1. Introduction
Extreme snowstorms are common in North America (e.g., Changnon and Changnon 2005; G. Chen et al. 2019). They produce catastrophic weather such as heavy snowfall, strong winds, low temperature, freezing rain, and storm surge, which cause significant damages to infrastructure and threaten human welfare and lives. For example, 12 United States snowstorms in 1980–2013 averaged $2.9 billion in losses (Smith and Matthews 2015). Although the general temperature increase associated with global warming tends to reduce the snowfall, the snowfall of extreme snowstorms is less affected (O’Gorman 2014; Zarzycki 2018). Moreover, regions that used to be dominated by ice-phased precipitation could have more near-0°C precipitation, wherein the increase of freezing rain/drizzle (Lambert and Hansen 2011; Groisman et al. 2016) and rain-on-snow (e.g., Jeong and Sushama 2018; Musselman et al. 2018) incur even higher threats to traffic safety and electricity delivery.
The northeast United States (NEUS) is of our particular interest for two reasons. First, the majority of United States extreme snowstorms tend to happen over this region (Changnon and Changnon 2005, 2006). Second, several mega-cities (such as Boston, Massachusetts; New York City, New York; Philadelphia, Pennsylvania; and Washington, D.C.) with large populations and economies are located along the coast (i.e., close to the climatological storm tracks), making the region even more vulnerable to impacts from extreme snowstorms (Kocin and Uccellini 2004). As the global temperature increased during the period of 1965–2005, the regional-mean winter climate in the NEUS was observed to have higher temperature, fewer snow-covered days, and less snowfall (Burakowski et al. 2008); the modeling study by Ashley et al. (2020) further showed that both the frequency and affected area of snowstorms tend to be significantly decreased by future warming in the twenty-first century. However, for the longer time period of 1901–2000, the NEUS was observed to have more occurrences of extreme snowstorms, which is in line with the increasing trend of cyclonic activity (Changnon et al. 2006). It is projected by climate models that both the number and the intensity of winter extratropical cyclones over the land area along the NEUS coast could be increased by global warming despite the fact that the cyclone track density over the western Atlantic storm track has tended to decrease (e.g., Colle et al. 2013). Given the conflicting results, more work is needed to understand how extreme storms affecting the NEUS may respond to future climate change.
Two questions are usually raised when investigating extreme weather within the context of climate change using climate models. The first is how climate change affects the occurrence frequency of the given synoptic weather patterns. While this question is straightforward, it is difficult to answer because of the limitation of model capabilities. Global climate models (GCMs) are frequently used to simulate the long-term climate change in a given scenario and do not require the supply of lateral boundary forcing (e.g., Lambert and Hansen 2011; Colle et al. 2013, 2015; Tsou et al. 2016; G. Chen et al. 2019). However, GCMs generally employ coarse resolution, and thus could have large uncertainties in the regional climate and high-frequency variability. For example, the coarse resolution tends to simulate extratropical cyclone tracks too close to the coast of NEUS (Colle et al. 2015), and this bias is still evident when the horizontal resolution is around 25 km (G. Chen et al. 2019). Particularly, the precipitation is even sensitive to the terrain height, so GCMs have difficulties over certain places due to the coarse horizontal resolution. In contrast, regional climate models (RCMs) employ higher resolution, but they cannot simulate the long-term climate change by themselves and rely on prescribed lateral boundary forcing. They usually dynamically downscale GCM simulations to provide detailed regional information (e.g., Huang et al. 2016a,b, 2019; Jeong and Sushama 2018). Nevertheless, this direct downscaling inevitably transfers GCM unrealistic daily (Lynn et al. 2009a) and mesoscale (such as mesoscale terrain forcing, gust fronts, drylines, and MCSs) variance to RCM simulations, which could yield large uncertainties in the simulation of extreme weather.
