1. Introduction
Near-surface (10 m) wind speed (NWS) is one of the key variables in climate studies; its changes have substantial impacts on human society and the natural environment (e.g., Pryor et al. 2006). The long-term slowdown of NWS can lead to large losses in the wind power industry (Vautard et al. 2010; Pryor and Barthelmie 2011; Sterl et al. 2015). The intensification of NWS may exacerbate soil erosion, which generates more severe dust storms (Alizadeh-Choobari et al. 2014; Guan et al. 2017; Wang et al. 2017). Moreover, the NWS changes could affect the agriculture and ecosystem, as winds play a critical role in hydrological cycles through governing the atmospheric evaporation demand (Rayner 2007; McVicar et al. 2012a,b). Over the oceans, changes in NWS are important drivers for the oceanic circulations (Zhai et al. 2012), which modulate heat and moisture fluxes at the air–sea interface (Renault et al. 2017). In addition, persistent strengthening of the ocean NWS is likely to destroy coastal infrastructure (Lyddon et al. 2019). For all these reasons, it is important to understand the changes in global NWS as well as their associated physical mechanisms in order to be able to adapt to the expected new circumstances.
Previous studies have indicated that global NWS over land has weakened significantly in the past few decades, a phenomenon known as “terrestrial stilling” (e.g., Vautard et al. 2010; Brázdil et al. 2017; Kramm et al. 2019). For example, using a new dataset consisting of 652 stations, Guo et al. (2011) stated that the average rate of decrease in annual mean NWS over China was −0.18 m s−1 decade−1 during 1969–2005. By reviewing 148 studies dealing with NWS trends from across the globe, McVicar et al. (2012a) reported an average decline of terrestrial NWS in −0.14 m s−1 decade−1 for more than 30 years. Nevertheless, the Northern Hemisphere (NH) terrestrial NWS has shown a recovery from this decline during recent years. Kim and Paik (2015) found that the decreasing trends of NWS over South Korea ceased and became unclear after 2003. Azorin-Molina et al. (2018a) reported a break in the wind stilling over Saudi Arabia during 2002–13. Zhang and Wang (2020) also suggested a weak recovery of NWS over China from 2005 to 2017. Based on in situ observations, Zeng et al. (2019) illustrated that the break in terrestrial stilling existed across the NH, especially in Europe, East Asia, and North America, becoming prominent since around 2010.
On the other hand, global NWS over the ocean areas has been found to intensify over the past few decades. Using the Special Sensor Microwave/Imager (SSM/I), Wentz et al. (2007) reported an increasing NWS trend of +0.08 m s−1 decade−1 over the global oceans during 1987–2006. Tokinaga and Xie (2011) documented an increase in NWS of the same magnitude using adjusted ship-based anemometer readings, and +0.13 m s−1 decade−1 (1988–2008) using SSM/I. Moreover, Swart and Fyfe (2012), Zieger et al. (2014), and Zheng et al. (2016) all reported that the marine NWS trends were not evenly distributed in the oceans, with more noticeable positive trends over the Pacific low-latitude waters than the waters of higher latitudes. Consistently, Young and Ribal (2019) also analyzed the ocean NWS trends over 33-yr period from 1985 to 2018 using satellite altimeter observations, and found that the largest increases in ocean NWS occurred in the Southern Hemisphere (SH).
It is still unclear what causes the terrestrial stilling and its recent reversal, as well as the ocean wind acceleration, although a number of possible causes have been proposed and studied at regional scales. For example, the break in terrestrial stilling is proposed to be linked to the phase changes in, for example, the North Atlantic Oscillation (Azorin-Molina et al. 2018a) and the Pacific decadal oscillation (PDO) (e.g., Zeng et al. 2019). Besides, instrumental issues such as the aging and renewal of anemometer could also lead to the drift of NWS trends (Azorin-Molina et al. 2018b). On global scales, Vautard et al. (2010), Wever (2012) and Z. Zhang et al. (2019) proposed that the terrestrial stilling could be partly attributed to the increase in surface roughness (e.g., forest growth, land use changes, and urbanization). Nevertheless, using in situ observations and model simulations, Zeng et al. (2018) argued that the land surface greening (i.e., enhanced vegetation leaf area index) had limited impacts on the reduction in land NWS, implying that there could be additional physical drivers that modulate the changes in global NWS.
Other possible factors driving global NWS changes include the variations of large-scale atmospheric circulation, such as the Hadley cell and midlatitude westerly jet stream, which are influenced by both the tropical and polar climate change (Staten et al. 2018). For example, the Arctic has experienced a faster warming compared to the lower latitudes, which leads to a weakened equator-to-pole air temperature gradient and associated weakening of the midlatitude westerly jet stream (e.g., Barnes and Polvani 2013; Coumou et al. 2015). Meanwhile, on the tropical side, the Hadley cell has been reported to intensify significantly (Sohn and Park 2010; Hu et al. 2011) and to expand poleward (Hu and Fu 2007; Seidel et al. 2008) during recent decades. Although the polar warming may lead to a decrease in midlatitude NWS (Francis et al. 2017), the Hadley cell strengthening tends to enhance the trade winds and westerly jet streams (Webster 2004; Ceppi and Hartmann 2013; Deng et al. 2018a,b). This phenomenon is described as a termed as “tug of war” on the midlatitude winds (e.g., Perlwitz 2011). It should be noted that the response of the Hadley cell to a warming climate is asymmetric in the two hemispheres, which is more likely to strengthen in the colder hemisphere (Hack et al. 1989; McGee et al. 2014). Given that the land surface warms faster than the ocean surface, the uneven distribution of global land–sea cover may cause stronger warming in the NH than the SH (Kang et al. 2015). This may result in an intensified Hadley cell in the SH and a more noticeable increase in the SH NWS.
So far, the connections between global NWS trends and the Hadley cell and their relationships with anthropogenic forcing have been rarely studied. It is uncertain whether the changes in Hadley cell could lead to the asymmetric NWS trends in the two hemispheres. If so, did the relationships between global NWS and large-scale atmospheric circulation experience abrupt changes before and after 2010? What factors could drive the multidecadal changes in global NWS? To answer these questions, we first examine the trends in global NWS during the periods 1980–2010 and 2010–19 using eight reanalysis datasets. Then, we investigate the impacts of anthropogenic forcing on global NWS using global climate model simulations. The remainder of this paper is organized as follows: The data and method used in this study are described in section 2. The spatial pattern and temporal variations of reanalyzed NWS are presented in section 3. Model simulations are depicted in section 4. In section 5, we discuss the possible causes for global NWS changes, followed by a summary and discussion in section 6.
2. Data and method
a. Reanalysis datasets
Compared to ground-based observations, reanalysis datasets have a few advantages, such as complete spatial coverage and consistent temporal resolution, that are suitable for analyzing the global NWS. Previous studies have fully assessed the reliability of twentieth-century historical reanalyses (e.g., Wohland et al. 2019, 2020) and modern-era reanalyses (e.g., Torralba et al. 2017; Ramon et al. 2019) in representing the NWS from in situ observations. The performances of different reanalysis products in representing the observed NWS trends are strongly dependent on regions and time scales. For example, Miao et al. (2020) reported that the Japanese 55-year Reanalysis (JRA-55) can well represent the NWS trends in Asia; the Climate Forecast System Reanalysis (CFSR) has good performance in Europe; the Modern-Era Retrospective Analysis for Research and Applications version 2 (MERRA-2) can consistently reproduce the observed NWS in central North America. Yu et al. (2019) also stated that all of the NCEP–DOE II, MERRA-2, JRA-55, the ECMWF interim reanalysis (ERA-Interim, herein ERA-I) and the National Oceanic and Atmospheric Administration’s Twentieth Century Reanalysis (NOAA-20C) can reasonably reproduce the spatial pattern of linear trends in observed NWS over China. In an effort to identify which reanalysis products best represent the NWS, Ramon et al. (2019) analyzed five state-of-the-art global reanalyses, and revealed that the ERA5 datasets offer the best agreement with the observed winds on daily and interannual variability.
