Modeling Canopy Height of Forest–Savanna Mosaics in Togo Using ICESat-2 and GEDI Spaceborne LiDAR and Multisource Satellite Data DOI Creative Commons
Arifou Kombate,

Guy Armel Fotso Kamga,

Kalifa Goı̈ta

et al.

Remote Sensing, Journal Year: 2024, Volume and Issue: 17(1), P. 85 - 85

Published: Dec. 29, 2024

Quantifying forest carbon storage to better manage climate change and its effects requires accurate estimation of structural parameters such as canopy height. Variables from remote sensing data machine learning models are tools that being increasingly used for this purpose. This study modeled the height forest–savanna mosaics in Sudano–Guinean zone Togo. Relative heights were extracted GEDI ICESat-2 products, which combined with optical, radar, topographic variables modeling. We tested four methods: Random Forest (RF), Support Vector Machine (SVM), Extreme Gradient Boosting (XGBoost) Deep Neural Network (DNN). The RF algorithm obtained best predictions using 98% relative (RH98). best-performing result was (r = 0.84; RMSE 4.15 m; MAE 2.36 m) compared 0.65; 5.10 3.80 m). Models developed during can be applied over large areas mosaics, enhancing dynamics monitoring line REDD+ objectives. provides valuable insights future spaceborne LiDAR other applications similar complex ecosystems offers local decision-makers a robust tool management.

Language: Английский

A Review: Potential of Earth Observation (EO) for Mapping Small-Scale Agriculture and Cropping Systems in West Africa DOI Creative Commons

Niklas Heiss,

Jonas Meier, Ursula Geßner

et al.

Land, Journal Year: 2025, Volume and Issue: 14(1), P. 171 - 171

Published: Jan. 15, 2025

West Africa faces a complex range of challenges arising from climatic, social, economic, and ecological factors, which pose significant risks. The rapidly growing population, coupled with persistently low agricultural yield, further exacerbates these A state-of-the-art monitoring data derivation systems are crucial for improving livelihoods enhancing food security. Despite smallholder farming accounting 80% cultivated cropland area providing about 42% the total employment in Africa, there exists lack comprehensive overview Remote Sensing (RS) products studies specifically tailored to systems, this review aims address. Through systematic literature comprising 163 SCI papers sourced Web Science database (Filter I), followed by full-text II), we analyze RS sensors, spatiotemporal distribution, temporal scales, crop types examined, thematic foci employed existing research. Our findings highlight predominance high very high-resolution, multispectral sensors as primary source observe that wide array available datasets, along increasing computing capacities, have shaped field over last years. By highlighting knowledge, study identifies potential pinpoints key research gaps. This sets stage future investigations aimed at addressing critical African systems.

Language: Английский

Citations

0
Canopy Height Mapper: a Google Earth Engine application for predicting global canopy heights combining GEDI with multi-source data. DOI Creative Commons
Cesar Alvites, Hannah O’Sullivan, Saverio Francini

et al.

Environmental Modelling & Software, Journal Year: 2024, Volume and Issue: 183, P. 106268 - 106268

Published: Nov. 12, 2024

Language: Английский

Citations

1
Estimation of canopy height based on multi-source remote sensing data using forest structure aided sample selection DOI

Yinpeng Zhao,

Shouhang Du, Kangning Li

et al.

International Journal of Remote Sensing, Journal Year: 2024, Volume and Issue: 45(7), P. 2235 - 2268

Published: March 20, 2024

Forest canopy height data are crucial for estimating forest carbon storage and assessing ecology. By utilizing satellite imagery, obtained from airborne or spaceborne LiDAR have been expanded footprint plot levels to spatially continuous elevation mapping of forests. However, current research suggests that without type presents a challenge in how effectively integrate multi-source ensure the samples adequately represent various types higher estimation accuracy. Therefore, this study proposes method considers structure integrates overcome challenge. First, stratified sampling based on (SSMFS) was proposed select training enhance their representativeness. Second, we combined GEDI ATL08 create dataset, enhancing geographic coverage increasing samples. Third, LiDAR-based model incorporates previously unconsidered openness features uses SSMFS Finally, improved accuracy by creating residual correction adjusts differences between scanner (ALS) estimates. This study, conducted Zhangwu County, achieved an R2 = 0.71, MAE 1.20 m, RMSE 1.71 m. These results show 51.06% increase R2, 26.38% decrease MAE, 24.00% compared recent research. In summary, profoundly amplifies predictive accuracy, providing clear advantage delineation regional maps.

Language: Английский

Citations

0
Modeling Canopy Height of Forest–Savanna Mosaics in Togo Using ICESat-2 and GEDI Spaceborne LiDAR and Multisource Satellite Data DOI Creative Commons
Arifou Kombate,

Guy Armel Fotso Kamga,

Kalifa Goı̈ta

et al.

Remote Sensing, Journal Year: 2024, Volume and Issue: 17(1), P. 85 - 85

Published: Dec. 29, 2024

Quantifying forest carbon storage to better manage climate change and its effects requires accurate estimation of structural parameters such as canopy height. Variables from remote sensing data machine learning models are tools that being increasingly used for this purpose. This study modeled the height forest–savanna mosaics in Sudano–Guinean zone Togo. Relative heights were extracted GEDI ICESat-2 products, which combined with optical, radar, topographic variables modeling. We tested four methods: Random Forest (RF), Support Vector Machine (SVM), Extreme Gradient Boosting (XGBoost) Deep Neural Network (DNN). The RF algorithm obtained best predictions using 98% relative (RH98). best-performing result was (r = 0.84; RMSE 4.15 m; MAE 2.36 m) compared 0.65; 5.10 3.80 m). Models developed during can be applied over large areas mosaics, enhancing dynamics monitoring line REDD+ objectives. provides valuable insights future spaceborne LiDAR other applications similar complex ecosystems offers local decision-makers a robust tool management.

Language: Английский

Citations

0