Abstract

The methods used in earlier research focusing on the province of Ontario, Canada, were adapted for the current paper to expand the study area over the entire nation of Canada where various industries (e.g., transportation, agriculture, energy, and commerce) and infrastructure are at risk of being impacted by extreme wind gust events. The possible impacts of climate change on future wind gust events across Canada were assessed using a three-step process: 1) development and validation of hourly and daily wind gust simulation models, 2) statistical downscaling to derive future station-scale hourly wind speed data, and 3) projection of changes in the frequency of future wind gust events. The wind gust simulation models could capture the historically observed daily and hourly wind gust events. For example, the percentage of excellent and good validations for hourly wind gust events ≥90 km h−1 ranges from 62% to 85% across Canada; the corresponding percentage for wind gust events ≥40 km h−1 is about 90%. For future projection, the modeled results indicated that the frequencies of the wind gust events could increase late this century over Canada using the ensemble of the downscaled eight-GCM simulations [Special Report on Emissions Scenarios (SRES) A2 and B1]. For example, the percentage increases in future daily wind gust events ≥70 km h−1 from the current condition could be 10%–20% in most of the regions across Canada; the corresponding increases in future hourly wind gust events ≥70 km h−1 are projected to be 20%–30%. In addition, the inter-GCM and interscenario uncertainties of future wind gust projections were quantitatively assessed.

1. Introduction

It is well known that high wind speeds or gust wind extremes from nontornadic storms are among the most destructive natural hazards over the world that cause considerable economic and social costs, as well as damage to properties, infrastructure, agriculture, power lines, and trees (Dore 2003; Changnon 2009; Lopes et al. 2009; Pinto et al. 2010). The damage to structures can increase significantly when wind gusts exceed a certain collective wind pressure design or “breaking point.” Climate information of wind gusts is required to determine wind loads for infrastructure design codes and standards, which play an important role in preventing damage and reducing risks to lives from extreme wind events. In addition, potential changes to future wind regimes under a changing climate have also received great attention from the climate community (see some of the references listed in Table 1). Under global warming, the severity and frequency of future wind gust events could be expected to change late this century. Potential impacts of climate change on wind gusts are of great interest to many economic sectors, especially for the wind-energy industry (Jiang et al. 2010; Najac et al. 2011). The scientific information on the projections of future wind gust events is essential for decision makers to improve the adaptive capacity of infrastructure at the risk of being impacted by extreme wind gust events. In light of this concern, Environment Canada has recently completed a climatological research project to assess the possible impacts of climate change on wind gust events over the province of Ontario, Canada (Cheng et al. 2012). That study employed a three-step process: 1) development and validation of wind gust simulation models to verify historical hourly and daily wind gust events, 2) statistical downscaling to derive future station-scale hourly wind speed data, and 3) projections of changes in the frequency of future hourly and daily wind gust events. These methods were adapted in the current paper to expand the study area over the entire nation of Canada. We hope this study provides further scientific information on projections of future hourly and daily wind gust events across Canada. This information could be useful for various industries (e.g., transportation, agriculture, energy, and commerce) and communities to consider climate change in revising engineering infrastructure design standards, developing adaptation strategies/policies, and reducing the associated risks.

Table 1.

Examples of previous studies on climate change and wind regimes. RCM denotes regional climate model.

Examples of previous studies on climate change and wind regimes. RCM denotes regional climate model.
Examples of previous studies on climate change and wind regimes. RCM denotes regional climate model.

This paper is organized as follows: In section 2, data sources and treatments are described. Section 3 summarizes analysis techniques. Section 4 presents classification of wind gust stations. Section 5 includes the results and discussion on historical hourly and daily wind gust simulation and validation, changes in frequency of future hourly and daily wind gust events, and the uncertainty of future projections. The conclusions from the study are summarized in section 6.

2. Data sources and treatment

A variety of data sources, including surface meteorological observations, reanalysis data, and GCM simulations, are needed in order to assess possible impacts of climate change on future wind gust events. The surface meteorological observations include hourly and daily air temperature, sea level air pressure, 10-m wind gust, u wind, and υ wind during the period 1955–2009 (hourly wind gust during 1994–2009). With the exception of wind gust, when the observed hourly data were missing for three consecutive hours or less, missing data were interpolated using a temporal linear method; days with data missing for four or more consecutive hours were excluded from the analysis. Across the study area, 0.22% of the total hours required missing data interpolation; after interpolation, the dataset was 99.85% complete. The interpolation treatment cannot be applied to wind gust data since the wind gust values were recorded only on the hours and dates when the event occurred. An hourly wind gust is defined as a sudden increase in wind speed occurring during the 10-min period prior to the observation with a ≥28 km h−1 speed and measured at 9 km h−1 greater than the 2-min-average wind speed prior to the observation (Environment Canada 1977). A daily wind gust is defined as a daily peak wind that is greater than or equal to 28 km h−1 measured during the entire 24-h period of a day.

