Projected Emergence Seasons of Year‐Maximum Near‐Surface Wind Speed

Global warming is expected to have far‐reaching impacts on the frequency and intensity of extreme events, but the effects of anthropogenic warming on the emergence seasons of year‐maximum near‐surface wind speed (NSWS) remain poorly understood. We provide a comprehensive map of the changing emergence seasons of year‐maximum NSWS using Coupled Model Intercomparison Project Phase 6 projections. Our analysis reveals a rapid response of synoptic‐scale extreme NSWS to global warming, with consistent spatial patterns observed across various periods and warming scenarios. The most significant increase (∼16%) in the emergence season is projected to occur in December‐January‐February (DJF) over Mid‐high‐latitude Asia by the end of the 21st century. The study also anticipates changes in the emergence seasons of year‐maximum NSWS at a regional scale. These results deepen our understanding of the complex and interconnected nature of global climate change and underscore the need for concerted efforts in addressing this pressing challenge.

modeling, which simulates atmospheric dispersion and pollutant transport (Pryor et al., 2012;Vose et al., 2014;Yan et al., 2006).Lastly, it serves as an indicator for gauging the intensity and frequency of extreme weather events, such as tropical storms and hurricanes (Alford et al., 2019;Usbeck et al., 2010).Thus, predicting extreme NSWS helps to forecast these events, which can have substantial social, economic, and environmental consequences (Brooks et al., 2003).
In recent years, the world has experienced an increase in extreme weather conditions, and anthropogenic warming has been identified as a major contributor to this phenomenon (Sisco et al., 2017;Wallace et al., 2014).However, most previous studies focused on changes in extreme NSWS strength (as reviewed by Wu et al. (2018)), neglecting variations between seasons when extreme NSWS occurs.In view of (a) the minimal number of studies investigating the emergence seasons of year-maximum NSWS over land; (b) the potential linkage between the emergence seasons of year-maximum NSWS and extreme weather events; and (c) better understanding contributions of the effect of global warming on year-maximum NSWS.The principle aim of this research is to investigate how the emergence seasons of year-maximum NSWS would change under different global warming scenarios.To achieve this aim, a combination of state-of-the-art global reanalysis data and modeling outputs are used.The results of this research will have important implications for wind energy management and extreme weather forecasting.The paper is organized as follows: The data and methodology are described in Section 2. Section 3 gives the results of Coupled Model Intercomparison Project Phase 6 (CMIP6) evaluation and projection.Conclusions are summarized in Section 4.

ERA5 and CMIP6
The fifth generation of the European Centre for Medium-Range Weather Forecasts Reanalysis (ERA5) is a state-of-the-art global atmospheric reanalysis (Hersbach et al., 2020;Olauson, 2018).ERA5 is widely recognized as the most accurate reanalysis data for representing observed NSWS (Ramon et al., 2019).We utilize the daily 10 m NSWS data from 1985 to 2014, with a high spatial resolution of 0.25° × 0.25° in latitude and longitude.
Eighteen global climate models from CMIP6 (O'Neill et al., 2016) under four shared socioeconomic pathways (SSP) are selected (Table S1 in Supporting Information S1), including the low-forcing scenario SSP1-2.6, the medium-forcing scenario SSP2-4.5, the relative-high-forcing scenario SSP3-7.0, and the high-forcing scenario SSP5-8.5.A model is chosen if it provides daily 10-m NSWS for both historical and four SSP experiments.To assure an equal weight when calculating the multi-model ensemble (MME) mean of models, we used the "first" ensemble member (r1i1p1f1) in each model for fair comparison (Shen et al., 2021)).All data sets are converted to 2° × 2° latitude and longitude resolution to ensure that they are comparable (Shen et al., 2022).

Definition of the Emergence Season of Year-Maximum NSWS
For ERA5 and each CMIP6 model, the emergence season is defined as the time when daily NSWS reaches the maximum in a year.While for MME of CMIP6, we first calculate the numbers of models that reach year-maximum NSWS in December-January-February (DJF), March-April-May (MAM), June-July-August (JJA), and September-October-November (SON) for each year and each grid-point, respectively.For example, the numbers of models are 5, 9, 3, and 1 for DJF, MAM, JJA, and SON at 60°N, 110°E in 2050.Then, we define the emergence season of year-maximum NSWS is MAM at this year and grid-point.If the maximum numbers for different seasons are repeated at a grid (e.g., 6 for DJF, 6 for MAM, 3 for JJA, and 3 for SON), we choose the season with a larger averaged NSWS as the emergence season.This method is different from the classical MME mean of all CMIP6 models, in that the latter way could ignore the model biases in synoptic NSWS.Lastly, the percentages of emergence season of year-maximum NSWS are summed up over North America (10°-70°N, 165°-60°W), Europe (35°-70°N, 10°W-50°E), Mid-high-latitude Asia (30°-75°N, 50°−180°E), Low-latitude Asia (0°-30°N, 50° −140°E), South America (60°S-10°N, 90°-30°W), Africa (40°S-35°N, 20°W-50°E), and Australia (45°-0°S, 110°−160°E), respectively (Table S2 in Supporting Information S1).