A reanalysis dataset that stands out in one region may have worse performance in other regions. Therefore, this study analyzes the global NWS changes, not merely based on one specific reanalysis dataset, but based on the multiple reanalysis sources. There are eight reanalysis datasets used in this study: the NOAA-20C over 1900–2015 (Slivinski et al. 2019), the European Centre for Medium-Range Weather Forecasts’ twentieth-century reanalysis (ERA-20C) over 1900–2010 (Poli et al. 2016), JRA-55 during 1958–2019 (Kobayashi et al. 2015), MERRA-2 over 1980–2019 (Gelaro et al. 2017), NCEP–DOE-2 (Kanamitsu et al. 2002), ERA-I (Dee et al. 2011), ERA5 (Hersbach et al. 2020), and CFSR (Saha et al. 2010) over 1979–2019. In addition, we have examined observed NWS trends to assess the reliability of reanalysis products (Fig. S1 in the online supplemental material). The station observations are retrieved from the Global Surface Summary of the Day (GSOD; 28 149 stations) database (ftp://ftp.ncdc.noaa.gov/pub/data/gsod; last accessed on 1 August 2020). The original records were subjected to extensive quality control procedures and the hourly observations were processed into mean daily values (e.g., Zeng et al. 2019). We selected 2912 stations in the NH, where wind observations are complete for 1979–2019. Annual mean NWS were derived by averaging daily observations. Linear trends were calculated for 1980–2010 and 2010–19, respectively.
We also analyze other atmospheric variables, such as three-dimensional velocity, air temperature, and geopotential height at multiple pressure levels, as well as 2-m air temperatures at surface level, to explore the possible causes for global NWS changes. These atmospheric variables are primarily from ERA5. The NOAA interpolated monthly outgoing longwave radiation (OLR) (Liebmann and Smith 1996) is used as a proxy of tropical convection, which has a horizontal resolution of 2.5° × 2.5° for the period of 1974–2019. The OLR is a useful tool to diagnose the tropical convection, where positive anomalies of OLR mean decreased cloud cover and weakened convection, and vice versa. The NOAA extended reconstructed monthly sea surface temperature (SST) V5 data are also utilized to investigate the SST anomalies associated with global NWS changes (Huang et al. 2017), with a horizontal resolution of 2° × 2° for the period 1854–2019. Both the OLR and SST datasets are provided by the NOAA/OAR/ESRL PSL, Boulder, Colorado, from their website https://www.esrl.noaa.gov/psd/. In addition, we also explore the linkages between recent reversals of global NWS trends and the PDO. The PDO index can be obtained at https://ds.data.jma.go.jp/tcc/tcc/products/elnino/decadal/annpdo.txt.
b. Outputs from model simulations
The historical simulations from phase 6 of the Coupled Model Intercomparison Projection (CMIP6) are applied in this study to assess the relationships between global NWS changes and anthropogenic forcing. The CMIP6 historical experiments are designed to simulate the observed variables in climate system from 1850 to near present driven by historical all/individual forcings, based on observed records such as emissions of greenhouse gases (GHGs), global gridded land-use changes, natural variability (solar forcing and volcanic eruption), and stratospheric aerosols. Detailed information of the CMIP6 experimental design and organization can be found in Eyring et al. (2016). Although there are many climate models participating in the CMIP6, there are only three models that have generated at least 10 ensemble members for the experiments used in this study. These models include the Canadian Earth System Model version 5 (CanESM5) (Swart et al. 2019), the Centre National de Recherches Météorologiques Coupled Model for phase 6 (CNRM-CM-6-1) (Voldoire et al. 2019), and version 6 of the Institute Pierre–Simon Laplace (IPSL) climate model (IPSL-CM6A-LR) (Boucher et al. 2018). We note that, although the coupled models may have underestimated the NWS trends (Jiang et al. 2017; Krishnan and Bhaskaran 2020), they are able to simulate the signs of the NWS trends. The CMIP6 model historical simulations provide various types of experiments, driven by individual forcings, which are useful for the attribution analysis.
The relationships between Arctic warming amplification and lower-latitude NWS are analyzed based on the model simulations from the Polar Amplification Model Intercomparison Project (PAMIP) (Smith et al. 2019). The PAMIP model simulations are forced with different combinations of SST and/or sea ice concentration (SIC). For example, the pdSST-piArcSIC experiments represent atmosphere-only time slice forced by climatological monthly mean sea surface temperature (SST) for the present day and preindustrial (1850) Arctic SIC. The experiments used in this study include the pdSST-pdSIC and pdSST-piArcSIC, which allow us to examine the impacts of the Arctic SIC changes on the local air temperatures and remote atmospheric circulation. The simulations begin on 1 April and run for 14 months, where the first 2 months are ignored for the initial model spinup. Therefore, the time period analyzed in this study is from 1 June 2000 to 1 June 2001. To be consistent with the CMIP6 models, the PAMIP model we selected is the CanESM5. There are also 10 ensemble members for each of the PAMIP experiments.
c. Methods
For comparison purposes, all of the above datasets (reanalyses and model simulations) are interpolated into 2.5° × 2.5° (144 × 73) grid resolution using a bilinear interpolation method. The temporal resolution for model simulations and reanalysis products used in this study is monthly. The linear trends in near-surface NWS are estimated by the ordinary regression method. For independent variable (X) and dependent variable (Y), their relationship can be expressed as Y = BX + A. The regression coefficient B (the slope) indicates the linear rate of change of the NWS. The unit of NWS trend used in this study is meters per second per decade (m s−1 decade−1). Positive values of B correspond to increasing NWS, whereas negative values denote decreasing NWS. The statistical significance of the trends is tested using Student’s t test (e.g., Deng et al. 2019, 2020). Trends are calculated from the anomaly series (i.e., deviation from the 1980–2010 mean, which is the common period of the eight reanalysis datasets and model simulations). Additionally, to better understand the NWS trends, we separate the research domain into four regions: NH land, NH ocean, SH land, and SH ocean. In this study, the NH and SH are simply defined as 0°–70°N and 70°S–0°, respectively, because 1) the uncertainties of NWS trends over the polar region are relatively larger than middle and low latitudes (<70°) and 2) the NWS trends and their recent reversals are more noticeable over the latitudes of 70°S–70°N. Over each region, we calculate the NWS trends during 1980–2010 and 2010–19, respectively.
3. Global NWS trends from reanalysis datasets
First, we present the trends in annual mean NWS during the common period (1980–2010) of the eight reanalysis datasets. As shown in Fig. 1, the results from different reanalysis datasets show general agreement on the spatial patterns of NWS trends, with significant (p < 0.1) negative trends over land and noticeable positive trends over the oceans, with the latter being about twice as strong as the former. More specifically, the negative NWS trends are found across the NH land, particularly over western Europe and East Asia. Besides, negative trends are also observed over the NH oceans, such as the North Atlantic and the North Pacific, with the largest negative trend of about −0.20 m s−1 decade−1. In contrast, the increase in NWS mainly appears over the SH, especially over the eastern Pacific, the subtropics (30°S–0°), and the sub-Antarctic regions (around 60°S), where the maximum positive trends exceed +0.35 m s−1 decade−1. In addition to these similarities, there also exist a few discrepancies in trends among the eight reanalysis datasets. For example, the CFSR reveals strong negative NWS trends over the equatorial central Pacific, whereas the other datasets show positive trends. Over the eastern Pacific, NOAA-20C, ERA-20C, and NCEP–DOE-2 show large negative NWS trends, while such trends in the other datasets are much weaker. The weakening of NH land NWS and the intensification of SH ocean NWS, as revealed by Fig. 1, are basically consistent with the findings of previous studies based on in situ observations (Young and Ribal 2019; Zeng et al. 2019).