In addition to the meteorological observations, for statistical downscaling the 6-hourly reanalysis data at 0600, 1200, 1800, and 0000 UTC for the period 1958–2009 from the U.S. National Centers for Environmental Prediction (NCEP) website were included in the analysis. To combine the gridded reanalysis data with the surface observations, similar to data treatment in the study (Cheng et al. 2012), the reanalysis data from the surrounding four-grid domain fields were interpolated to the selected weather stations using the inverse-distance method (Shen et al. 2001). Cheng et al. (2010) have tested two other domain sizes (i.e., 16 and 36 grid) and concluded that the four-grid interpolation is best based on its correlation with radiosonde data.

The 104 stations across Canada with the hourly and daily surface meteorological observations, as shown in Fig. 1, were selected for the study based on the length of the available data record. Of the selected stations, 60% possess at least a 30-yr data record of daily wind gust observations, of which more than half have a data record of 40–55 years. To consider the spatial density of the stations for some of the regions, the stations with a data record less than 20 years were also included in the study, which is 9% of the selected stations. The detailed information on the length of daily wind gust records for each of the selected stations is illustrated in Fig. 1. As shown in Fig. 1, there are a certain number of stations distributed in each of the wind gust regions across Canada, except for the most northern region (i.e., region N1, only two stations). For more information regarding the classification of wind gust stations refer to section 4.

Fig. 1.

Study area and locations of the selected stations as well as wind gust regions over Canada.

Fig. 1.

Study area and locations of the selected stations as well as wind gust regions over Canada.

As pointed out by Cheng et al. (2012), of all measured and archived meteorological fields, wind data are probably among the most variable in Canada for protocols, instrument sitting, anemometer heights and types, measurement averaging periods, instrument maintenance, and uncertainties, which are described in the wind metadata. It is important to keep the consistency of observed wind data in terms of instrument sitting and anemometer heights over the time period in order to more effectively develop wind gust simulation models. Therefore, a certain time period of the wind data record with a certain level consistency in wind data observations needs to be identified. To achieve this, in addition to the wind metadata, annual total frequencies of wind speed or wind gust greater than the certain value (e.g., ≥43 or ≥60 km h−1) were analyzed for each of the 104 selected stations over Canada. Based on the time series of the frequency of wind speed or wind gust events and wind metadata, the daily wind gust and hourly wind speed data observed for a certain time period at each of the stations were selected for the analysis in order to keep the consistency of observed wind data in terms of the frequency of wind speed or wind gust events and instrument sitting and anemometer heights (e.g., 10 m). For example, the time periods of daily wind gust data records selected for Resolute Cars, St. Johns, Victoria, and Windsor, Canada, are 1982–2009, 1976–2009, 1965–2009, and 1964–2009, respectively. Detailed information on the length of daily wind gust data records for each of the stations is presented in Fig. 1. Hourly wind gust data observed for the time period 1994–2009 were used for all stations studied since the data are available only back to 1994. Details of selecting the time period of the wind data record are not presented in the current paper owing to the limitations of space; refer to Cheng et al. (2012) for more detailed information.

In addition to observed meteorological data, daily future climate data projected from eight GCMs with two Intergovernmental Panel on Climate Change (IPCC) scenarios [Special Report on Emissions Scenarios (SRES)A2 and B1] were used in the analysis (see Table 2 for a complete list of all models and expansions). The GCM simulations were retrieved from the website of the Program for Climate Model Diagnosis and Intercomparison (PCMDI 2010). The GCMs are CGCM3.1/T63, CNRM-CM3, CSIRO Mk3.0, ECHAM5 (from the Max Planck Institute), ECHO-G, GFDL CM2, and MIROC3.2 (medres). As pointed out by Cheng et al. (2012), these eight GCMs were selected simply because the simulations of surface and upper-air temperature, pressure, u wind, and υ wind are available from the website, which were needed in our climate change impact studies. Scenarios A2 and B1 consider different assumptions of future greenhouse gas (GHG) emissions derived from a distinctly different direction for future population growth, economic development, and technological change. Consequently, both scenarios were used in the study to generate a range of projections of possible changes in the frequency of future hourly and daily wind gust events over Canada.

Table 2.

GCM acronyms and expansions.

GCM acronyms and expansions.
GCM acronyms and expansions.