Model Evaluation
Given that human habitats and most wind farms are located onshore, we focus on terrestrial NSWS changes, excluding the Antarctic.For the climatologic emergence seasons of year-maximum NSWS during 1985-2014 (Figure 1), the spatial correlation coefficient between ERA5 and MME mean of CMIP6 is 0.62 (P < 0.01), the highest among the 18 CMIP6 models (ranging from 0.36 to 0.54).For the MME, spatial distributions of the model numbers of four seasons are shown in Figure S1 of Supporting Information S1.There exist a few grids where the numbers of two (∼2.6% of total land grids) or three (less than 0.1%) seasons are the same and the maximum, these grids on the conclusion.These indicate that the MME best captures the climatologic emergence seasons of year-maximum NSWS, and suggests the models' inherent biases could be reduced when we use more models.Generally, the emergence seasons are DJF for Northwest North America, Europe, Greenland, Central Africa, and Central Australia; MAM for the United States, North Africa, Mid-high-latitude Asia, and parts of Central Asia; JJA for Low-latitude South America, South Asia, and parts of South Africa; and SON for Northeast North America, Mid-latitude South America and Mid-latitude South Africa, respectively.These spatial patterns of emergence seasons are potentially associated with regional weather and climate processes.For instance, the DJF emergence season for Europe and Greenland may be influenced by Arctic cyclones and storm track activities (Screen & Simmonds, 2014;Screen et al., 2018).The MAM emergence season for the United States and Mid-high-latitude Asia could be related to the occurrences of American tornadoes (Brooks et al., 2003; Brotzge & Donner, 2013) and Asian dust storms (Shao & Dong, 2006;Sun & Zhao, 2008).And the JJA emergence season for South Asia may reflect the prevailing strong South Asian Summer Monsoon induced by the meridional temperature gradient between the Asia continent and the Indian Ocean (Li et al., 2021).Therefore, changes in emergence seasons may hold important implications for the government addressing extreme climate change resulting from anthropogenic global warming.

Projection of Year-Maximum Near-Surface Wind Speed
In this subsection, the MME of 18 CMIP6 models is used to perform future projections.Temporal evolutions of the emergence seasons percentage of year-maximum NSWS for both globally and across seven subregions under SSP5-8.5 are shown in Figure 2. Different subregions typically exhibit distinct characteristics in changing emergence seasons.For North America and Europe (Figures 2a and 2b), interannual variations in emergence seasons can reach up to ∼20% for DJF.Over Mid-high-latitude Asia (Figure 2c), it is evident that the year-maximum NSWS is more likely to occur in DJF in the future, while less likely in MAM.Over Low-latitude Asia (Figure 2e), year-maximum NSWS will be more likely to occur in JJA.Over South America (Figure 2f), year-maximum NSWS is more likely to occur in SON and less likely in JJA.Over Africa (Figure 2g), year-maximum NSWS will be less likely to occur in DJF.Over Australia (Figure 2h), the largest interannual variabilities of emergence seasons are observed compared to other subregions.Besides, similar temporal evolutions can be found in SSP1-2.6,SSP2-4.5, and SSP3-7.0(Figures S2-S4 in Supporting Information S1), albeit with weaker amplitudes.
Figure 3 illustrates the distributions of anomalous emergence seasons during the near-term (2021-2040), mid-term (2041-2060), and long-term (2081-2100) relative to the historical period (1995-2014) under four SSPs.Interestingly, the emergence seasons of year-maximum NSWS exhibit significant changes even in lower warming scenarios, with similar spatial patterns among them.Except for Africa, more year-maximum NSWS will occur in DJF and MAM in the mid-high-latitude Northern Hemisphere, and in SON in the mid-high-latitude Southern Hemisphere.The increased emergence seasons of JJA are primarily located in West Africa, potentially associated with an amplification of synoptic variability of the West African Summer Monsoon under global warming (Akinsanola et al., 2020).Overall, these projections provide obvious evidence of shifting emergence seasons of year-maximum NSWS in the future.The consistent and clustered changes suggest connections between regional synoptic to interannual variabilities of climate change and extreme NSWS.
Figure 4 provides a quantification of changing emergence seasons to enhance our understanding of year-maximum NSWS changes on a global scale and across seven subregions.Generally, stronger global warming corresponds to larger changing rates of emergence seasons.Compared to 1995-2014, the global year-maximum NSWS is more likely to occur in DJF (1%-3.5%)and less likely to occur in MAM (0%-3.5%)(Figure 4d).Over North America (Figure 4a), the increasing rates in DJF are comparable to the decreasing rates in SON, with percentages range from 2%-7%.Over Mid-high-latitude Asia (Figure 4c), emergence seasons of year-maximum NSWS will only be more frequent in DJF, with increasing rates of 6%-16%.Over Low-latitude Asia (Figure 4e), it is evident that the emergence seasons will linearly increase in JJA (0%-5%), while decreasing in MAM (0.5%-4.5%).Over South America and Australia (Figures 4f and 4h), the emergence seasons of year-maximum NSWS will occur more frequently in SON, with increasing rates of 0%-8% and 2%-7%, respectively.Over Africa (Figure 4g), there will be fewer extreme NSWS in DJF, with percentage ranges of 2%-7%.Besides, we also noted there are some large differences for regional scales, such as MAM in North America under SSP1-2.6 (Figure 4a), and JJA in Australia under SSP2-4.5 (Figure 4h).These discrepancies could be induced by the effects of internal variabilities under lower global warming levels.