Trends in global near-surface (10 m) wind speed (NWS) (unit: m s−1 decade−1) over their common period (1980–2010), for (a) NOAA-20C, (b) ERA-20C, (c) JRA-55, (d) NCEP–DOE-2, (e) ERA5, (f) ERA-I, (g) CFSR, and (h) MERRA-2. The statistically significant (p < 0.1) grid points are stippled.
Citation: Journal of Climate 34, 6; 10.1175/JCLI-D-20-0310.1
Figure 2 presents the temporal variations of annual mean NWS, which are averaged over the NH land, NH ocean, SH land, and SH ocean, respectively. In general, relatively large discrepancies in NWS series existed before 1980. From 1980 to 2010, the NWS series from the various datasets have very similar magnitudes, showing general agreement on the year-to-year variability. The NWS series from the CFSR is an obvious outlier from the other reanalysis products during 2010–19, which could be related to the change in system version of the CFSR in 2011 (Saha et al. 2014). As shown by Fig. 2a, NWS over the NH land experienced significant (p < 0.1) decreasing trends during 1980–2010, which has been captured by all reanalysis datasets, with an average trend of −0.02 m s−1 decade−1. During 2010–19, most reanalysis datasets (except for the CFSR) revealed positive trends in NH land NWS, recalling the recent reversal of the so-called terrestrial stilling (e.g., Zeng et al. 2019). Over the SH land (Fig. 2b), the NWS trends are simply opposite to those over the NH, experiencing increasing trends during 1980–2010 and slightly decreasing trends during 2010–19. Over the NH oceans (Fig. 2c), the NWS trends are insignificant at p < 0.1 level during the period 1980–2010. Finally, Fig. 2d shows the temporal variation of NWS over the SH oceans, with significant (p < 0.1) positive trends during 1980–2010 and significant (p < 0.1) negative trends during 2010–19, and averaged trends of +0.09 and −0.11 m s−1 decade−1, respectively. To summarize, the NWS trends over the NH land and the SH oceans are much more prominent than those over the SH land and NH oceans. In addition, the NWS trends over the NH land are opposite to those over the SH oceans, suggesting an interhemispheric asymmetry of the NWS trends over the last decades.
Annual mean NWS (unit: m s−1) over (a) the NH land, (b) the SH land, (c) the NH ocean, and (d) the SH ocean, which are expressed as anomalies from the climatology (1980–2010). Color curves indicate the results from eight reanalysis datasets, while black lines denote the linear regressions of averaged NWS during 1980–2010 (trend: r1) and 2010–19 (trend: r2). Significances of the averaged NWS trends during the two periods (p1 and p2) are shown separately in the panels.
Citation: Journal of Climate 34, 6; 10.1175/JCLI-D-20-0310.1
Figure 3 displays the zonal means of NWS trends over land and ocean during the common period 1980–2010. As shown in Fig. 3a, the negative NWS trends are mainly concentrated over the NH land (0°–70°N), whereas the positive NWS trends are found over the SH land. Over the oceans (Fig. 3b), however, negative NWS trends appear over the NH middle latitudes (30°–70°N), while large positive trends occur over the SH. Overall, there are two peaks of positive NWS trends over the SH oceans: one in the SH subtropical trade winds and the other in the sub-Antarctic westerly winds, where the NWS trends can be as high as +0.20 m s−1 decade−1. Table 1 lists the NWS trends obtained from the eight reanalysis datasets over the NH land, NH ocean, SH land, and SH ocean during the common period 1980–2010. All reanalysis datasets capture the negative NWS trends over NH land during 1980–2010, with an average magnitude of −0.02 m s−1 decade−1. It is interesting that the decreasing NWS trends over NH land from modern-era reanalysis datasets seem to be larger than those from the twentieth-century reanalyses, suggesting that the modern-era reanalysis products better represent the phenomenon of “terrestrial stilling” during 1980–2010. By comparison, most reanalysis datasets (except for the CFSR) have revealed significant (p < 0.1) positive NWS trends over the SH oceans. The NWS trends over the SH land and NH ocean are less consistent among the reanalyses. For example, JRA-55, MERRA-2, and CFSR indicate negative NWS trends over SH land, whereas the others show positive trends. For the NH oceans, there are four reanalysis datasets displaying positive NWS trends, while negative NWS trends are found in the other four reanalysis products.
Zonal mean of NWS trends (unit: m s−1 decade−1) for (a) land and (b) ocean, where color curves denote the trends from the eight reanalysis dataset and shadings indicate their average. (c) Ratios of land and ocean areas at different latitudes.
Citation: Journal of Climate 34, 6; 10.1175/JCLI-D-20-0310.1
Trends in near-surface (10 m) wind speed (NWS) (unit: m s−1 decade−1) over land and ocean in the Northern (0°–70°N) and Southern (70°S–0°) Hemispheres, obtained from eight reanalysis datasets during their common period (1980–2010). The trends significant at levels of p < 0.1 are shown in boldface.
Global NWS trends have experienced abrupt changes since 2010. Figure 4 further displays the zonal mean of NWS trends over land and ocean during 2010–19. Unlike the period during 1980–2010, the NWS trends during 2010–19 were reversed in both hemispheres. Over the land (Fig. 4a), NWS over the NH showed positive trends, while the NWS over the SH were characterized by negative trends. Similarly, over the oceans (Fig. 4b), the NH NWS experienced increasing trends, while the SH subtropical trade winds showed noticeable decreasing trends. The changes in NWS trends over land and the oceans at different latitudes suggest the breaks in the wind stilling/wind strengthening, which are consistent with previous studies based on station observations (e.g., Azorin-Molina et al. 2018b; Zeng et al. 2019). The possible causes of the NWS changes and their recent reversals are further discussed in section 5.
It must be highlighted that although the sign of NWS trends obtained from the reanalysis products is basically consistent with that estimated from the station observations, their magnitudes of the trends are much smaller than those from the station observations. As shown in Figs. S1a and S1b, the observed NWS trends during 1980–2010 are as high as −0.40 m s−1 decade−1, while the trends from the merged reanalysis datasets are around −0.05 m s−1 decade−1. The NH mean reanalyzed NWS trend over land is less than −0.03 m s−1 decade−1, which is smaller than the observed mean trend of around −0.10 m s−1 decade−1 as revealed by Zeng et al. (2019). By comparison, the reanalyzed NWS trends are more similar to the observed ones for 2010–19 than the earlier period (Figs. S1c,d), both of which show a trend of more than +0.30 m s−1 decade−1. We note that despite the shortcomings of the reanalysis products (underestimation of the NH terrestrial stilling for 1980–2010), they have a few advantages, such as the complete spatial coverage and consistent temporal resolution. Actually, the global reanalysis datasets have mostly represented the signs of NWS trends correctly across the globe before and after 2010. Therefore, the reanalysis products are still considered useful for wind studies especially on global scales. In particular, due to improved quality in recent decades, the reanalysis products may be closer to reality than before.