3. Analysis techniques

A recent study by Cheng et al. (2012) employed wind gust simulation models with downscaled future station-scale wind speed to project changes in the frequency of future daily and hourly wind gust events focusing on the province of Ontario, Canada. The methods comprise three steps: 1) development and validation of hourly and daily wind gust simulation models, 2) statistical downscaling to derive future station-scale hourly wind speed data, and 3) projection of changes in the frequency of future hourly and daily wind gust events. These methods were adapted in the current paper to expand the study area from the province of Ontario to the entire nation of Canada. The principal methods and steps used in the study are summarized in Fig. 2. Each of the steps is outlined as follows; for more detailed information, refer to Cheng et al. (2012).

Fig. 2.

Flowchart of methodologies and steps used in the study.

Fig. 2.

Flowchart of methodologies and steps used in the study.

The wind gust factor, representing a relationship between the peak gust wind speed and mean wind speed, was employed to estimate wind gusts. The wind gust factor kg(V) used in this study is defined as follows:

 
formula

where Vg and V are the gust wind speed and mean wind speed, respectively. Two kinds of wind gust factors were used in the study: 1) hourly wind gust factor defined by the ratio of hourly gust wind speed and hourly wind speed and 2) daily wind gust factor defined by the ratio of daily gust wind speed and 1-h maximum wind speed (the maximum wind speed of hourly observations during a day). These two wind gust factors were used to develop the simulation models for hourly and daily gust wind speed based on the mean wind speed, respectively. As described in the study (Cheng et al. 2012), the wind gust simulation models were developed separately for four seasons in order to assess climate change impacts on two different-scale windstorm systems: synoptic windstorms typically occurring in the winter season and localized convective windstorms generally occurring in the summer season. The wind gust simulation models were also validated using a leave-one-year-out cross-validation procedure, in which the procedure was repeatedly run to develop a wind gust simulation model that would validate one year of independent data for each year in the dataset. The calibrated and validated hourly and daily gust wind speeds were then compared with observations to evaluate model performance.

To project future hourly and daily wind gusts, future hourly wind speed data at the station scale are necessary for the use of wind gust simulation models. The statistical downscaling method developed by Cheng et al. (2008, 2012) was employed to downscale future hourly wind speed. As described by Cheng et al. (2012), the downscaling method comprises a two-step process: 1) spatially downscaling daily u wind and υ wind from the GCM domain field to the selected weather stations across Canada, as shown in Fig. 1, and 2) temporally downscaling daily u-wind and υ-wind speeds to hourly time steps. The downscaling transfer functions were constructed using three regression techniques: multiple stepwise regression, stepwise orthogonal regression, and autocorrelation correction regression. The predictors selected in the development of daily and hourly wind downscaling transfer functions include daily and hourly u wind and υ wind, west–east and south–north sea level pressure gradients across the individual weather stations, as well as u wind and υ wind at the previous hour. One of the conclusions from the study is that these regression techniques are suitable and work well to develop wind speed downscaling transfer functions that can remove most of the GCM biases. Most of the daily downscaling transfer functions for surface winds possess coefficients of determination R2 greater than 0.8; the corresponding R2 of hourly downscaling transfer functions range from 0.69 to 0.92, with half of them greater than 0.89.

Following the downscaling of the future hourly wind speed, future hourly and daily gust wind speeds are able to be projected using wind gust simulation models, expressed in Eq. (1). Cheng et al. (2012) also analyzed future wind gust projections to ascertain whether the models are suitable for future projection through comparing hourly and daily wind gusts projected by downscaled GCM historical runs with observations over a comparable time period (1961–2000). The results showed that the data distributions of both datasets are similar; so it can be concluded that the methods used in the study are suitable for projecting future hourly and daily wind gust information at a local or station scale. These comparison results are not presented in this current paper owing to the limitations of space; refer to Cheng et al. (2008, 2012) for details.

4. Classification of wind gust stations

To more effectively present projected results on changes in future hourly and daily wind gust events for the study area, it is necessary to classify the 104 stations over Canada shown in Fig. 1 into some regions with unique wind gust–related climatological characteristics. The historical monthly mean number of days observed with wind gusts ≥50, ≥60, and ≥70 km h−1 was used as an indicator to classify the wind gust stations. The classification method is comprised of the correlation matrix–based principal component analysis (PCA) and average linkage clustering procedure. The PCA was applied to the 36 variables for the 104 stations, producing a four-component solution that explained 96% of the total variance within the dataset. The remainder of the components with eigenvalues less than one was discarded. The average linkage clustering procedure was performed on the spatial four-component scores and resulted in 13 wind gust regions plus two individual stations (i.e., Sandspit, British Columbia, and Lethbridge, Alberta) for the study area (Fig. 1). This classification method attempts to find a suitable classification solution as it minimizes within-category variances and maximizes between-category variances (Boyce 1996; Cheng et al. 2007, 2011). The classification of wind gust stations used in this study focused on climatological characteristics of daily wind gust events rather than regional circulation patterns. However, the regional circulation patterns should be similar within a region for the geographically close stations with unique climatological characteristics of the observed frequency of daily wind gust events.

