, landscape‐level approaches may overlook heterogeneity within regions. To address this challenge, we combined remote sensing with inventories to comprehensively map analyze attributes patches across Minneapolis‐St. Paul Metropolitan Area (MSPMA) a multistep process. First, developed predictive machine learning models by integrating data from (from 40 12.5‐m‐radius plots) Global Ecosystem Dynamics Investigation (GEDI) observations Sentinel‐2‐derived land surface phenology (LSP). These enabled accurate predictions attributes, specifically nine metrics plant (tree species richness, tree abundance, understory abundance), (average canopy height, dbh, density), structural complexity (variability density) relative errors ranging between 11% 21%. Second, applied these predict 804 additional plots GEDI Sentinel‐2. Finally, Bayesian multilevel predicted assess influence multiple factors—patch dimensions, landscape plot position, jurisdictional agency—on plots. The showed all predictors have some degree effect on presenting varying explanatory power R 2 values 0.071 0.405. Overall, characteristics (e.g., distance nearest trail, proximity edge) agency explained large portion variability patches, whereas patch did not. versus management sets marginal Δ was heterogeneous ecological subsections (an classification designation). multiplicity influencing forests emphasizes intricate nature ecosystems highlights nuanced, relationships anthropogenic factors that determine properties. Effectively enhancing requires assessments, management, conservation strategies tailored context‐specific characteristics.

Language: Английский

Citations

1
Effects of environmental conditions on ICESat-2 terrain and canopy heights retrievals in Central European mountains DOI
Vítězslav Moudrý, Kateřina Gdulová, Lukáš Gábor

et al.

Remote Sensing of Environment, Journal Year: 2022, Volume and Issue: 279, P. 113112 - 113112

Published: June 17, 2022

Language: Английский

Citations

31
Forest Structure Characterization in Germany: Novel Products and Analysis Based on GEDI, Sentinel-1 and Sentinel-2 Data DOI Creative Commons
Patrick Kacic, Frank Thonfeld, Ursula Geßner

et al.

Remote Sensing, Journal Year: 2023, Volume and Issue: 15(8), P. 1969 - 1969

Published: April 7, 2023

Monitoring forest conditions is an essential task in the context of global climate change to preserve biodiversity, protect carbon sinks and foster future resilience. Severe impacts heatwaves droughts triggering cascading effects such as insect infestation are challenging semi-natural forests Germany. As a consequence repeated drought years since 2018, large-scale canopy cover loss has occurred calling for improved disturbance monitoring assessment structure conditions. The present study demonstrates potential complementary remote sensing sensors generate wall-to-wall products combination high spatial temporal resolution imagery from Sentinel-1 (Synthetic Aperture Radar, SAR) Sentinel-2 (multispectral) with novel samples on Global Ecosystem Dynamics Investigation (GEDI, LiDAR, Light detection ranging) enables analysis dynamics. Modeling three-dimensional GEDI machine learning models reveals recent changes German due disturbances (e.g., degradation, salvage logging). This first consistent data set Germany 2017 2022 provides information height, biomass allows estimating at 10 m resolution. maps support better understanding post-disturbance

Language: Английский

Citations

20
StrucNet: a global network for automated vegetation structure monitoring DOI Creative Commons
Kim Calders, Benjamin Brede, Glenn Newnham

et al.

Remote Sensing in Ecology and Conservation, Journal Year: 2023, Volume and Issue: 9(5), P. 587 - 598

Published: April 14, 2023

Abstract Climate change and increasing human activities are impacting ecosystems their biodiversity. Quantitative measurements of essential biodiversity variables (EBV) climate used to monitor carbon dynamics evaluate policy management interventions. Ecosystem structure is at the core EBVs stock estimation can help inform assessments species diversity. also as an indirect indicator habitat quality expected richness or community composition. Spaceborne provide large‐scale insight into monitoring structural ecosystems, but they generally lack consistent, robust, timely detailed information regarding full three‐dimensional vegetation local scales. Here we demonstrate potential high‐frequency ground‐based laser scanning systematically changes in vegetation. We present a proof‐of‐concept high‐temporal ecosystem time series 5 years temperate forest using terrestrial (TLS). data from automated that allow upscaling scanning, overcoming limitations typically opportunistic TLS measurement approach. Automated will be critical component build network field sites required calibration for satellite missions effectively over large areas. Within this perspective, reflect on how could designed discuss implementation pathways.

Language: Английский

Citations

18