4. NWS trends from CMIP6 historical simulations
Figure 5 shows the changes in NWS simulated by the CMIP6 models driven by historical all forcings. Generally, the CMIP6 models have successfully reproduced the significant (p < 0.1) decreasing trends over the NH land (Fig. 5a) and the significant (p < 0.1) increasing trends over the SH oceans (Fig. 5d) during 1980–2010. Meanwhile, the CMIP6 models have also simulated noticeable upward NWS trends over the SH land (Fig. 5b) and weak negative NWS trends over the NH oceans (Fig. 5c). The magnitudes of the CMIP6 simulated NWS trends are smaller than the reanalyzed trends, likely due to the multimodel mean that reduces the atmospheric internal variability (Deser et al. 2012).
Annual mean NWS (unit: m s−1) from the CMIP6 historical all-forcing simulations and JRA-55 over (a) the NH land, (b) the SH land, (c) the NH ocean, and (d) the SH ocean. Blue and red curves represent the multimodel mean (MMM), whereas the shadings denote the intermodel spreads. Black lines denote the linear regressions of MMM NWS during 1980–2010, where the trends (r) and their significance levels (p) are shown in corresponding panels. The JRA-55 is selected to be compared with the model simulations, as the simulation period illustrated in this figure started from 1960.
Citation: Journal of Climate 34, 6; 10.1175/JCLI-D-20-0310.1
Figure 6 illustrates the spatial patterns of CMIP6 NWS trends during 1980–2010, which are driven by the individual forcing. As seen from Fig. 6a, under the all-forcing scenario, which was forced by a combination of anthropogenic and natural forcings, significant (p < 0.1) negative trends are found over NH land, the North Atlantic, and the North Pacific, together with significant (p < 0.1) positive trends in the SH subtropical trade winds and the sub-Antarctic westerly winds. As shown in Fig. 6b, under the GHG-only forcing, negative trends in NWS occur over the NH land areas, such as Europe and East Asia. Meanwhile, significant (p < 0.1) positive trends are found over the SH subtropics and sub-Antarctic regions. Overall, the NWS trends forced by GHGs are very similar to those driven by the all-forcing scenario and those in the reanalyzed, suggesting that the NWS trends are primarily attributable to the forcing of GHGs. Figure 6c shows the spatial distributions of NWS trends due to the aerosol-only forcing, where significant (p < 0.1) decreasing trends mainly existed in the NH, particularly over the tropical North Atlantic and the North Pacific, which implies that the aerosol forcing tend to contribute to the declining of NH NWS. In contrast, under the natural-only forcing (Fig. 6d), the NWS trends simulated by the CMIP6 models are weak and insignificant at p < 0.1 over most regions, no matter over the land or oceans. The above attribution analysis demonstrates that anthropogenic forcing (e.g., emissions of GHGs and aerosols) could be important drivers for the NWS changes during 1980–2010.
Spatial distributions of annual NWS trends (unit: m s−1 decade−1) during 1980–2010, from the CMIP6 MMM for (a) all forcings, (b) GHG-only forcing, (c) aerosol-only forcing, and (d) natural-only forcing. The statistically significant (p < 0.1) grid points are stippled.
Citation: Journal of Climate 34, 6; 10.1175/JCLI-D-20-0310.1
Figure 7 shows the averaged NWS trends during 1980–2010 as well as their uncertainties, which are calculated from the multiple reanalyses and model simulations. The reanalyzed NWS over NH land showed an average trend of −0.02 m s−1 decade−1, with a standard deviation of ±0.01 m s−1 decade−1 (Fig. 7a). By comparison, the CMIP6 all-forcing simulations produced an average trend of −0.01 m s−1 decade−1, with the same spread of ±0.01 m s−1 decade−1. We detect that the negative NWS trends over NH land are attributed to the combined effects of GHGs and aerosols, both of which have induced negative NWS trends over NH land. Nevertheless, the averaged trend in reanalyzed NWS over the SH land is +0.01 m s−1 decade−1, with a standard deviation of ±0.02 m s−1 decade−1, implying that the signs of NWS trends over the SH land are much more uncertain.
Annual NWS trends (bar; unit: m s−1 decade−1) over (a) land and (b) ocean during 1980–2010 and their uncertainties (error bars) for the eight reanalysis datasets and CMIP6 model simulations. Filled (open) rectangles indicate the trends averaged over the NH (SH). The uncertainty spread is determined by one standardized deviation of the NWS trends.
Citation: Journal of Climate 34, 6; 10.1175/JCLI-D-20-0310.1
The NWS trends over the oceans and their uncertainties are shown in Fig. 7b. Over the NH oceans, the uncertainty spreads of NWS trends are so large that the signs of NWS trends are unsure. Such issues exist not only in the reanalyzed datasets but also in the CMIP6 model simulations. In contrast, NWS over the SH oceans have shown unanimously positive trends, with averaged trends of +0.09 m s−1 decade−1 for the reanalyses and more than +0.03 m s−1 decade−1 for model simulations. In general, the CMIP6 models have underestimated the reanalyzed NWS trends, although they have correctly simulated the signs of the NWS trends, except for the NH oceans. In addition, the model simulations demonstrate that positive NWS trends over the SH ocean could be mainly attributed to the GHG forcing, whereas the aerosol forcing tend to force negative NWS trends over both hemispheres.
5. Possible causes of global NWS changes
The above analysis reveals that NWS over the NH land (SH oceans) experienced decreasing (increasing) trends during 1980–2010. Moreover, a reversal of global NWS appeared in around 2010. In this section, we further discuss the possible causes of global NWS changes, based on the ERA5 products and model simulations.
a. Intensifying NWS over the SH oceans during 1980–2010
The SH is dominated by oceans, while the NH is mainly covered by land (Fig. 3c). As land surfaces warm faster than the ocean ones, more warming is expected in the NH than the SH in a changing climate (Kang et al. 2015). Hack et al. (1989) and McGee et al. (2014) suggested that the Hadley cell is more likely to strengthen in the colder hemisphere, with an opposite response in the warmer hemisphere. That means, the changes in Hadley cell could lead to the intensification of the SH NWS.
Figure 8a presents the trends in meridional atmospheric circulation during 1980–2010. It shows that ascending trends in air motions appeared over the equatorial regions, which were connected with descending trends over the SH middle latitudes, implying an intensified Hadley cell over the SH. Correspondingly, airflows that originated from the tropics gradually accumulated over the SH middle latitudes, which led to the increases in pressure (contours) in the upper troposphere around 40°S. Due to the fluid continuity, air masses accumulated in upper troposphere were forced to sink in the SH middle latitudes. As a result, on the one hand, warmer air temperatures were found over these regions owing to a stronger adiabatic heating. One the other hand, the sinking airflow was split into two branches at the surface, one moving toward the tropics and the other moving toward high latitudes. Under the Coriolis effects, the equatorward and poleward branches eventually caused the intensification of the SH trade winds and the sub-Antarctic westerly winds, respectively.
(a) Zonal averages of trends in annual meridional and vertical (multiplied by −100) winds (vectors; unit: m s−1 decade−1), geopotential heights (contours; unit: m decade−1), and air temperatures (shading; unit: °C decade−1) during 1980–2010. (b) Spatial distributions of trends in annual OLR (shading; unit: W m−2 decade−1) and 300-hPa velocity potential (contours; unit: 104 decade−1). The trends in winds and OLR are obtained from the ERA5 and NOAA interpolated OLR datasets, respectively.
Citation: Journal of Climate 34, 6; 10.1175/JCLI-D-20-0310.1
Figure 8b displays the spatial patterns of trends in convection and 300-hPa velocity potential during 1980–2010. We can see that negative OLR trends mainly appeared over the western Pacific and the tropical west Atlantic, indicating enhanced convection over these regions. Meanwhile, decreasing trends in the velocity potential were found over the upper troposphere, implying stronger upper-level wind divergences. The enhanced tropical convection and the associated upper-level divergent winds over the west Pacific and the tropical west Atlantic acted as a driving force that led to the intensification of the SH Hadley cell, which resulted in the strengthening of the SH NWS over the past decades.