Table 3 shows the region-averaged seasonal total number of days observed with wind gusts ≥50, ≥60, and ≥70 km h−1. From Table 3, the major wind gust climatological characteristics can be summarized as follows, which can be supported by the previous studies on analyses of mean seasonal wind speed (Thomas 1953) and 30-yr-return-period extreme winds (Yip et al. 1995):

  1. Across the country, the wind gust events most frequently occur in the Atlantic provinces (regions A1 and A2), especially in Newfoundland (region A2). Conversely, the wind gust events least frequently occur in region W1, which covers Yukon Territory, northern British Columbia, the southern Northwest Territories, northern Prairie provinces (Alberta, Saskatchewan, and Manitoba), and northern Ontario across the region from northwest to southeast. In addition, the wind gust events more frequently occur in southern Ontario and southern Quebec (region C3) in central Canada, in eastern British Columbia (region W4), in the southern Prairie provinces (region 5), and in the northern Canada regions (especially region N3) than other regions. These regional wind gust climatological characteristics are consistent with the previous studies (Thomas 1953; Yip et al. 1995).

  2. In the seasonality perspectives, the wind gust events least frequently occur in summer months for most of the regions across Canada, especially for the Atlantic regions (A1 and A2), central Canada regions C2 and C3, and northern Canada regions N2 and N3. All five wind gust regions in western Canada, except for region W4, usually experience the similar occurrence frequencies of the wind gust events from season to season. However, in region W4, the wind gust events least frequently occur in summer months.

  3. Two individual stations (Sandspit, British Columbia, and Lethbridge, Alberta) are isolated by the classification method from regions W4 and W5, respectively, since the wind gust events more frequently occur at the stations than others within the regions.

Table 3.

Region-averaged seasonal total number of days observed with daily wind gust events ≥50, ≥60, and ≥70 km h−1 (data record length shown in Fig. 1). Periods include December–February (DJF), March–May (MAM), June–August (JJA), and September–November (SON).

Region-averaged seasonal total number of days observed with daily wind gust events ≥50, ≥60, and ≥70 km h−1 (data record length shown in Fig. 1). Periods include December–February (DJF), March–May (MAM), June–August (JJA), and September–November (SON).
Region-averaged seasonal total number of days observed with daily wind gust events ≥50, ≥60, and ≥70 km h−1 (data record length shown in Fig. 1). Periods include December–February (DJF), March–May (MAM), June–August (JJA), and September–November (SON).

5. Results and discussion

a. Historical hourly and daily wind gust simulation and validation

The four correctness levels of excellent, good, fair, and poor were subjectively defined based on absolute difference between observed and simulated hourly or daily gust wind speeds in order to more effectively evaluate the performance of wind gust simulation models. The definition of four correctness levels is shown in Table 4. The results of the model performance validation for hourly wind gust events over 13 regions across Canada are summarized in Fig. 3. Daily wind gust model performance is slightly worse than the hourly model validation results, which is not shown in the paper owing to the limitations of space. The model evaluation and future projection in this study were undertaken for four wind gust categories based on observed wind gusts greater than or equal to certain thresholds (i.e., ≥90, ≥70, ≥40, and ≥28 km h−1). The thresholds of 90 and 70 km h−1 were selected to correspond with the criteria currently used by Environment Canada for issuing wind gust warnings (Environment Canada 2013). The threshold of 40 km h−1 was selected since a wind turbine delivers most of its power only once the wind speed is greater than this value (Moyer 2009). The threshold of 28 km h−1 was selected since it is the minimum value to record wind gust observations.

Table 4.

Criteria used to evaluate hourly and daily wind gust simulation models. Diff is the absolute difference between observed and simulated wind gust.

Criteria used to evaluate hourly and daily wind gust simulation models. Diff is the absolute difference between observed and simulated wind gust.
Criteria used to evaluate hourly and daily wind gust simulation models. Diff is the absolute difference between observed and simulated wind gust.
Fig. 3.

Evaluation results on validation performance of hourly wind gust simulation models for wind gust regions across Canada during the period 1994–2009.

Fig. 3.

Evaluation results on validation performance of hourly wind gust simulation models for wind gust regions across Canada during the period 1994–2009.