The second question is how a certain large-scale synoptic weather pattern (e.g., a cyclone) would change if it happened in a future climate scenario, focusing on effects of changes in the large-scale dynamic and thermodynamic environment on synoptic-scale weather behaviors. The approach of pseudo–global warming (PGW; Schär et al. 1996; Kimura and Kitoh 2007), or mean-signal nesting as for the broader concept of climate change, is frequently used to investigate this question in many studies [e.g., on regional precipitation (Frei et al. 1998; Sato et al. 2007; Hara et al. 2008; Rasmussen et al. 2011; Lackmann 2013; Musselman et al. 2018; Ashley et al. 2020); on the East Asian summer monsoon (Kawase et al. 2009; Jung et al. 2015); on winter extratropical cyclones over the North Atlantic (Marciano et al. 2015; Michaelis et al. 2017); on hurricanes (Lynn et al. 2009b; Mallard et al. 2013; Lackmann 2015; Gutmann et al. 2018; Jung and Lackmann 2019); and on extreme tornadic storms (Trapp and Hoogewind 2016)]. This approach combines advantages of both GCM and RCM simulations. It extracts the change of mean climate state from GCM historical and future simulations, superimposes it on the historical climate events in reanalysis to project future events, and then uses RCMs to simulate the historical events and projected events. As a result, the simulation of historical events is close to the observation because of more realistic high-frequency variance in reanalysis; and the differences shown in the simulation of projected events are caused and can be explained by effects of mean climate change.
This study aims to investigate responses of extreme snowstorms affecting the NEUS to future global warming using the PGW approach. G. Chen et al. (2019) identified 110 extreme snowstorms causing heavy snowfall over mega-cities along the NEUS coast during 1980–2015 and found that the High Resolution Atmospheric Model (HiRAM) could well capture the observed characteristics of occurrence and circulation of those events in the historical simulation. Therefore, this study extracted mean-climate change from HiRAM historical and future simulations and dynamically downscaled those events using the WRF Model with a high horizontal resolution (5 km) for comparative analyses. The events, unevenly distributed in 35 continuous winters (November–March), exhibited remarkable variability in snowfall coverage and the synoptic patterns, so the modeling results, especially the statistical characteristics, are robust and representative of global warming effects on regional weather extremes.
Compared with previous PGW studies by Marciano et al. (2015) and Michaelis et al. (2017) that were focused on dynamic responses of extratropical cyclones associated with snowstorms, this study is mainly concerned about surface precipitation changes in aspects of spatial distribution, intensity, and phase (i.e., rain or snow). Meanwhile, because of the high horizontal resolution, this study simulates better topography and thus more reliable regional precipitation changes than studies based on coarser-resolution GCMs (e.g., Colle et al. 2013; Zarzycki 2018), which are valuable for policy makers to take possible mitigation and adaptation measures.
The rest of this paper is arranged as follows: section 2 describes the HiRAM simulation data and provides an overview of the mean-climate change over the eastern United States caused by global warming; section 3 depicts the setup of WRF dynamic downscaling and evaluates the simulated surface precipitation against observation and reanalysis data; and section 4 presents the main results, followed by the summary and discussion in section 5.
2. HiRAM historical and future simulations
a. Model configuration
The HiRAM historical (1979–2015) and future (2075–2100) simulations used in this study were conducted by and archived at the Research Center for Environmental Change (RCEC), Academia Sinica of Taiwan. While details of the model configuration and simulation setup can be found in Tsou et al. (2016) and G. Chen et al. (2019), a brief description is given below.
In both simulations, the HiRAM, described in Zhao et al. (2009) and Chen and Lin (2011), was set up with a C384 grid, corresponding to a horizontal resolution of about 25 km, and 32 vertical levels with the model top at 10 hPa, and was driven with prescribed sea surface temperature (SST) and sea ice concentration (SIC) using the time-slice method (Bengtsson et al. 1996). In the historical simulation, SST and SIC were from the monthly Hadley Center SST1 data (1° × 1°; Rayner et al. 2003); in the future simulation, SST and SIC were composed using a PGW approach following Mizuta et al. (2008, 2014) and Kusunoki (2018), which superimposed three components: 1) the future change of 2075–2100 versus 1979–2004 projected by the multimodel mean of 28 models that participated in phase 5 of the Coupled Model Intercomparison Project (CMIP5) for the representative concentration pathway 8.5 (RCP8.5) scenario; 2) the linear trend during 2075–2100 projected by the multimodel ensemble; and 3) the detrended monthly anomalies during 1979–2004 from the observation. Greenhouse gas concentrations and aerosol forcing were set up accordingly to be consistent with the simulation periods and scenarios.
