Li et al. (2026) Atmospheric Drivers and Spatiotemporal Variability of Pan Evaporation Across China (2002–2018)

Identification

• Journal: Atmosphere
• Year: 2026
• Date: 2026-01-10
• Authors: Shuai Li, X. Z. Li
• DOI: 10.3390/atmos17010073

Research Groups

• Campus Security Department, Xinjiang College of Science and Technology, Korla, China
• College of Geography and Remote Sensing Science, Xinjiang University, Urumqi, China

Short Summary

This study investigated the spatiotemporal variability and climatic controls of pan evaporation (PE) across China from 2002 to 2018 using data from 759 meteorological stations. It revealed a nationwide decreasing interannual PE trend with strong regional contrasts, particularly declines in northern China, and demonstrated that dominant climatic drivers vary significantly across different climate zones.

Objective

• To identify dominant temporal trends and periodic components of PE.
• To characterize the spatial heterogeneity of PE variations across seven major climate zones.
• To quantify the climatic drivers responsible for the observed regional differences in PE.

Study Configuration

• Spatial Scale: Mainland China (approximately 20°N–53°N and 73°E–135°E), encompassing seven major climate zones, utilizing daily records from 759 meteorological stations.
• Temporal Scale: 2002–2018 (17 years) for daily and monthly pan evaporation and meteorological variables.

Methodology and Data

• Models used:
  • Mann–Kendall (MK) test
  • Sen’s slope estimator
  • Pettitt test
  • Seasonal-Trend decomposition using Loess (STL)
  • Continuous Wavelet Transform (CWT)
  • Ensemble Empirical Mode Decomposition (EEMD)
  • Empirical Orthogonal Function (EOF) analysis
  • Global Moran’s I
  • Local Indicators of Spatial Association (LISA)
  • Random Forest (RF) regression model
  • Multiple Linear Regression
  • Ridge Regression
• Data sources:
  • National Meteorological Information Center (NMIC) of the China Meteorological Administration (CMA) "Daily Dataset of Surface Meteorological Variables from Chinese National Ground Stations".
  • Variables: daily pan evaporation (PE), precipitation (P), mean air temperature (Tair), mean wind speed (WS), mean relative humidity (RH), and sunshine duration (SSD).
  • Nighttime lights (NTL) as an urbanization proxy (for sensitivity analysis).

Main Results

• A nationwide significant monotonic decreasing trend in annual mean PE was observed from 2002 to 2017, with a Sen’s slope of -0.015 mm day⁻¹ yr⁻¹ (MK p-value = 0.0074).
• Spatial analysis revealed a clear north–south contrast, with the strongest PE declines in northern, northeastern, and central China, while southern and coastal regions showed weakly positive or non-significant changes.
• Temporal decomposition techniques (STL, CWT, EEMD) consistently identified a dominant annual cycle (explaining 83.7% of total variance) and a pronounced peak in PE in 2018, attributed to anomalous warming (+3.388 °C) and widespread precipitation deficits (-59.404 mm) nationally.
• EOF analysis revealed a dominant dipole pattern (EOF1, 46.0% variance) with positive loadings over northern and eastern regions and negative loadings over western China/Tibetan Plateau, indicating contrasting PE variability between monsoon-influenced eastern China and westerly/topography-controlled western interior.
• The Random Forest model achieved good predictive performance (R² = 0.493, RMSE = 1.777 mm) on an independent test set, outperforming linear baselines (R² ≈ 0.429).
• Dominant climatic drivers of PE varied substantially across climate zones: sunshine duration and air temperature primarily controlled PE in humid regions, while wind speed and relative humidity exerted stronger influences in arid and semi-arid regions.
• Sensitivity analyses showed that within-month extremes (95th percentile of daily wind speed and air temperature) provided incremental explanatory power but did not alter the primary driver structure, and an urbanization proxy (Nighttime Lights) also had a small additional contribution without changing the dominant meteorological drivers.

Contributions

• Provides a comprehensive, climate-zone-resolved assessment of PE variability across China for 2002–2018, integrating multi-scale decomposition, spatial statistics, and machine learning methods.
• Quantifies the nonlinear contributions of meteorological variables to monthly PE across diverse climate zones, revealing region-specific dominant drivers along humidity gradients.
• Offers insights into the spatiotemporal heterogeneity of PE responses to climate forcing, which is crucial for drought assessment and water resource management in a warming climate.

Funding

• National Natural Science Foundation of China (42301414)
• China Postdoctoral Science Foundation (2023M732959)

Citation

@article{Li2026Atmospheric,
  author = {Li, Shuai and Li, X. Z.},
  title = {Atmospheric Drivers and Spatiotemporal Variability of Pan Evaporation Across China (2002–2018)},
  journal = {Atmosphere},
  year = {2026},
  doi = {10.3390/atmos17010073},
  url = {https://doi.org/10.3390/atmos17010073}
}