From Fig. 3, it can be observed that the proportion of the model validations that fell into excellent and good correctness levels is much greater than the proportion that fell into fair and poor levels. In addition, the percentage of excellent and good simulations slightly increases when the wind gust categories move from extreme events (e.g., ≥90 km h−1) to the weaker wind gust events (e.g., ≥40 km h−1). For example, the percentage of excellent and good validations for hourly wind gust events ≥90 km h−1 ranges from 62% to 85% over 13 regions across Canada; the corresponding percentage for wind gust events ≥40 km h−1 is about 90%. As pointed out in the previous study (Cheng et al. 2012), it can be concluded that the wind gust simulation models can capture historically observed hourly and daily wind gust events and are suitable to assess changes in the frequency of future wind gust events at a local scale or station scale.

b. Changes in frequency of future hourly and daily wind gust events

The projections of percentage changes in the future annual-mean number of hours and days with wind gust events (greater than or equal to certain thresholds) from the historical baselines were analyzed for each of 12 regions, as shown in Figs. 4 and 5. Region N1 was analyzed separately since it has only two members of the stations so that it is unable to calculate the 95% confidence interval, as shown in Fig. 6 with the two other individual stations. The historical baselines—the region-averaged annual-mean frequencies of historically observed wind gust events—are presented in Table 5 as the references for future projections, which also illustrates frequency distributions of daily and hourly wind gust events. It is immediately apparent in Figs. 46 that the entire nation of Canada could experience more wind gust events late this century. The major findings of future wind gust projections can be summarized as follows:

  1. The magnitude of the percentage increases in the frequency of future hourly and daily wind gust events would be greater for more severe wind gust events. For example, the percentage increases in the frequency of future hourly and daily wind gust events ≥28 km h−1 are projected to be less than 10%, generally for most of the regions. The corresponding increases for future hourly and daily wind gust events ≥70 km h−1 in most of the regions could be 20%–30% and 10%–20%, respectively. The percentage increases in the frequency of future daily wind gust events ≥90 km h−1 are projected to be 20%–40%. The corresponding increases for future hourly wind gust events ≥90 km h−1 are projected to be more than double for all regions, except regions A1 and A2. The projected greater percentage increases for future extreme wind gust events (i.e., ≥90 km h−1) are partially due to rare cases of the events, especially for regions W1 and C1.

  2. The projected percentage increases in the frequency of future hourly and daily wind gust events derived from the A2 scenario differs from the B1 scenario. The increases derived from downscaled A2 scenario are usually slightly greater than those derived from downscaled B1 scenario. The A2 scenario projects higher emissions and more future warming than the B1 scenario. For more detailed differences regarding interscenario and inter-GCM uncertainties, refer to section 5c.

Fig. 4.

Projected percentage changes in annual-mean frequency of future hourly wind gust events derived from downscaled eight-GCM ensemble, summarized by regions (four bars in each of the panels: the first two for scenario A2 over the periods 2046–65 and 2081–2100; the last two for scenario B1 over the periods 2046–65 and 2081–2100). The 95% confidence interval is indicated.

Fig. 4.

Projected percentage changes in annual-mean frequency of future hourly wind gust events derived from downscaled eight-GCM ensemble, summarized by regions (four bars in each of the panels: the first two for scenario A2 over the periods 2046–65 and 2081–2100; the last two for scenario B1 over the periods 2046–65 and 2081–2100). The 95% confidence interval is indicated.

Fig. 5.

As in Fig. 4, but for daily wind gust events.

Fig. 5.

As in Fig. 4, but for daily wind gust events.

Fig. 6.

As in Figs. 4 and 5, but for Eureka and Resolute Cars in region N1, Sandspit near region W4, and Lethbridge near region W5.

Fig. 6.

As in Figs. 4 and 5, but for Eureka and Resolute Cars in region N1, Sandspit near region W4, and Lethbridge near region W5.

Table 5.

Region-averaged annual-mean number of hours and days observed with wind gust events greater than or equal to the thresholds (hourly wind gust period 1994–2009; daily wind gust: record length shown in Fig. 1).

Region-averaged annual-mean number of hours and days observed with wind gust events greater than or equal to the thresholds (hourly wind gust period 1994–2009; daily wind gust: record length shown in Fig. 1).
Region-averaged annual-mean number of hours and days observed with wind gust events greater than or equal to the thresholds (hourly wind gust period 1994–2009; daily wind gust: record length shown in Fig. 1).

