Station records over land show a widespread multi-decadal decline of near-surface wind speed since the 1960s–1980s ("terrestrial stilling"), with direct consequences for wind-energy resources because wind power density scales with the cube of wind speed. The unresolved question, posed quantitatively for China by Li et al. (2008) via observation-minus-reanalysis differencing, is how much of the observed decline in wind speed and wind-energy density is caused by local-to-regional anthropogenic land-surface modification (urbanization, afforestation and vegetation growth, irrigation and agricultural expansion) versus large-scale atmospheric circulation variability and climate change. Later work produced widely conflicting partitions (land-use shares from roughly 25% to 85% depending on region, period, wind metric, and method), showed that vegetation-greening forcing alone fails to reproduce the stilling in Earth-system models, attributed the post-2010 recovery mainly to internal ocean–atmosphere oscillations, and revealed that a large fraction of the raw observed trend is non-climatic instrumental inhomogeneity. A method-consistent, uncertainty-quantified attribution that includes the observational-inhomogeneity contribution remains an open problem.

Background

Near-surface wind speed (SWS) is the wind measured by anemometers at the standard meteorological height of about 10 m at surface stations; it drives evaporation and wind-energy resources, and wind power density (available wind energy per unit area swept by a turbine) scales as ρv3/2\rho v^{3}/2 with air density ρ\rho and speed vv, so a small wind-speed decline implies a large energy loss. Beginning in the 2000s, analysts noticed that station records over most land areas showed a systematic multi-decadal SWS decline — a phenomenon later called terrestrial stilling — while atmospheric reanalyses (state estimates produced by weather-forecast models constrained by assimilated observations) showed little or no such decline over land.

Two families of explanations were proposed. The first is local-to-regional anthropogenic modification of the underlying surface: urbanization, afforestation and vegetation growth, irrigation, and agricultural expansion raise the aerodynamic roughness length (the height at which the logarithmic surface-layer wind profile extrapolates to zero), increasing drag and slowing the wind measured nearby. The second is large-scale atmospheric variability and change: weakening of the horizontal pressure-gradient force, shifts of storm tracks and circulation modes, and decadal ocean–atmosphere oscillations, which reanalyses capture because they are driven by pressure and upper-air observations.

A key diagnostic is the observation-minus-reanalysis (OMR) method introduced by Kalnay and Cai (2003): because the NCEP/NCAR reanalysis assimilates pressure and upper-air data but is largely insensitive to local land-surface properties, the difference between station observations and the reanalysis at the same location is taken as an estimate of the surface-change signal. Applying OMR to 604 Chinese stations for 1960–1999, Li, Wang, Chu and Tang (2008) found that observed mean wind speed weakened and near-surface wind power declined at about −4.51 W m⁻² per decade, of which about −3.84 W m⁻² per decade was attributed to anthropogenic land-use change — that is, land-surface modification could account for the large majority of China's observed wind-energy decline.

Subsequent work complicated this picture in three ways. First, Vautard, Cattiaux, Yiou, Thépaut and Ciais (2010) extended the question to the northern mid-latitudes (822 stations, 1979–2008, SWS down 5–15% over almost all land) and bounded the partition rather than settling it: circulation changes captured by reanalyses explain 10–50%, and increased surface roughness from biomass growth and land-use change could explain 25–60%, of the stilling. Second, counter-evidence emerged: Zeng et al. (2018) showed that AMIP-type ensemble simulations (atmosphere–land model runs driven by observed sea-surface temperature and sea ice rather than by a coupled interactive ocean) of two Earth-system models (IPSL-CM and CESM) forced with observed vegetation greening do not reproduce significant land stilling, and Zeng et al. (2019) documented that global stilling reversed around 2010, attributing the decadal wind-speed variations mainly to internal ocean–atmosphere oscillations rather than vegetation growth or urbanization. Third, and most consequentially, the raw observations themselves were shown to be contaminated: Zhang and Wang (2020, 2023) documented anemometer replacements, instrument automation, and station relocations in Chinese records (inhomogeneities at 59% of over 2000 stations), such that homogenization — adjusting series for non-climatic breakpoints using, for example, geostrophic wind speed (the wind implied by surface-pressure gradients, which reflects large-scale forcing only) as a homogeneity reference — reduces China's 1970–2017 SWS trend from −0.15 to −0.05 m s⁻¹ per decade, and a follow-up global homogenization of more than 3600 stations (1973–2022) finds inhomogeneities at about 35% of stations and halves the Asian and European trends. Yan, Wu, Chao and Sun (2026) further show that 60.5% of Chinese stations' discontinuities are tied to relocations and that the apparent post-2005 recovery largely disappears after homogenization.