Figure 9 depicts the trends in meridional atmospheric circulation and OLR simulated by the CMIP6 model simulations. As shown in Fig. 9a, under the forcing of GHGs, upward (downward) trends appeared over the tropics (SH middle latitudes), indicating that the SH Hadley cell was intensified during 1980–2010. Correspondingly, negative trends in OLR were found over the west Pacific (Fig. 9b), suggesting stronger convection over these regions in response to the increased GHGs. The results simulated by CMIP6 models are well consistent with those from the reanalysis datasets, further increasing the confidence that the GHG forcing might affect the global NWS changes through modulating the meridional atmospheric circulation. Opposite to the GHG, the aerosol forcing tended to suppress the convection over the west Pacific and caused a weakened NWS globally, and the natural variability forcing was not able to simulate any significant (p < 0.1) trends in the meridional atmospheric circulation (figure not shown) during 1980–2010. In other words, the strengthening of SH NWS during 1980–2010 was mainly attributed to the GHG forcing through altering the SH Hadley cell.
(a) Zonal means of trends in meridional and vertical (multiplied by −100) winds (vectors; unit: m s−1 decade−1) and vertical pressure velocity (shading; unit: 10−4 Pa s−1 decade−1) for 1980–2010, which are obtained from the CMIP6 MMM simulations driven by GHG-only forcing. (b) Spatial distributions of trends in MMM OLR (unit: W m−2 decade−1).
Citation: Journal of Climate 34, 6; 10.1175/JCLI-D-20-0310.1
b. Declining NWS over the NH land during 1980–2010
Although the changes in SH Hadley cell can reasonably account for the strengthening of the SH NWS during 1980–2010, they failed to explain the declining of NWS over the NH land. Over the last decades, human land-use activities have resulted in large changes to the surface roughness, which could affect the NWS over NH land. As seen from Fig. 10a, model simulations show that the historical land-use changes could cause negative NWS trends over the NH, particularly over northern Eurasia, northern Africa, and South Asia, which induced a negative trend of more than −0.02 m s−1 decade−1 during 1980–2010. However, the land-use changes alone are not sufficient to explain the declining of NWS over the NH land. For example, over other regions, such as the United States and central Asia, the effects of land-use changes on the NWS seem to be opposite to the reanalyzed trends. Therefore, there could be additional factors that affect the NWS changes over NH land.
(a) Annual NWS trends during 1980–2010 from CMIP6 MMM simulations, driven by historical land-use- only forcing. (b),(c) The changes in PAMIP simulated annual 2-m air temperatures (T2m; unit: °C) and NWS (unit: m s−1), respectively, due to the losses of Arctic sea ice concentration (SIC) relative to the preindustrial level.
Citation: Journal of Climate 34, 6; 10.1175/JCLI-D-20-0310.1
c. Recent reversals of global NWS since 2010
The breaks in terrestrial stilling/wind acceleration were observed around 2010. As shown in Fig. 8b, the enhanced convection over the west Pacific and tropical Atlantic is followed by a strong suppressed convection over the east Pacific. The spatial distribution of convection trends is reminiscent of the Pacific multidecadal variability.
Figure 11 shows trends in global annual SSTs during 1980–2010 and 2010–19. As seen from Fig. 11a, warming SST trends during 1980–2010 appeared in the North Atlantic and North Pacific, while cooling SST trends were found in the central and eastern Pacific, which are similar to the negative phase of the PDO. During 2010–19 (Fig. 11b), however, positive SST trends were observed in the central and eastern Pacific, while negative SST trends existed in the high-latitude waters of the Pacific, implying a positive phase of the PDO. Figure 11c presents the PDO index and their 5-yr running mean over the last decades, where we detect that the PDO experienced an obvious phase change around 2010.
Trends in annual sea surface temperature (SST; unit: °C decade−1) during (a) 1980–2010 and (b) 2010–19, where the statistically significant (p < 0.1) grid points are stippled. (c) Index of the annual mean Pacific decadal oscillation (black) and its 10-yr running mean (blue).
Citation: Journal of Climate 34, 6; 10.1175/JCLI-D-20-0310.1
The phase changes in the PDO coincided well with the recent reversals of global NWS trends. More specifically, the PDO experienced a positive-to-negative phase changes during 1980–2010, corresponding to warming SST trends in the western Pacific and cooling SST trends in the central and eastern Pacific, which resulted in an intensified equatorial Pacific zonal SST gradient and enhanced convection over the western Pacific, leading to an intensified Hadley cell and strengthening NWS over the SH. Conversely, during 2010–19, the PDO underwent a negative-to-positive phase change, characterized by warming SST trends in the eastern Pacific and cooling SST trends in the western Pacific, which reduced the equatorial Pacific zonal SST gradient and suppressed the western Pacific convection. As a result, the SH Hadley cell was weakened, which might cause the break in SH wind acceleration. Moreover, during 1980–2010, warming SST trends in the North Atlantic and North Pacific led to decreased meridional temperature gradients, which could induce weakened westerly winds over the NH. During 2010–19, however, cooling SST trends in the North Atlantic and North Pacific actually increased the meridional temperature gradients that could trigger the intensification of the NH westerly winds (e.g., Coumou et al. 2015). That is, recent phase changes in the PDO may have played a crucial role in the reversals of global NWS trends.
Finally, Fig. 12 shows the zonal mean of trends in meridional atmospheric circulation and air temperature during 2010–19. It is shown that the SH Hadley cell experienced a weakening trend during this period, with strong ascending motions over the SH middle latitudes and sinking motions over the SH subtropics (Fig. 12a), which was accompanied by the suppression of the western Pacific convection (Fig. 12b). These results are consistent with the previous assumptions that the phase changes of the PDO can inhabit the western Pacific convection and weaken the SH Hadley cell, which ultimately lead to the decreases in the SH trade winds and sub-Antarctic westerly winds. By comparison, over the NH strong upward motions appeared over the NH subtropics and the sub-Arctic regions, accompanied by weak descending motions over the middle latitudes, which indicated the intensification of meridional atmospheric circulation over the NH. The descending trends over the NH middle latitudes could further induce the warm air temperature trends through the adiabatic warming processes (Fig. 12a), and consequently altered the NH meridional air temperature gradient that could reverse the NWS trends over the NH. We note that although the results are mainly obtained from the ERA5 dataset, similar changes in large-scale atmospheric circulations are also detected from the other reanalysis datasets.
As in Fig. 9, but for the period 2010–19, obtained from the ERA5 and NOAA interpolated OLR datasets.
Citation: Journal of Climate 34, 6; 10.1175/JCLI-D-20-0310.1
6. Summary and discussion
This study investigates the long-term trends in global annual near-surface wind speed (NWS), based on reanalysis datasets and global climate model simulations. It is found that the NH land NWS experienced significant (p < 0.1) decreasing trends during 1980–2010, whereas the SH ocean NWS showed noticeable increasing trends. The average trend in NWS over the NH land during 1980–2010 is −0.02 m s−1 decade−1, while the average trend over the SH ocean is +0.09 m s−1 decade−1. Both of these mean trends are robust among the reanalysis and simulated datasets. The breaks in the NH wind stilling/SH wind strengthening appears from around 2010. That is, NWS over the NH land (SH ocean) underwent increasing (decreasing) trends during 2010–19. Figure 13 displays a schematic summary for the possible causes of global NWS changes. During 1980–2010 (Fig. 13a), the SH trade winds and sub-Antarctic westerly winds experienced intensified trends, which was caused by the enhanced Hadley cell over the SH. The enhancement of the SH Hadley cell is primarily caused by the increased GHG forcing and the negative phases of the PDO. Meanwhile, the NH land NWS experienced weakening trends, likely due to the joint impacts of vegetation/land-use changes and the accelerating Arctic warming. During 2010–19 (Fig. 13b); however, the NWS trends in the two hemispheres were reversed, which is suggested to be linked to the phase changes in the PDO and associated changes in large-scale atmospheric circulation.