As pointed out in the previous study (Cheng et al. 2012), the wind gust events that occur in the study area are usually triggered by two different-scale systems: synoptic and localized convective windstorms. Synoptic windstorms typically occur in the winter season, while localized convective windstorms generally occur in the summer season. The seasonal projections of future wind gust events were analyzed in order to assess climate change impacts on two different-scale windstorm systems. The projected percentage changes in the frequency of future daily wind gust events ≥70 km h−1 from the historical baselines for scenario A2 over 2081–2100 were analyzed and presented by seasons in Fig. 7. It can be observed from Fig. 7 that the projected percentage increases in future daily wind gust events across Canada for the summer season are generally greater than those for rest of the seasons, with 42 stations possessing a relative increase rate greater than 30%. In other words, all regions across Canada could experience more localized convective windstorms in summer season late this century due to warmer temperatures under a future changing climate. The corresponding increase rate in winter can be witnessed mostly in central Canada (i.e., regions C1 and C3), western Canada (i.e., regions W1, W2, and W5), and Newfoundland (i.e., region A2), where they could experience more synoptic windstorms in the winter season in the future. The possible reason for this could be that the future synoptic windstorms in the winter season could move northward and northeastward into the area. This conclusion could be supported by a previous study (Lambert and Fyfe 2006), which suggests that the number of intense midlatitude cyclones, projected by the GCMs participating in the IPCC Fourth Assessment Report (AR4) diagnostic exercise under the A2 and B1 scenarios, would increase by the periods 2046–65 and 2081–2100. It is noteworthy that the current study did not consider the seasonal shift of the future localized convective and synoptic windstorm systems over Canada. Further analysis in a future study needs to be conducted to address this seasonal shifting, with more focus on identification and analysis of the typical climatological characteristics between both windstorm systems. It is also necessary to ascertain whether those analyses are suitable for the projection of and differentiation between both future windstorm systems.

Fig. 7.

Projected percentage changes in seasonal-mean frequency of future daily wind gust events ≥70 km h−1 for scenario A2 during the period 2081–2100.

Fig. 7.

Projected percentage changes in seasonal-mean frequency of future daily wind gust events ≥70 km h−1 for scenario A2 during the period 2081–2100.

c. Uncertainty of the future projections

As discussed in the previous study (Cheng et al. 2012), through the downscaling process, most of the GCM model bias can be removed using about 50-yr historical relationships between regional-scale predictors and station-scale observed hourly and daily wind data. As a result, the quality of future wind gust projections was improved. However, conclusions made in this study about the impacts of climate change on future wind gust events still rely on GCM scenarios and projections. Consequently, the inter-GCM and interscenario uncertainties still exist in the projections of future hourly and daily wind gust events. To quantitatively assess inter-GCM uncertainties on future wind gust projections, we analyzed the region-averaged absolute difference among eight selected GCMs for each of the A2 and B1 scenarios. Similarly, the interscenario uncertainties were quantitatively assessed using the region-averaged absolute difference between both scenarios A2 and B1. For inter-GCM uncertainty calculation, there are 28 pairs in total combined from eight GCM models included in the analysis. As shown in Table 6, it is immediately apparent that the inter-GCM uncertainties of projected percentage increases in the frequency of future hourly and daily wind gust events are generally similar to or greater than the interscenario uncertainties. To indicate the statistical significance of the values shown in Table 6, the standard deviation was analyzed and the results showed the standard deviations were usually much smaller than the means. For example, the standard deviations of the interscenario uncertainties of percentage increases in the frequency of future hourly wind gust events ≥28 and ≥40 km h−1 are about 10%–40% of the means; the corresponding standard deviations for hourly wind gust events ≥70 km h−1 are about 30%–60% of the means.

Table 6.

Region-averaged inter-GCM and interscenario uncertainties of percentage increases in the frequency of future hourly and daily wind gust events (≥28, ≥40, and ≥70 km h−1) from the current conditions.

Region-averaged inter-GCM and interscenario uncertainties of percentage increases in the frequency of future hourly and daily wind gust events (≥28, ≥40, and ≥70 km h−1) from the current conditions.
Region-averaged inter-GCM and interscenario uncertainties of percentage increases in the frequency of future hourly and daily wind gust events (≥28, ≥40, and ≥70 km h−1) from the current conditions.

Comparing Figs. 46 with Table 6, it can be seen that across the study area for all regions, the projected percentage increases in the frequency of future hourly and daily wind gust events ≥28 and ≥40 km h−1 are double or more than double the inter-GCM or interscenario uncertainties. For wind gust events ≥70 km h−1, the corresponding projected percentage increases are similar to the inter-GCM or interscenario uncertainties. However, for wind gust events ≥90 km h−1, the inter-GCM and interscenario uncertainties generally greater than the projected percentage increases, which is not shown in the table owing to the limitations of space. The results on the inter-GCM and interscenario uncertainties derived from this study are consistent with conclusions made by the previous study geographically focusing on the province of Ontario (Cheng et al. 2012).

6. Conclusions

The overarching purpose of this study was to assess possible impacts of climate change on future hourly and daily wind gust events over Canada. To achieve this goal, a three-step process was used: 1) development and validation of hourly and daily wind gust simulation models, 2) statistical downscaling to derive future station-scale hourly wind speed data, and 3) projections of changes in the frequency of future hourly and daily wind gust events. The study results indicated that the wind gust simulation models can capture historically observed hourly and daily wind gust events in the past 50 years. Furthermore, it can be concluded that the methods used in the study are suitable to assess changes in the frequency of future hourly and daily wind gust events at a local scale or station scale.