Because of its high resolution, the HiRAM model has exhibited good performance on many climate topics in historical and future climate regimes [e.g., on tropical cyclones Chen and Lin 2011, 2013; Tsou et al. 2016); extreme precipitation over East Asia (Freychet et al. 2017a); seasonal precipitation change in western North Pacific and East Asia (C.-A. Chen et al. 2019); and snowfall over the European Alps (Freychet et al. 2017b]. Particularly for climate over the NEUS, Zhao et al. (2009) showed that the model-simulated year-to-year variations in Atlantic hurricane frequency were correlated with observations with a coefficient of 0.83, and the simulated upward trend of hurricane frequency over North Atlantic during 1981–2005 was also consistent with observations; Hsu and Chen (2020) showed that the simulated atmospheric rivers along the North American northeast coast were close to observation in occurrence frequency, intensity, and circulation characteristics; and our previous study (G. Chen et al. 2019) showed that the model well captured the observed characteristics of extreme snowstorms affecting the NEUS in aspects of occurrence frequency, subseasonal and multiple-year variations, snow spatial coverage, and the associated circulation patterns. Therefore, it is believed that HiRAM simulations are trustworthy for estimating the change of future climate.
Compared with PGW studies based on ensemble mean of multiple CMIP5 models, this study heavily depends on one model, which is a compromise between simulations that are more realistic and affordable computation cost. Results of dynamic downscaling are highly influenced by the large-scale background flows used to drive the regional model. If the models providing the background flow could not well capture the characteristics of synoptic disturbances associated with the snowstorms, the downscaled results could not reflect the finer structures such as precipitation embedded in the disturbances. As shown in our previous studies (G. Chen et al. 2019; Hsu and Chen 2020), the HiRAM well captured the characteristics of these synoptic disturbances, partly because of the fine spatial scale at 25 km. CMIP5 models were typically conducted with the 100–200-km spatial scale and were likely to miss certain important characteristics of background states, such as temperature and moisture gradients and related vertical structures even in the long-term mean state changes that were used in the PGW experiments. Ensemble mean changes of many CMIP5 models would further smooth the mean states. In view of this drawback, we decided to use the HiRAM projection as the background flows for driving the regional model. Ideally, regional downscaling driven by the outputs of multiple high-resolution models should be conducted. However, long-term climate change projections using high-resolution models are limited and data are not easy to get. Even if the data were available, it would take tremendous amount of computing time to complete a suite of downscaling simulations driven by the outputs of several high-resolution models. Thus, the approach taken in this study is still the most affordable method for PGW simulations.
b. Change of mean climate state in global warming
Figure 1 presents the change of November–March mean climate state over the eastern United States caused by global warming under the RCP8.5 scenario (2075–99 minus 1980–2004). Clearly, the spatial distribution of warming is not uniform. The lower troposphere (e.g., 850 hPa in Fig. 1b) is warmed more at high latitudes than at low latitudes and more over land than over the oceans. The magnitude and spatial pattern of warming are similar to results shown in Maloney et al. (2014), with the maximum warming at the surface located in Canada (greater than 10°C; figure not shown). In contrast, the upper troposphere (e.g., 300 hPa in Fig. 1a) is more warmed at low latitudes than at high latitudes. The vertical and meridional structures of warming (Fig. 1c) are consistent with those in IPCC AR5 (Collins et al. 2013), implying that the atmospheric instability tends to be enhanced at high latitudes (including the NEUS this study concerned) but reduced at low latitudes. Meanwhile, it is also noticed that the warming is more remarkable in the early winter (November–December) than the late winter (January–March) (Fig. 1d).
Change of November–March mean climate state over the eastern United States caused by global warming estimated from HiRAM simulations (future minus historical): (a),(b) changes of temperature (color shadings; K) and relative humidity (contours; %) at 300 and 850 hPa, (c) temperature change along the meridional cross section averaged between 79° and 70°W, and (d) subseasonal variation of temperature changes over the coastal NEUS [37°–44°N, 79°–70°W; dashed rectangles in (a) and (b)]. Black dots in (a) and (c) indicate positions of four coastal mega-cities (Boston, New York City, Philadelphia, and Washington, D.C.) successively from north to south.