Schematic diagrams of possible causes of global NWS changes during (a) 1980–2010 and (b) 2010–19.
Citation: Journal of Climate 34, 6; 10.1175/JCLI-D-20-0310.1
The CMIP6 simulations demonstrate that the GHG forcing was a major force for global NWS changes during 1980–2010 via a strengthened Hadley cell over the SH and accelerated Arctic warming over the NH. Nevertheless, since 2010, the phase changes of PDO could have played a more dominant role in modulating the NWS trends than the GHG forcing, which reasonably explains the reversed trends in global NWS. In addition, we note that this study discusses the causes of global NWS changes, mainly from the perspective of anthropogenic forcing and the natural multidecadal mode in SST. Other factors such as the discontinuity in data assimilating system and the aging and renewal of anemometers may have also affected the changes in NWS; these are not explicitly explored in this study.
Over the past decades, the weakening of NH terrestrial winds triggered concerns about whether we can rely on wind power as a main renewable energy. The recent reversal of wind stilling, as revealed by previous literature and this study, has provided new hope for the wind power industry (Zeng et al. 2019). Nevertheless, the strengthened terrestrial NWS could also bring other issues, such as more severe soil erosion (e.g., G. Zhang et al. 2019). Therefore, it is necessary to monitor and assess the future changes in NWS, which could aid the decision making of public and private sectors to better adapt to the new circumstances.
Acknowledgments
We greatly appreciate the three anonymous reviewers for their careful reading of our manuscript and their many constructive comments and suggestions. We acknowledge supports from the Swedish Research Council (VR-2017-03780 and VR-2019-03954), Ramon y Cajal fellowship (RYC-2017-22830), RTI2018-095749-A-I00 (MCIU/AEI/FEDER, UE), the Second Tibetan Plateau Scientific Expedition and Research Program (Grant 2019QZKK0606), and Swedish MERGE and BECC. The computational resources were provided by the Swedish National Infrastructure for Computing (SNIC) at the National Supercomputer Centre in Sweden (NSC).
Data availability statement
The reanalysis datasets that support the findings of this study are openly available at https://www.ecmwf.int/en/forecasts/datasets/browse-reanalysis-datasets/ for the ERA-20C, ERA-I, and ERA5; at https://rda.ucar.edu/ for the JRA-55 and CFSR; at https://psl.noaa.gov/data/gridded/ for the NOAA-20C, NCEP/DOE-2, SST, and OLR; and at https://disc.gsfc.nasa.gov/datasets?page=1&keywords=MERRA-2 for the MERRA-2. The last time we accessed these datasets was 1 July 2020. In situ observations for Fig. S1 are available from the Global Surface Summary of the Day database. The model simulation datasets can be publicly downloaded from a website at https://esgf-node.llnl.gov/search/cmip6/.
REFERENCES
Alizadeh-Choobari, O., P. Zawar-Reza, and A. Sturman, 2014: The “wind of 120 days” and dust storm activity over the Sistan Basin. Atmos. Res., 143, 328–341, https://doi.org/10.1016/j.atmosres.2014.02.001.
Azorin-Molina, C., S. Rehman, J. A. Guijarro, T. R. McVicar, L. Minola, D. Chen, and S. M. Vicente-Serrano, 2018a: Recent trends in wind speed across Saudi Arabia, 1978–2013: A break in the stilling. Int. J. Climatol., 38, e966–e984, https://doi.org/10.1002/joc.5423.
Azorin-Molina, C., J. Asin, T. R. McVicar, L. Minola, J. I. Lopez-Moreno, S. M. Vicente-Serrano, and D. Chen, 2018b: Evaluating anemometer drift: A statistical approach to correct biases in wind speed measurement. Atmos. Res., 203, 175–188, https://doi.org/10.1016/j.atmosres.2017.12.010.
Barnes, E. A., and L. Polvani, 2013: Response of the midlatitude jets, and of their variability, to increased greenhouse gases in the CMIP5 models. J. Climate, 26, 7117–7135, https://doi.org/10.1175/JCLI-D-12-00536.1.
Boucher, O., S. Denvil, A. Caubel, and M. A. Foujols, 2018: IPSL IPSL-CM6A-LR model output prepared for CMIP6 CMIP. Earth System Grid Federation, https://doi.org/10.22033/ESGF/CMIP6.1534.
Brázdil, R., A. Valík, P. Zahradníček, L. Řezníčková, R. Tolasz, and M. Možný, 2017: Wind-stilling in the light of wind speed measurements: The Czech experience. Climate Res., 74, 131–143, https://doi.org/10.3354/cr01492.
Ceppi, P., and D. Hartmann, 2013: On the speed of the eddy-driven jet and the width of the Hadley cell in the Southern Hemisphere. J. Climate, 26, 3450–3465, https://doi.org/10.1175/JCLI-D-12-00414.1.
Cohen, J., and Coauthors, 2020: Divergent consensuses on Arctic amplification influence on midlatitude severe winter weather. Nat. Climate Change, 10, 20–29, https://doi.org/10.1038/s41558-019-0662-y.
Coumou, D., J. Lehmann, and J. Beckmann, 2015: The weakening summer circulation in the Northern Hemisphere mid-latitudes. Science, 348, 324–327, https://doi.org/10.1126/science.1261768.
Dee, D., and Coauthors, 2011: The ERA-Interim reanalysis: Configuration and performance of the data assimilation system. Quart. J. Roy. Meteor. Soc., 137, 553–597, https://doi.org/10.1002/qj.828.
Deng, K., S. Yang, M. Ting, Y. Tan, and S. He, 2018a: Global monsoon precipitation: Trends, leading modes, and associated drought and heat wave in the Northern Hemisphere. J. Climate, 31, 6947–6966, https://doi.org/10.1175/JCLI-D-17-0569.1.
Deng, K., M. Ting, S. Yang, and Y. Tan, 2018b: Increased frequency of summer extreme heat waves over Texas area tied to the amplification of Pacific zonal SST gradient. J. Climate, 31, 5629–5647, https://doi.org/10.1175/JCLI-D-17-0554.1.
Deng, K., S. Yang, M. Ting, P. Zhao, and Z. Wang, 2019: Dominant modes of China summer heat waves driven by global sea surface temperature and atmospheric internal variability. J. Climate, 32, 3761–3775, https://doi.org/10.1175/JCLI-D-18-0256.1.
Deng, K., S. Yang, D. Gu, A. Lin, and C. Li, 2020: Record-breaking heat wave in southern China and delayed onset of South China Sea summer monsoon driven by the Pacific subtropical high. Climate Dyn., 54, 3751–3764, https://doi.org/10.1007/s00382-020-05203-8.
Deser, C., A. S. Phillips, V. Bourdette, and H. Teng, 2012: Uncertainty in climate change projections: The role of internal variability. Climate Dyn., 38, 527–546, https://doi.org/10.1007/s00382-010-0977-x.
