Fine-scale monitoring of emergent plant based on spatiotemporal-spectral fusion

A geospatial dataset mapping the annual dynamics of emergent plant communities in Honghu Lake South, China, at a 2-meter resolution. The data was generated by Yifei Han using a spatiotemporal-spectral fusion model and a transformer-based classifier applied to long-term multispectral image sequences. The method achieved an overall mapping accuracy of 88.21% and was last updated in May 2026.

Use Cases

• Monitoring wetland restoration effectiveness based on annual plant community dynamics.
• Estimating potential carbon storage changes in wetlands based on fine-scale plant mapping.
• Developing or benchmarking remote sensing fusion methods for ecological monitoring based on the described spatiotemporal-spectral fusion approach.
• Analyzing shifts in plant community composition over time based on high-resolution annual maps.

Strengths

• Achieves a fine 2-meter spatial resolution for wetland plant mapping.
• Reports a high overall classification accuracy of 88.21%.
• Provides annual temporal dynamics, enabling time-series analysis.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• The dataset is small (2.5 MB), which may indicate limited spatial or temporal coverage.

Provenance

Source

Yifei Han via figshare.

Collection Method

Generated using a TemPanSharpening net fusion model and a transformer-based classifier on multispectral image sequences.

Time Range

Annual dynamics, but specific years are not stated.

Freshness

Last updated 2026-05-06 06:40:04.

Geography

Honghu Lake South, China.