Citation: Journal of Climate 34, 6; 10.1175/JCLI-D-20-0197.1
The change of relative humidity (ΔRH) is generally small (contours in Figs. 1a,b), with patterns similar to the multimodel mean of CMIP5 results given in Liu et al. (2016) and Laîné et al. (2014). For lower levels (e.g., 850 hPa) where most water vapor resides, ΔRH is no more than ±4% for most regions, showing a slight decrease over land and slight increase over the oceans. At higher levels (e.g., 300 hPa) where the water vapor content is small, ΔRH becomes bigger and can be as much as ±8%, showing a decrease at low latitudes and an increase at high latitudes. This is because the saturation mixing ratio of water vapor is sensitive to temperature changes. However, it is worth noting that the water vapor content is increased at all levels in fact, and the relative increase is larger (smaller) over regions with more (less) warming. For example, the specific humidity at 850 hPa is increased by 30%–40% over the continental NEUS and 20%–30% over the oceans to the east (figure not shown). This tends to enhance the latent heat release, invigorate the upward vertical motion, and increase the precipitation.
With the temperature increase, the 0°C contour is shifted northward, which may greatly change the phase (i.e., solid or liquid) of precipitation associated with winter storms. However, it is worth emphasizing that the northward retreat of 0°C in storms is not as much as that in the mean climate state. To illustrate this, we examined the meridional distribution of temperature at the surface averaged between 79° and 70°W (Fig. 2c). In the mean climate state (dashed lines), the 0°C line is shifted northward by 4.9° at the surface (dashed blue vs dashed red). By contrast, for the average of the 93 snowstorm events, the northward shift of 0°C is only 1.9° (solid blue vs solid red). This is because the meridional temperature gradient in storm days is usually much larger than that in nonstorm days, which dominates the mean climate state. Also note that the zone with temperature between −5° and 0°C, the temperature range favoring freezing precipitation (Stuart and Isaac 1999), is shifted northward and markedly expanded. In historical events (solid blue), the zone covers an area of 1.6° at the surface; in the future events (solid red), the area is enlarged to 3.1°. The southern area tends to have less occurrence of temperature falling in this range while the northern area, especially the inland area, tends to have more (Figs. 2a,b). This shift means that freezing precipitation may decrease in the south, increase in the north, and threaten a broader area, which will be shown below in section 4a. Specifically, at four coastal mega-cities, the occurrence frequency of temperature in this range clearly decreases at all four cities (Fig. 2d).
Temperature differences at the surface between the 93 historical events (Table S1 in the online supplemental material) based on the ERA-Interim data (ERA_historical) and the projected future events (ERA_future; i.e., historical events superimposed with differences between HiRAM historical and future simulations): (a),(b) temperature at the surface (contours) and number of days when at least one of the 6-hourly temperature data falls in the range from −5° to 0°C (shadings), (c) temperature averaged between 79° and 70°W, and (d) probability distribution of near-surface temperature at four cities in historical (blue) and future (red) events. In (a) and (b), the blue (red) lines indicate 0°C contours in historical (future) cases while black circles represent the 50-km boundaries from centers of four coastal mega-cities; in (d), ticks indicate the 0th, 25th, 50th, 75th, and 100th percentiles, and filled dots indicate means.
Citation: Journal of Climate 34, 6; 10.1175/JCLI-D-20-0197.1
3. WRF dynamical downscaling
a. Observational datasets
Station-observed snowfall and total precipitation data (mm day−1) from Global Historical Climatology Network (GHCN) daily data (Menne et al. 2012a,b) in 1980–2015 were used to define extreme snow events and validate simulated snowfall and total precipitation at the surface. To be consistent with G. Chen et al. (2019), only stations where the total length of snowfall records is longer than 10 years (120 months) were included (station locations given in Fig. 3).
WRF domain setup. The horizontal grid spacing is 15 km in D1 and 5 km in D2 (solid black rectangle). Black circles represent the 50-km boundaries from centers of four coastal mega-cities; blue dots indicate GHCN stations (1971 stations in total) whose observed snowfall and precipitation were used to select events and evaluate WRF historical simulations. The simulated snowfall over the coastal NEUS (37°–44°N, 79°–70°W; dashed rectangle) tends to be significantly underestimated for 17 discarded events.