The projected results clearly show that Canada could possibly receive more wind gust events late this century than has been historically experienced. The magnitude of the projected percentage increases in the frequency of future wind gust events would be generally greater for more severe wind gust events. For example, the percentage increases in the frequency of future hourly wind gust events ≥28 and ≥70 km h−1 are projected to be less than 10% and 20%–30%, respectively, in most of the regions. The corresponding increases for future hourly wind gust events ≥90 km h−1 are projected to be more than double for most of the regions. The implications of these increases should be taken into consideration and integrated into policies and planning for adaptation strategies, including measures to incorporate climate change into engineering infrastructure design standards and disaster-risk-reduction measures.

The inter-GCM and interscenario uncertainties of future hourly and daily wind gust projections were quantitatively assessed. The results on inter-GCM and interscenario uncertainty analysis derived from this study are consistent with the previous study geographically focusing on the province of Ontario (Cheng et al. 2012). The analyzed results clearly indicated that for hourly and daily wind gust events ≥28 and ≥40 km h−1, both the inter-GCM and interscenario uncertainties are about half or less than half of the projected percentage increases in the frequency of future wind gust events. For wind gust events ≥70 km h−1, the uncertainties are similar to the corresponding projected percentage increases. However, for wind gust events ≥90 km h−1, the inter-GCM and interscenario uncertainties generally greater than the projected percentage increases, partially due to rare cases of the extreme wind gust events, especially for regions W1 and C1.

Acknowledgments

The authors thank three anonymous referees for providing detailed constructive comments that significantly improved the original manuscript.