Regional and modeling studies within the lineage include Han, Tang, Zhang, Xu and Kou (2016) on irrigated agriculture around stations in Northwest China; Li, Song, Ma, Xiao, Wang and Chen (2018) on urbanization in East China using OMR (observed decline −0.16 m s⁻¹ per decade over 1991–2015, with no trend in the ERA-Interim reanalysis, and weakening scaling with urbanization rate); Zha, Zhao, Wu and Zhang (2019) using the WRF (Weather Research and Forecasting) regional weather model to quantify land-use effects (about −0.17 m s⁻¹ between 1980s and 2010s land cover, with strong regional heterogeneity); Minola, Reese, Lai, Azorin-Molina, Guijarro, Son and Chen (2022) on Sweden, where the cessation of stilling followed North Atlantic Oscillation (a large-scale sea-level-pressure seesaw between the Icelandic low and the Azores high that steers Atlantic storm tracks) and cyclone changes while the reanalysis misses the forest-cover-related slowdown; Miao et al. (2024) linking the Northeast China wind reversal to aerosol emission changes; and Xu and Zhang (2025), whose attribution for daily maximum wind over northern China gives trend contributions of 81.6% from roughness versus 14.2% from circulation but variance contributions of 71.6% from circulation versus 10.8% from roughness — showing that the quantitative partition depends on the wind metric and statistic chosen. The review of Zha, Zhao, Wu and Shen (2021) states the field's consensus position: the mechanisms (pressure-gradient weakening from large-scale ocean–atmosphere circulation versus roughness increase from land-use and cover change) are qualitatively established, but the contributions of the different forcing factors remain unknown.

Problem Statement

Provide a quantitative, uncertainty-quantified attribution of the observed multi-decadal decline in near-surface wind speed and wind-energy density over land — in the first instance the declines documented since the 1960s–1980s, as in the source case of China for 1960–1999 — separating the contributions of: (i) local-to-regional anthropogenic land-surface modification (urbanization, afforestation and vegetation growth, irrigation and agricultural expansion, which raise surface roughness); and (ii) large-scale atmospheric circulation variability and climate change (including decadal ocean–atmosphere oscillations and changes in the pressure-gradient force). Because later homogenization studies showed that non-climatic observational inhomogeneity (station relocations, anemometer replacement and instrument automation) contaminating the raw station record is of the same order as the climatic signal itself, the attribution must additionally quantify this inhomogeneity as a third component of the partition: the reported shares are shares of the raw observed trend, and the three components — land-surface modification, large-scale atmospheric circulation, and non-climatic observational inhomogeneity — must together account for it. This is the source question — the partition posed for China by Li et al. (2008) — now necessarily including the inhomogeneity accounting. Generalization beyond the source case (for example to the northern mid-latitudes as in Vautard et al. 2010), the sensitivity of the partition to the attribution method chosen (observation-minus-reanalysis differencing, homogeneity-adjusted observational analysis, geostrophic-wind or pressure-gradient-based diagnostics, regional or global model experiments) and to the wind metric chosen (mean versus strong or maximum wind, versus wind power density), and the implications of the attribution for wind-energy planning (for example the siting of afforestation or of urban growth), remain documented complications and motivations of the open problem rather than required deliverables. The partition itself remains required for both near-surface wind speed and wind-energy density; only its dependence on the attribution method and on the wind metric is optional.