Citation: Journal of Climate 34, 6; 10.1175/JCLI-D-20-0197.1
The ERA-Interim reanalysis (Dee et al. 2011; https://rda.ucar.edu/datasets/ds627.0/) was used to provide initial conditions and lateral boundary forcing for WRF simulations shown below. It has a horizontal resolution of 0.703° longitude × 0.703° latitude and 36 vertical levels. Additionally, the 12-h forecasts of snow water equivalent (SWE; mm day−1) and total precipitation (mm day−1) in this dataset were also used for validating simulated surface precipitation.
b. Case setup
The simulated events were based on the observed extreme snowstorm events in 1980–2015 from G. Chen et al. (2019). In the study, an extreme event was defined as a day when one or more of the four coastal mega-cities—Boston, New York City, Philadelphia, and Washington, D.C.—have GHCN observed daily snowfall exceeding the respective local 95th percentile thresholds (i.e., 137, 118, 105, and 111 mm for cities from north to south) of daily snowfall amount. There were 110 events in total. Therein, 17 events were discarded because the simulated snowfall tends to be significantly underestimated over the coastal NEUS, making them no longer representative of extreme snowstorms (see details in the online supplemental material). For each of the remaining 93 events, two cases were simulated with WRF using the PGW method: a historical case and a future case. The historical simulations were driven by the original ERA-Interim data, while the future simulations were driven by ERA-Interim data that were superimposed with mean-climate change estimated from HiRAM simulations using the following procedure. First, the multiyear monthly-mean climate state was calculated for each HiRAM simulation using 25-yr data (1980–2004 in the historical simulation versus 2075–99 in the future simulation) and linearly interpolated to each day of the year; then, daily differences in surface temperature (Tsrf) and profiles of temperature (T), relative humidity (RH), zonal and meridional winds (u and υ), and geopotential height (ph) between the two HiRAM simulations were added to corresponding ERA fields during the extreme events. Additionally, greenhouse gas concentrations in two cases were prescribed with the values of 2005 and 2090, respectively.
Unlike many other PGW studies that maintain constant RH between historical and future simulations (e.g., Lackmann 2013; Marciano et al. 2015; Michaelis et al. 2017; Jung and Lackmann 2019), this study chose to alter RH based on GCM simulations, similar to Liu et al. (2016). In doing so, water vapor does not scale according to the Clausius–Clapeyron relation. As shown in section 2b, RH is decreased over the land but increased over the oceans by the warming, which is opposite to the land–ocean contrast of temperature changes (i.e., more warming over land than over the oceans). As a result, the RH changes not only directly affect the surface precipitation (shown below in section 4a), but also further decrease the land–ocean thermodynamic contrast, which affects the cyclonic evolution and the surface precipitation.
c. WRF model configuration
The WRF model version 3.9.1 was used for dynamical downscaling. There were two two-way nested domains (Fig. 3). The outer domain (D1) had 200 × 170 grids with a horizontal resolution of 15 km while the inner domain (D2) had 259 × 259 grids with a horizontal resolution of 5 km. Both domains used the default setup of 38 vertical levels with the model top at 50 hPa (a setup of 50 vertical levels was also tested and did not show any nonnegligible change in the precipitation). Each simulation started one day before the extreme-snow day, leaving one day for model spinup and one day for analysis. Results in D1 and D2 were stored every 6 h and 3 h, respectively.
The CONUS physics suite was used for major physical parameterizations. It consists of the Thompson cloud microphysics scheme (Thompson et al. 2008), the modified Tiedtke cumulus convection scheme (only in D1) (Tiedtke 1989; Zhang et al. 2011), the RRTMG shortwave and longwave radiation scheme (Iacono et al. 2008), the Mellor–Yamada–Janjic PBL scheme (Janjić 1994), the Monin–Obukhov (Janjic) surface-layer scheme (Janjić 2001), and the unified Noah land surface scheme (Chen and Dudhia 2001). SST was held constant throughout the simulation. In addition, the spectral nudging was applied in the outer domain (D1) throughout the simulation as in Hara et al. (2008) and Liu et al. (2016). The horizontal wind speeds (u and υ), temperature (T), and geopotential height (ph) above the boundary layer were nudged with the nudging coefficient of 0.0003, and the wavenumber was set to 2 for both directions (i.e., the wavelength is 1500 km in the west–east direction and 1275 km in the south–north direction). The use of spectral nudging was to constrain the model internal variability such as circulation adjustment, forcing the temperature and circulation fields consistent with the enforced data, and thus focus the analysis on precipitation changes. Meanwhile, it saves the necessity of longer spinup time in simulations and suppressing uncertainties associated with surface properties other than the surface temperature (such as surface albedo and snowpack), whose future changes were not considered in our approach, although effects of these properties on precipitation should be minor in short-term WRF simulations (e.g., Cedilnik et al. 2012; Kumar et al. 2014).