REFERENCES

REFERENCES
Beniston
,
M.
, and
Coauthors
,
2007
:
Future extreme events in European climate: An exploration of regional climate model projections
.
Climatic Change
,
81
,
71
95
.
Bogardi
,
I.
, and
I.
Matyasovszky
,
1996
:
Estimating daily wind speed under climate change
.
Sol. Energy
,
57
,
239
248
.
Boyce
,
A. J.
,
1996
: Mapping diversity: A comparative study of some numerical methods. Numerical Taxonomy, A. J. Cole, Ed., Academic Press, 1–30.
Changnon
,
S. A.
,
2009
:
Temporal and spatial distributions of wind storm damages in the United States
.
Climatic Change
,
94
,
473
482
.
Cheng
,
C. S.
,
M.
Campbell
,
Q.
Li
,
G.
Li
,
H.
Auld
,
N.
Day
,
D.
Pengelly
, and
D.
Yap
,
2007
:
A synoptic climatological approach to assess climatic impact on air quality in south-central Canada. Part I: Historical analysis
.
Water Air Soil Pollut.
,
182
,
131
148
.
Cheng
,
C. S.
,
G.
Li
,
Q.
Li
, and
H.
Auld
,
2008
:
Statistical downscaling of hourly and daily climate scenarios for various meteorological variables in south-central Canada
.
Theor. Appl. Climatol.
,
91
,
129
147
.
Cheng
,
C. S.
,
G.
Li
,
Q.
Li
, and
H.
Auld
,
2010
:
A synoptic weather typing approach to simulate daily rainfall and extremes in Ontario, Canada: Potential for climate change projections
.
J. Appl. Meteor. Climatol.
,
49
,
845
866
.
Cheng
,
C. S.
,
G.
Li
, and
H.
Auld
,
2011
:
Possible impacts of climate change on freezing rain using downscaled future climate scenarios: Updated for eastern Canada
.
Atmos.–Ocean
,
49
,
8
21
.
Cheng
,
C. S.
,
G.
Li
,
Q.
Li
,
H.
Auld
, and
C.
Fu
,
2012
:
Possible impacts of climate change on wind gusts under downscaled future climate conditions over Ontario, Canada
.
J. Climate
,
25
,
3390
3408
.
Dore
,
M. H. I.
,
2003
:
Forecasting the conditional probabilities of natural disasters in Canada as a guide for disaster preparedness
.
Nat. Hazards
,
28
,
249
269
.
Environment Canada
,
1977
: MANOBS: Manual of surface weather observations. 7th ed. Meteorological Service of Canada, Environment Canada, 386 pp.
Environment Canada
, cited
2013
: Public alerting criteria. [Available online at http://www.ec.gc.ca/meteo-weather/default.asp?lang=En&n=D9553AB5-1.]
Gregow
,
H.
,
K.
Ruosteenoja
,
N.
Pimenoff
, and
K.
Jylhä
,
2011
:
Changes in the mean and extreme geostrophic wind speeds in northern Europe until 2100 based on nine global climate models
.
Int. J. Climatol.
,
32
,
1834
1846
.
Jiang
,
Y.
,
Y.
Luo
,
Z.
Zhao
,
Y.
Shi
,
Y.
Xu
, and
J.
Zhu
,
2010
:
Projections of wind changes for 21st century in China by three regional climate models
.
Chin. Geogr. Sci.
,
20
,
226
235
.
Lambert
,
S. J.
, and
J. C.
Fyfe
,
2006
:
Changes in winter cyclone frequencies and strengths simulated in enhanced greenhouse warming experiments: Results from the models participating in the IPCC diagnostic exercise
.
Climate Dyn.
,
26
,
713
728
.
Lopes
,
A.
,
S.
Oliveira
,
M.
Fragoso
,
J. A.
Andrade
, and
P.
Pedro
,
2009
: Wind risk assessment in urban environments: The case of falling trees during windstorm events in Lisbon. Bioclimatology and Natural Hazards, K. Strelcová et al., Eds., Springer, 55–74.
McInnes
,
K. L.
,
T. A.
Erwin
, and
J. M.
Bathols
,
2011
:
Global climate model projected changes in 10 m wind speed and direction due to anthropogenic climate change
.
Atmos. Sci. Lett.
,
12
,
325
333
.
Moyer
,
M.
,
2009
:
The way the wind blows
.
Sci. Amer.
,
301
,
27
28
.
Najac
,
J.
,
J.
Boé
, and
L.
Terray
,
2009
:
A multi-model ensemble approach for assessment of climate change impact on surface winds in France
.
Climate Dyn.
,
32
,
615
634
.
Najac
,
J.
,
C.
Lac
, and
L.
Terray
,
2011
:
Impact of climate change on surface winds in France using a statistical-dynamical downscaling method with mesoscale modelling
.
Int. J. Climatol.
,
31
,
415
430
.
Nikulin
,
G.
,
E.
Kjellström
,
U.
Hansson
,
G.
Strandberg
, and
A.
Ullerstig
,
2011
:
Evaluation and future projections of temperature, precipitation and wind extremes over Europe in an ensemble of regional climate simulations
.
Tellus
,
63A
,
41
55
.
PCMDI
, cited
2010
: About the WCRP CMIP3 multi-model dataset archive. [Available online at http://www-pcmdi.llnl.gov/ipcc/about_ipcc.php.]
Pinto
,
J. G.
,
C. P.
Neuhaus
,
G. C.
Leckebusch
,
M.
Reyers
, and
M.
Kerschgens
,
2010
:
Estimation of wind storm impacts over western Germany under future climate conditions using a statistical–dynamical downscaling approach
.
Tellus
,
62A
,
188
201
.
Pryor
,
S. C.
,
J. T.
Schoof
, and
R. J.
Barthelmie
,
2005
:
Climate change impacts on wind speeds and wind energy density in northern Europe: Empirical downscaling of multiple AOGCMs
.
Climate Res.
,
29
,
183
198
.
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,
doi:10.1029/2006GL026000
.
Pryor
,
S. C.
,
R. J.
Barthelmie
,
N. E.
Clausen
,
M.
Drews
,
N.
MacKellar
, and
E.
Kjellström
,
2012a
:
Analyses of possible changes in intense and extreme wind speeds over northern Europe under climate change scenarios
.
Climate Dyn.
,
38
,
189
208
.
Pryor
,
S. C.
,
R. J.
Barthelmie
, and
J. T.
Schoof
,
2012b
:
Past and future wind climates over the contiguous USA based on the North American Regional Climate Change Assessment Program model suite
.
J. Geophys. Res.
,
117
, D19119,
doi:10.1029/2012JD017449
.
Rauthe
,
M.
,
M.
Kunz
, and
C.
Kottmeier
,
2010
:
Changes in wind gust extremes over central Europe derived from a small ensemble of high resolution regional climate models
.
Meteor. Z.
,
19
,
299
312
.
Schwierz
,
C.
,
P.
Köllner-Heck
,
E. Z.
Mutter
,
D. N.
Bresch
,
P.-L.
Vidale
,
M.
Wild
, and
C.
Schär
,
2010
:
Modelling European winter wind storm losses in current and future climate
.
Climatic Change
,
101
,
485
514
.
Shen
,
S. S. P.
,
P.
Dzikowski
,
G.
Li
, and
D.
Griffith
,
2001
:
Interpolation of 1961–97 daily temperature and precipitation data onto Alberta polygons of ecodistrict and soil landscapes of Canada
.
J. Appl. Meteor.
,
40
,
2162
2177
.
Thomas
,
M. K.
,
1953
:
Climatological Atlas of Canada.
National Research Council, 256 pp.
Yip
,
T. C.
,
H.
Auld
, and
W.
Dnes
,
1995
: Updating wind pressures for the 1995 National Building Code of Canada. Proc. Ninth Int. Conf. on Wind Engineering, New Delhi, India, International Association for Wind Engineering, 1871–1877.