Time series changes and influencing factors of fractional vegetation coverage under weed competition in wheat field ecosystems through remote sensing

Fractional Vegetation Cover (FVC) is a crucial indicator for assessing the vegetation status of terrestrial ecosystems. However, challenges remain in analyzing FVC time series changes and influencing factors. This study tries to address key challenges in FVC assessment by analyzing over 300 wheat ge...

Full description

Saved in:
Bibliographic Details
Main Authors: Guofeng Yang, Yong He, Zhenjiang Zhou, Lingzhen Ye, Hui Fang, Xuping Feng
Format: Article
Language:English
Published: Taylor & Francis Group 2025-08-01
Series:International Journal of Digital Earth
Subjects:
Online Access:https://www.tandfonline.com/doi/10.1080/17538947.2025.2504134
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Fractional Vegetation Cover (FVC) is a crucial indicator for assessing the vegetation status of terrestrial ecosystems. However, challenges remain in analyzing FVC time series changes and influencing factors. This study tries to address key challenges in FVC assessment by analyzing over 300 wheat germplasms using UAV remote sensing, multispectral imaging, and semantic segmentation. The Transformer-based PoolFormer model outperformed convolutional neural networks, achieving a two-year average mIoU of 93.1% using full-band multispectral data. Visualization confirmed the ability of PoolFormer to track FVC changes over time. While wheat germplasms exhibited consistent growth trends across regions, environmental factors influenced growth rates and durations. High sampling frequencies were essential for capturing dynamic FVC trends. FVC trends in wheat germplasms varied across growth stages, with Argentinian germplasm declining the least. European germplasms exhibited the highest maximum FVC, Oceanic germplasms showed high variability, and Asian and American germplasms had intermediate maximum FVC. Early-stage FVC correlated strongly with plant height, but this correlation weakened over time, while the leaf area index shifted from a positive to a negative correlation. Lodging led to FVC overestimation, with errors increasing over time, and weed interference significantly affected accuracy. These findings provide insights for smart breeding, supporting sustainable wheat ecosystems.
ISSN:1753-8947
1753-8955