Eyring, V., S. Bony, G. A. Meehl, C. A. Senior, B. Stevens, R. J. Stouffer, and K. E. Taylor, 2016: Overview of the Coupled Model Intercomparison Project Phase 6 (CMIP6) experimental design and organization. Geosci. Model Dev., 9, 1937–1958, https://doi.org/10.5194/gmd-9-1937-2016.
Francis, J. A., S. J. Vavrus, and J. Cohen, 2017: Amplified Arctic warming and mid-latitude weather: New perspectives on emerging connections. Wiley Interdiscip. Rev.: Climate Change, 8, e474, https://doi.org/10.1002/wcc.474.
Gelaro, R., and Coauthors, 2017: The Modern-Era Retrospective Analysis for Research and Applications, version 2 (MERRA-2). J. Climate, 30, 5419–5454, https://doi.org/10.1175/JCLI-D-16-0758.1.
Guan, Q., X. Sun, J. Yang, B. Pan, S. Zhao, and L. Wang, 2017: Dust storms in northern China: Long-term spatiotemporal characteristics and climate controls. J. Climate, 30, 6683–6700, https://doi.org/10.1175/JCLI-D-16-0795.1.
Guo, H., M. Xu, and Q. Hu, 2011: Changes in near-surface wind speed in China: 1969–2005. Int. J. Climatol., 31, 349–358, https://doi.org/10.1002/joc.2091.
Hack, J. J., W. H. Schubert, D. E. Stevens, and H.-C. Kuo, 1989: Response of the Hadley circulation to convective forcing in the ITCZ. J. Atmos. Sci., 46, 2957–2973, https://doi.org/10.1175/1520-0469(1989)046<2957:ROTHCT>2.0.CO;2.
Hersbach, H., and Coauthors, 2020: The ERA5 global reanalysis. Quart. J. Roy. Meteor. Soc., 146, 1999–2049, https://doi.org/10.1002/qj.3803.
Hu, Y., and Q. Fu, 2007: Observed poleward expansion of the Hadley circulation since 1979. Atmos. Chem. Phys., 7, 5229–5236, https://doi.org/10.5194/acp-7-5229-2007.
Hu, Y., C. Zhou, and J. Liu, 2011: Observational evidence for the poleward expansion of the Hadley circulation. Adv. Atmos. Sci., 28, 33–44, https://doi.org/10.1007/s00376-010-0032-1.
Huang, B., and Coauthors, 2017: Extended Reconstructed Sea Surface Temperature, version 5 (ERSSTv5): Upgrades, validations, and intercomparisons. J. Climate, 30, 8179–8205, https://doi.org/10.1175/JCLI-D-16-0836.1.
Jiang, Y., X. Xu, H. Liu, X. Dong, W. Wang, and G. Jia, 2017: The underestimated magnitude and decline trend in near-surface wind over China. Atmos. Sci. Lett., 18, 475–483, https://doi.org/10.1002/asl.791.
Kanamitsu, M., W. Ebisuzaki, J. Woollen, S.-K. Yang, J. J. Hnilo, M. Fiorino, and G. L. Potter, 2002: NCEP–DOE AMIP-II Reanalysis (R-2). Bull. Amer. Meteor. Soc., 83, 1631–1644, https://doi.org/10.1175/BAMS-83-11-1631.
Kang, S. M., R. Seager, D. M. W. Frierson, and X. Liu, 2015: Croll revisited: Why is the Northern Hemisphere warmer than the Southern Hemisphere? Climate Dyn., 44, 1457–1472, https://doi.org/10.1007/s00382-014-2147-z.
Kim, J., and K. Paik, 2015: Recent recovery of surface wind speed after decadal decrease: A focus on South Korea. Climate Dyn., 45, 1699–1712, https://doi.org/10.1007/s00382-015-2546-9.
Kobayashi, S., and Coauthors, 2015: The JRA-55 reanalysis: General specifications and basic characteristics. J. Meteor. Soc. Japan, 93, 5–48, https://doi.org/10.2151/jmsj.2015-001.
Kramm, G., N. Mölders, J. Cooney, and R. Dlugi, 2019: Near-surface wind-speed stilling in Alaska during 1984–2016 and its impact on the sustainability of wind power. J. Power Energy Eng., 7, 71–124, https://doi.org/10.4236/jpee.2019.77006.
Krishnan, A., and P. K. Bhaskaran, 2020: Skill assessment of global climate model wind speed from CMIP5 and CMIP6 and evaluation of projections for the Bay of Bengal. Climate Dyn., 55, 2667–2687, https://doi.org/10.1007/s00382-020-05406-z.
Liebmann, B., and C. A. Smith, 1996: Description of a complete (interpolated) outgoing longwave radiation dataset. Bull. Amer. Meteor. Soc., 77, 1275–1277.
Lyddon, C. E., J. M. Brown, N. Leonardi, and A. J. Plater, 2019: Increased coastal wave hazard generated by differential wind and wave direction in hyper-tidal estuaries. Estuarine Coastal Shelf Sci., 220, 131–141, https://doi.org/10.1016/j.ecss.2019.02.042.
McGee, D., A. Donohoe, J. Marshall, and D. Ferreira, 2014: Changes in ITCZ location and cross-equatorial heat transport at the Last Glacial Maximum, Heinrich Stadial 1, and the mid-Holocene. Earth Planet. Sci. Lett., 390, 69–79, https://doi.org/10.1016/j.epsl.2013.12.043.
McVicar, T. R., and Coauthors, 2012a: Global review and synthesis of trends in observed terrestrial near-surface wind speeds: Implications for evaporation. J. Hydrol., 416-417, 182–205, https://doi.org/10.1016/j.jhydrol.2011.10.024.
McVicar, T. R., M. L. Roderick, R. Donohue, and T. G. Van Niel, 2012b: Less bluster ahead? Ecohydrological implications of global trends of terrestrial near-surface wind speeds. Ecohydrology, 5, 381–388, https://doi.org/10.1002/eco.1298.
Miao, H., D. Dong, G. Huang, K. Hu, Q. Tian, and Y. Gong, 2020: Evaluation of Northern Hemisphere surface wind speed and wind power density in multiple reanalysis datasets. Energy, 200, 117382, https://doi.org/10.1016/j.energy.2020.117382.
Perlwitz, J., 2011: Tug of war on the jet stream. Nat. Climate Change, 1, 29–31, https://doi.org/10.1038/nclimate1065.
Poli, P., and Coauthors, 2016: ERA-20C: An atmospheric reanalysis of the twentieth century. J. Climate, 29, 4083–4097, https://doi.org/10.1175/JCLI-D-15-0556.1.
Pryor, S. C., and R. J. Barthelmie, 2011: Assessing climate change impacts on the near-term stability of the wind energy resource over the United States. Proc. Natl. Acad. Sci. USA, 108, 8167–8171, https://doi.org/10.1073/pnas.1019388108.
Pryor, S. C., J. T. Schoof, and R. J. Barthelmie, 2006: Winds of change? Projections of near-surface winds under climate change scenarios. Geophys. Res. Lett., 33, L11702, https://doi.org/10.1029/2006GL026000.
Ramon, J., L. Lledó, V. Torralba, A. Soret, and F. Doblas-Reyes, 2019: What global reanalysis best represents near-surface winds? Quart. J. Roy. Meteor. Soc., 145, 3236–3251, https://doi.org/10.1002/qj.3616.
Rayner, D. P., 2007: Wind run changes: The dominant factor affecting pan evaporation trends in Australia. J. Climate, 20, 3379–3394, https://doi.org/10.1175/JCLI4181.1.
Renault, L., J. C. McWilliams, and S. Masson, 2017: Satellite observations of imprint of oceanic current on wind stress by air–sea coupling. Sci. Rep., 7, 17747, https://doi.org/10.1038/s41598-017-17939-1.