Conclusion and Discussion
Global warming projections suggest a wide range of impacts on weather systems and processes, including increased frequency and intensity of extreme weather events (Meehl et al., 2000;Scher & Messori, 2019).However, little is known about the potential effects of anthropogenic warming on the emergence seasons of year-maximum NSWS.In this study, we present a comprehensive map of changing emergence seasons of year-maximum NSWS under anthropogenic warming for the first time by using CMIP6 projections.The spatial patterns of changing emergence seasons are consistent across different time periods and warming scenarios, emphasizing a rapid response of synoptic-scale extreme NSWS to global warming.The most drastic change occurs in DJF over Mid-high-latitude Asia, with an increasing rate of ∼16% by the end of the 21st century under SSP5-8.5.Taking into account the model uncertainties, it is expected that more year-maximum NSWS will occur in DJF over North America, Mid-high-latitude Asia, and Low-latitude Asia; MAM over Africa; JJA over Low-latitude Asia, and west Africa; and SON over South America and Australia.This work, while substantial, it still contains some limitations and provides room for further improvement, which should be addressed in the future: (a) We only considered the changing emergence seasons of year-maximum NSWS but did not consider the changes in the strength of year-maximum NSWS.Kumar et al. (2015) found the MME of Coupled Model Intercomparison Project Phase 5 can well simulate the spatial pattern of year-maximum NSWS.Considering the global annual mean terrestrial NSWS is projected to decrease in the future under global warming (Shen et al., 2022), it may be anticipated a coherent decrease in year-maximum NSWS (Pryor et al., 2020).(b) The effects of internal variabilities and model uncertainties on the changing emergence seasons are not fully considered, due to the largest correlation coefficient between MME of CMIP6 and ERA5.Large ensemble models may play a good role in investigating this issue (Li et al., 2021).(c) The dominant physical mechanisms behind the projections across the subregions remain different and unclear, including the effects of temperature gradient, surface friction, land use, upper-level jet stream, and boundary layer processes (Pryor et al., 2012;Shen et al., 2023;Wu et al., 2018;Zha et al., 2021).(d) The scarcity of daily-resolution situ stations limits our ability to use observations to evaluate CMIP6 simulations (Wu et al., 2018).However, this may be not a problem in certain regions that have a lot of situ stations, such as Europe, East Asia, and the United States.
Whatever, these projections provide valuable information for governments and stakeholders to prepare for and adapt to extreme weather events under global warming, even if ambitious mitigation measures are taken to limit the warming below 2°C.01648), and the Swedish Research Council (VR: 2021-02163).Huishuang Yuan is supported by the Innovation and Entrepreneurship Training Program for College Students of Yunnan University (S202310673266).This study is also supported by the Sven Lindqvists Forskningsstiftelse and Research Fund Adlerbertska Stiftelse.

Figure 1 .
Figure 1.Emergence seasons of the year-maximum near-surface wind speed over land during 1985-2014 based on (a) ERA5, (b) multi-model ensemble of CMIP6, and (c-t) 18 CMIP6 models.Blue, yellow, red and purple represent December-January-February (DJF), March-April-May (MAM), June-July-August (JJA), and September-October-November (SON), respectively.R on the top-right denotes the pattern correlation coefficient between the models and ERA5.

Figure 4 .
Figure 4. (a) Projections of percentage changes in emergence seasons of the year-maximum near-surface wind speed over land in SSP1-2.6 (blue), SSP2-4.5 (yellow), SSP3-7.0 (red), and SSP5-8.5 (purple) relative to 1995-2014 based on the multi-model ensemble of CMIP6 over North America.For each season, three clusters of solid bars denote near-term, mid-term, and long-term, respectively.The thick solid lines denote the 5%-95% model range.(b-h) Same as (a), but for Europe, Mid-high-latitude Asia, Global, Low-latitude Asia, South America, Africa, and Australia, respectively.