Tang et al. (2025) Seamless Reconstruction of MODIS Land Surface Temperature via Multi-Source Data Fusion and Multi-Stage Optimization

⚠️ Warning: This summary was generated from the abstract only, as the full text was not available.

Identification

• Journal: Remote Sensing
• Year: 2025
• Date: 2025-10-07
• Authors: Yanjie Tang, Yanling Zhao, Yueming Sun, Sumei Ren, Zhibin Li
• DOI: 10.3390/rs17193374

Research Groups

Not specified in the provided text.

Short Summary

This study develops a seamless MODIS Land Surface Temperature (LST) reconstruction framework by integrating multi-source data fusion and a multi-stage optimization strategy, achieving high spatiotemporal fidelity and outperforming conventional methods.

Objective

• To develop a robust and generalizable framework for reconstructing seamless MODIS LST with high spatiotemporal continuity, effectively addressing data gaps caused by cloud contamination and atmospheric interference.

Study Configuration

• Spatial Scale: Regional scale, demonstrated over two representative regions (Huainan and Jining), with a focus on reconstructing MODIS LST (typically 1 km resolution) while retaining spatial detail confirmed by cross-validation with high-resolution Landsat LST.
• Temporal Scale: Designed to address frequent data gaps and enhance temporal consistency and seasonal fidelity, implying continuous reconstruction over periods sufficient to capture seasonal variations.

Methodology and Data

• Models used:
  • Topography- and land cover-constrained spatial interpolation
  • Random Forest (RF) modeling
  • HANTS (Harmonic ANalysis of Time Series) for temporal smoothing
  • Poisson-based image fusion
• Data sources:
  • MODIS LST products
  • Multi-source predictors: Normalized Difference Vegetation Index (NDVI), Normalized Difference Water Index (NDWI), surface reflectance, Digital Elevation Model (DEM), land cover
  • High-resolution Landsat LST for cross-validation

Main Results

• The proposed integrated framework demonstrated superior performance in reconstructing MODIS LST for both daytime and nighttime scenarios over Huainan and Jining.
• The integrated approach achieved high accuracy, with correlation coefficients (CCs) exceeding 0.95 and root mean square errors (RMSEs) below 2 K.
• The method significantly outperformed conventional HANTS and standalone interpolation techniques.
• Cross-validation using high-resolution Landsat LST confirmed the framework's ability to retain spatial detail and maintain cross-scale consistency.

Contributions

• Presents an innovative and robust seamless MODIS LST reconstruction framework that integrates multi-source data fusion and a multi-stage optimization strategy.
• Effectively addresses the critical issue of frequent data gaps in MODIS LST products, enhancing their applicability for various studies.
• Demonstrates superior accuracy and performance compared to existing conventional methods for LST reconstruction.
• Provides a generalizable solution with high spatial and temporal fidelity, offering strong potential for broad applications in land surface process modeling, regional climate studies, and urban thermal environment analysis.

Funding

Not specified in the provided text.

Citation

@article{Tang2025Seamless,
  author = {Tang, Yanjie and Zhao, Yanling and Sun, Yueming and Ren, Sumei and Li, Zhibin},
  title = {Seamless Reconstruction of MODIS Land Surface Temperature via Multi-Source Data Fusion and Multi-Stage Optimization},
  journal = {Remote Sensing},
  year = {2025},
  doi = {10.3390/rs17193374},
  url = {https://doi.org/10.3390/rs17193374}
}