Saha, S., and Coauthors, 2010: The NCEP Climate Forecast System Reanalysis. Bull. Amer. Meteor. Soc., 91, 1015–1058, https://doi.org/10.1175/2010BAMS3001.1.
Saha, S., and Coauthors, 2014: The NCEP Climate Forecast System version 2. J. Climate, 27, 2185–2208, https://doi.org/10.1175/JCLI-D-12-00823.1.
Seidel, D. J., Q. Fu, W. J. Randel, and T. J. Reichler, 2008: Widening of the tropical belt in a changing climate. Nat. Geosci., 1, 21–24, https://doi.org/10.1038/ngeo.2007.38.
Slivinski, L. C., and Coauthors, 2019: Towards a more reliable historical reanalysis: Improvements for version 3 of the Twentieth Century Reanalysis system. Quart. J. Roy. Meteor. Soc., 145, 2876–2908, https://doi.org/10.1002/qj.3598.
Smith, D. M., and Coauthors, 2019: The Polar Amplification Model Intercomparison Project (PAMIP) contribution to CMIP6: Investigating the causes and consequences of polar amplification. Geosci. Model Dev., 12, 1139–1164, https://doi.org/10.5194/gmd-2018-82.
Sohn, B. J., and S. C. Park, 2010: Strengthened tropical circulations in past three decades inferred from water vapor transport. J. Geophys. Res., 115, D15112, https://doi.org/10.1029/2009JD013713.
Staten, P. W., J. Lu, K. M. Grise, S. M. Davis, and T. Birner, 2018: Re-examining tropical expansion. Nat. Climate Change, 8, 768–775, https://doi.org/10.1038/s41558-018-0246-2.
Sterl, A., A. M. R. Bakker, H. W. van den Brink, R. Haarsma, A. Stepek, I. L. Wijnant, R. C. de Winter, 2015: Large-scale winds in the southern North Sea region: The wind part of the KNMI’14 climate change scenarios. Environ. Res. Lett., 10, 035004, https://doi.org/10.1088/1748-9326/10/3/035004.
Swart, N. C., and J. Fyfe, 2012: Observed and simulated changes in the Southern Hemisphere surface westerly wind-stress. Geophys. Res. Lett., 39, L16711, https://doi.org/10.1029/2012GL052810.
Swart, N. C., and Coauthors, 2019: The Canadian Earth System Model version 5 (CanESM5.0.3). Geosci. Model Dev., 12, 4823–4873, https://doi.org/10.5194/gmd-12-4823-2019.
Tokinaga, H., and S. P. Xie, 2011: Wave and anemometer-based sea-surface wind (WASWind) for climate change analysis. J. Climate, 24, 267–285, https://doi.org/10.1175/2010JCLI3789.1.
Torralba, V., F. J. Doblas-Reyes, and N. Gonzalez-Reviriego, 2017: Uncertainty in recent near-surface wind speed trends: A global reanalysis intercomparison. Environ. Res. Lett., 12, 114019, https://doi.org/10.1088/1748-9326/aa8a58.
Vautard, R., J. Cattiaux, P. Yiou, J. N. Thépaut, and P. Ciais, 2010: Northern Hemisphere atmospheric stilling partly attributed to an increase in surface roughness. Nat. Geosci., 3, 756–761, https://doi.org/10.1038/ngeo979.
Voldoire, A., and Coauthors, 2019: Evaluation of CMIP6 DECK experiments with CNRM-CM6-1. J. Adv. Model. Earth Syst., 11, 2177–2213, https://doi.org/10.1029/2019MS001683.
Wang, R., and Coauthors, 2017: Variation of strong dust storms events in northern China during 1978–2007. Atmos. Res., 183, 166–172, https://doi.org/10.1016/j.atmosres.2016.09.002.
Webster, P. J., 2004: The elementary Hadley circulation. The Hadley Circulation: Present, Past and Future, H. F. Diaz and R. S. Bradley, Eds., Advances in Global Change Research Series, Vol. 21, Springer, 9–60, https://doi.org/10.1007/978-1-4020-2944-8_2.
Wentz, F., L. Ricciardulli, and C. Mears, 2007: How much more rain will global warming bring? Science, 317, 233–235, https://doi.org/10.1126/science.1140746.
Wever, N., 2012: Quantifying trends in surface roughness and the effect on surface wind speed observations. J. Geophys. Res., 117, D11104, https://doi.org/10.1029/2011JD017118.
Wohland, J., N. E. Omrani, D. Witthaut, and N. S. Keenlyside, 2019: Inconsistent wind speed trends in current twentieth century reanalyses. J. Geophys. Res. Atmos., 124, 1931–1940, https://doi.org/10.1029/2018JD030083.
Wohland, J., H. Bloomfield, D. Brayshaw, S. Pfenninger, and M. Wild, 2020: Understanding multidecadal variability for energy system studies: Can current 20th century reanalyses do the job? EGU General Assembly 2020, Online, EGU2020-3008, https://doi.org/10.5194/egusphere-egu2020-3008.
Young, I. R., and A. Ribal, 2019: Multiplatform evaluation of global trends in wind speed and wave height. Science, 364, 548–552, https://doi.org/10.1126/science.aav9527.
Yu, J., T. Zhou, Z. Jiang, and L. Zou, 2019: Evaluation of near-surface wind speed changes during 1979 to 2011 over China based on five reanalysis datasets. Atmosphere, 10, 804, https://doi.org/10.3390/atmos10120804.
Zeng, Z., and Coauthors, 2018: Global terrestrial stilling: Does Earth’s greening play a role? Environ. Res. Lett., 13, 124013, https://doi.org/10.1088/1748-9326/aaea84.
Zeng, Z., and Coauthors, 2019: A reversal in global terrestrial stilling and its implications for wind energy production. Nat. Climate Change, 9, 979–985, https://doi.org/10.1038/s41558-019-0622-6.
Zhai, X., H. L. Johnson, D. P. Marshall, and C. Wunsch, 2012: On the wind power input to the ocean general circulation. J. Phys. Oceanogr., 42, 1357–1365, https://doi.org/10.1175/JPO-D-12-09.1.
Zhang, G., C. Azorin-Molina, P. Shi, D. Lin, J. A. Guijarro, F. Kong, and D. Chen, 2019: Impact of near-surface wind speed variability on wind erosion in the eastern agro-pastoral transitional zone of northern China, 1982–2016. Agric. For. Meteor., 271, 102–115, https://doi.org/10.1016/j.agrformet.2019.02.039.
Zhang, Z. T., and K. C. Wang, 2020: Stilling and recovery of the surface wind speed based on observation, reanalysis, and geostrophic wind theory over China from 1960 to 2017. J. Climate, 33, 3989–4008, https://doi.org/10.1175/JCLI-D-19-0281.1.
Zhang, Z. T., K. C. Wang, D. Chen, J. Li, and R. Dickinson, 2019: Increase in surface friction dominates the observed surface wind speed decline during 1973–2014 in the Northern Hemisphere lands. J. Climate, 32, 7421–7435, https://doi.org/10.1175/JCLI-D-18-0691.1.
Zheng, C. W., J. Pan, and C. Y. Li, 2016: Global oceanic wind speed trends. Ocean Coastal Manage., 129, 15–24, https://doi.org/10.1016/j.ocecoaman.2016.05.001.
Zieger, S., A. V. Babanin, and I. R. Young, 2014: Changes in ocean surface wind with a focus on trends in regional and monthly mean values. Deep-Sea Res., 86, 56–67, https://doi.org/10.1016/j.dsr.2014.01.004.