References

  1. Yan Li, Yuan Wang, HuiYun Chu, JianPing Tang (2008), 'The climate influence of anthropogenic land-use changes on near-surface wind energy potential in China', Chinese Science Bulletin 53(18):2859–2866, DOI 10.1007/s11434-008-0360-z, https://doi.org/10.1007/s11434-008-0360-z
  2. Eugenia Kalnay, Ming Cai (2003), 'Impact of urbanization and land-use change on climate', Nature 423:528–531, DOI 10.1038/nature01675, https://doi.org/10.1038/nature01675
  3. Robert Vautard, Julien Cattiaux, Pascal Yiou, Jean-Noël Thépaut, Philippe Ciais (2010), 'Northern Hemisphere atmospheric stilling partly attributed to an increase in surface roughness', Nature Geoscience 3:756–761, DOI 10.1038/ngeo979, https://doi.org/10.1038/ngeo979
  4. Songjun Han, Qiuhong Tang, Xuezhen Zhang, Di Xu, Lihang Kou (2016), 'Surface wind observations affected by agricultural development over Northwest China', Environmental Research Letters 11:054014, DOI 10.1088/1748-9326/11/5/054014, https://doi.org/10.1088/1748-9326/11/5/054014
  5. Jian Wu, Jinlin Zha, Deming Zhao, Qidong Yang (2018), 'Changes in terrestrial near-surface wind speed and their possible causes: an overview', Climate Dynamics 51(5–6):2039–2078, DOI 10.1007/s00382-017-3997-y, https://doi.org/10.1007/s00382-017-3997-y
  6. Zhengquan Li, Lili Song, Hao Ma, Jingjing Xiao, Kuo Wang, Lian Chen (2018), 'Observed surface wind speed declining induced by urbanization in East China', Climate Dynamics 50(3–4):735–749, DOI 10.1007/s00382-017-3637-6, https://doi.org/10.1007/s00382-017-3637-6
  7. Zhenzhong Zeng, Shilong Piao, Laurent Z X Li, Philippe Ciais, Yue Li, Xitian Cai, Long Yang, Maofeng Liu, Eric F Wood (2018), 'Global terrestrial stilling: does Earth's greening play a role?', Environmental Research Letters 13:124013, DOI 10.1088/1748-9326/aaea84, https://doi.org/10.1088/1748-9326/aaea84
  8. Zhenzhong Zeng, Alan D. Ziegler, Timothy Searchinger, Long Yang, Anping Chen, Kunlu Ju, Shilong Piao, Laurent Z. X. Li, Philippe Ciais, Deliang Chen, Junguo Liu, Cesar Azorin-Molina, Adrian Chappell, David Medvigy, Eric F. Wood (2019), 'A reversal in global terrestrial stilling and its implications for wind energy production', Nature Climate Change 9:979–985, DOI 10.1038/s41558-019-0622-6, https://doi.org/10.1038/s41558-019-0622-6
  9. Zhengtai Zhang, Kaicun Wang (2020), 'Stilling and Recovery of the Surface Wind Speed Based on Observation, Reanalysis, and Geostrophic Wind Theory over China from 1960 to 2017', Journal of Climate 33(10):3989–4008, DOI 10.1175/jcli-d-19-0281.1, https://doi.org/10.1175/jcli-d-19-0281.1
  10. Jinlin Zha, Deming Zhao, Jian Wu, Pengwei Zhang (2019), 'Numerical simulation of the effects of land use and cover change on the near-surface wind speed over Eastern China', Climate Dynamics 53(3–4):1783–1803, DOI 10.1007/s00382-019-04737-w, https://doi.org/10.1007/s00382-019-04737-w
  11. Jinlin Zha, Deming Zhao, Jian Wu, Cheng Shen (2021), 'Terrestrial Near-Surface Wind Speed Variations in China: Research Progress and Prospects', Journal of Meteorological Research 35(3):537–556, DOI 10.1007/s13351-021-0143-x, https://doi.org/10.1007/s13351-021-0143-x
  12. Qing Yang, Mingxing Li, Ziqing Zu, Zhuguo Ma (2021), 'Has the stilling of the surface wind speed ended in China?', Science China Earth Sciences 64(7):1036–1049, DOI 10.1007/s11430-020-9738-4, https://doi.org/10.1007/s11430-020-9738-4
  13. Lorenzo Minola, Heather Reese, Hui-Wen Lai, Cesar Azorin-Molina, Jose A. Guijarro, Seok-Woo Son, Deliang Chen (2022), 'Wind stilling-reversal across Sweden: The impact of land-use and large-scale atmospheric circulation changes', International Journal of Climatology 42(2):1049–1071, DOI 10.1002/joc.7289, https://doi.org/10.1002/joc.7289
  14. Zhengtai Zhang, Kaicun Wang (2023), 'Homogenization of Observed Surface Wind Speed Based on Geostrophic Wind Theory over China from 1970 to 2017', Journal of Climate 36(11):3667–3679, DOI 10.1175/jcli-d-22-0445.1, https://doi.org/10.1175/jcli-d-22-0445.1
  15. Haozeyu Miao, Haiming Xu, Kai Yang, Haosu Tang, Jiechun Deng, Meng Xu, Guicai Ning, Gang Huang (2024), 'The reversal of surface wind speed trend in Northeast China: impact from aerosol emissions', Climate Dynamics 63(1):43, DOI 10.1007/s00382-024-07544-0, https://doi.org/10.1007/s00382-024-07544-0
  16. Chang-Ai Xu, Zhengtai Zhang (2025), 'The Impact of Atmospheric Circulation and Surface Roughness on Changes in Maximum Surface Wind Speed: A Preliminary Study over Northern China', Journal of Applied Meteorology and Climatology 64(11):1651–1664, DOI 10.1175/jamc-d-25-0083.1, https://doi.org/10.1175/jamc-d-25-0083.1
  17. Zhengtai Zhang, Kaicun Wang (2025), 'Homogenization of Surface Wind Speed Over Global Land From 1973 to 2022', Journal of Geophysical Research: Atmospheres 130(11):e2024JD042939, DOI 10.1029/2024jd042939, https://doi.org/10.1029/2024jd042939
  18. Yan Yan, Jia Wu, Qingchen Chao, Ying Sun (2026), 'Overestimation of the recent observed near-surface wind speed recovery in China', npj Climate and Atmospheric Science 9:49, DOI 10.1038/s41612-026-01322-x, https://doi.org/10.1038/s41612-026-01322-x