Mo et al. (2026) Distinct temperature sensitivity indicators of regional vegetation greening across different satellite observations

Identification

• Journal: Ecological Indicators
• Year: 2026
• Date: 2026-01-01
• Authors: Jinglin Mo, Na Dong, Wenjuan Shen, Chong Liu, Zhen Liu, Huabing Huang
• DOI: 10.1016/j.ecolind.2025.114604

Research Groups

• School of Geospatial Engineering and Science, Sun Yat-sen University, Zhuhai, China
• Key Laboratory of Natural Resources Monitoring in Tropical and Subtropical Area of South China, Ministry of Natural Resources, Zhuhai, China
• College of Forestry, Nanjing Forestry University, Nanjing, China
• Earth, Ocean and Atmospheric Sciences (EOAS) Thrust, Function Hub, The Hong Kong University of Science and Technology (Guangzhou), Guangzhou, China

Short Summary

This study quantitatively assesses the biophysical effects of vegetation greening on land surface temperature (LST) and near-surface air temperature (T2) across three ecological restoration regions in China, clarifying uncertainties from different satellite datasets. It finds that openland-to-forest conversions produce substantially stronger cooling than within-type greening, with LST sensitivity generally 2–4 times stronger than T2, and highlights the importance of multi-dataset comparisons.

Objective

• To explicitly evaluate the efficiency of heterogeneous vegetation greening on land surface temperature (LST) and near-surface air temperature (T2).
• To clarify uncertainties in these biophysical effects arising from different satellite-derived Leaf Area Index (LAI) and Land Use/Land Cover (LULC) datasets.
• To analyze the differences in LST and T2 responses under two greening scenarios (within-type LAI increase and openland-to-forest conversion) using the space-for-time method.
• To provide a scientific basis for understanding the climate regulation mechanisms of vegetation greening in ecologically critical regions of China.

Study Configuration

• Spatial Scale: Three ecological restoration regions in China (Southwest China, the Loess Plateau, and Northeast China). Data resampled to 1 km spatial resolution, with a 7 km × 7 km moving window.
• Temporal Scale: Summer period (June–August, DOY 153–241) from 2003 to 2019.

Methodology and Data

• Models used: Space-for-time approach for biophysical sensitivity calculation; Spatially Varying Coefficient Model with Sign Preservation (SVCM-SP) algorithm for T2 data generation; Generalized Regression Neural Networks (GRNN) for GLASS LAI; Spatiotemporal filtering for BNUV6 LAI.
• Data sources:
  • LAI: GLASS LAI product (500 m, 8-day), reprocessed MODIS LAI dataset (BNUV6, 500 m, 8-day).
  • LULC: MODIS MCD12Q1 (500 m, annual), CNLUCC (30 m, multi-period), ESA CCI (300 m, annual).
  • LST: MODIS MYD11A2 (1 km, 8-day).
  • T2: Seamless 1 km daily near-surface air temperature dataset (1 km, daily).
  • ET: MODIS MOD16A2 (500 m, 8-day).
  • Albedo: GLASS albedo product (1 km, 8-day).
  • DEM: Copernicus DEM (30 m).

Main Results

• Vegetation greening generally exerts cooling effects on both LST and T2, with LST exhibiting a 2–4 times stronger response than T2 due to its direct reflection of surface energy balance.
• Openland-to-forest conversions produce substantially stronger cooling effects on both LST and T2 compared to within-type greening processes.
  • Mean LST sensitivity for growth greening (based on GLASS LAI and MCD12Q1) is −0.32 ± 0.03 K⋅(LAI)−1.
  • Mean T2 sensitivity for growth greening (based on GLASS LAI and MCD12Q1) is −0.08 ± 0.01 K⋅(LAI)−1.
  • For openland-to-forest conversion (CNLU dataset), ΔLST/ΔLAI values range from −0.38 ± 0.05 K⋅(LAI)−1 (Northeast China) to −0.80 ± 0.09 K⋅(LAI)−1 (Loess Plateau).
  • Corresponding ΔT2/ΔLAI values range from −0.14 ± 0.01 K⋅(LAI)−1 (Northeast China) to −0.32 ± 0.03 K⋅(LAI)−1 (Loess Plateau).
• Enhanced evapotranspiration (ET) is the dominant cooling driver, with ΔET/ΔLAI generally positive across all vegetation types and conversion scenarios, often exceeding 0.2 kg⋅m−2⋅day−1.
• Albedo generally decreases with LAI increase (negative ΔAlbedo/ΔLAI), but its potential warming effect is often offset or masked by strong ET-induced cooling.
• Temperature sensitivity estimates vary among satellite datasets: GLASS LAI generally estimates larger cooling magnitudes than BNUV6 LAI. Different LULC datasets (MCD, CNLU, ESA) show varying spatial distributions and sample sizes, affecting sensitivity estimates, particularly for within-type greening.

Contributions

• Uniquely quantifies temperature (LST, T2) – LAI sensitivities and their multi-source satellite uncertainties across both within-type greening and openland-to-forest conversion processes.
• Highlights how the choice of satellite product (LAI and LULC) fundamentally affects the estimated climate benefits of vegetation greening.
• Provides practical guidance for policymakers and restoration planners to optimize land-use planning and afforestation strategies for maximizing climate benefits, including prioritizing specific regions and conversion types.
• Offers empirical constraints on the biophysical cooling effects of vegetation greening and land-cover transitions, clarifying the dominant role of ET in driving temperature reductions.

Funding

• National Natural Science Foundation of China (Grant No. 42301027, 42405020)
• Guangdong Basic and Applied Basic Research Foundation (Grant ID. 2024A1515011093, 2022A1515111033)
• Governmental project of Guangdong Province, China (2023QN10L422)
• Guangzhou-HKUST(GZ) Joint Funding Program (2025A03J3949)
• Guangzhou Municipal Science and Technology Project for Maiden Voyage (no.2024A04J4523)

Citation

@article{Mo2026Distinct,
  author = {Mo, Jinglin and Dong, Na and Shen, Wenjuan and Liu, Chong and Liu, Zhen and Huang, Huabing},
  title = {Distinct temperature sensitivity indicators of regional vegetation greening across different satellite observations},
  journal = {Ecological Indicators},
  year = {2026},
  doi = {10.1016/j.ecolind.2025.114604},
  url = {https://doi.org/10.1016/j.ecolind.2025.114604}
}