d. Evaluation of surface precipitation
For each event, we compared the simulated total precipitation/snow in D2 with those from the GHCN observation and the ERA 12-h forecast, where the observation and forecast data were interpolated to WRF D2 grids. Hereafter, if not clearly stated otherwise, precipitation refers to total precipitation [i.e., the sum of liquid (rain) and solid precipitation], while snow refers to total snow (i.e., the sum of solid precipitation consisting of snow, ice, and graupel from the Thompson microphysics parameterization). Only land grids were included because GHCN data were only available over land.
Figure 4 compares the spatial distributions of simulated precipitation and snow with those from GHCN observations and ERA data. For the mean of 93 events, WRF simulations show larger precipitation and snow than the other two datasets and have larger spatial correlations with the ERA data (0.97 for total precipitation and 0.94 for snow) than with the GHCN observations (0.88 for both total precipitation and snow). For individual events, the WRF–GHCN correlation is larger than 0.6 for both total precipitation and snowfall in more than 50% of events, with means of 0.60 and 0.66; by contrast, the WRF–ERA correlation is much better, with a mean of 0.88 for total precipitation and 0.87 for snow. This indicates that the WRF simulations are mostly determined by the enforced reanalysis data. It is consistent with results of Chen et al. (2017) and confirms that the RCM dynamical downscaling is subjected to lateral boundary forcing from GCMs. In addition, this implies that the WRF–observation discrepancies are mostly attributable to uncertainties of observations and the enforced reanalysis. For example, the metric of snow is snowfall depth in GHCN but SWE in WRF simulations and ERA data, whereas the snow-to-water ratio varies significantly in the observation because of sensitivities to temperature and snow type (e.g., dry snow or wet snow). Overall, it is concluded that the WRF simulations well simulated the precipitation in both phases and spatial distributions.
Comparisons of mean total precipitation/snow from (a),(b) WRF historical simulations and with corresponding data from (c),(d) GHCN observations and (e),(f) ERA-Interim data, and (g),(h) the probability distribution of spatial correlations between WRF and GHCN/ERA results for 93 events, where ticks indicate the 0th, 25th, 50th, 75th, and 100th percentiles and filled dots indicate means. GHCN snowfall in (d) was converted to SWE by dividing 10, assuming a simple 10:1 snow-to-liquid ratio; numbers in the lower-right corners of (c)–(f) indicate spatial correlation between the plot with respective WRF results.
Citation: Journal of Climate 34, 6; 10.1175/JCLI-D-20-0197.1
4. Projected precipitation changes
This section first examines the regional perspective of precipitation changes and then precipitation changes at the four coastal mega-cities, where subseasonal characteristics of responses are also discussed. All analyses were based on the statistics of 93 events. The single-event analysis was not included, as it is less meaningful due to the limited representativeness of any single event and it is beyond the scope of this article to examine the changes event by event for 93 events.
a. Regional perspective
Figure 5 presents regional changes of total precipitation for the average of 93 events. Clearly, precipitation in the future is decreased in the northern coastal area but increased over the southern area and the north inland area (Fig. 5a). The relative change of precipitation per degree of warming is also the largest over the southern area (around ~10% K−1), decreases northward, and reaches the smallest values of around −3% K−1 at the northern coastal area (Fig. 5b). This pattern is markedly different from the values predicted by the Clausius–Clapeyron relation (Boer 1993), where the precipitation of daily extremes should increase around 7% per degree of warming. This is because the assumptions embedded in the estimation (i.e., unchanged relative humidity and constant ratio of precipitation to total water content) are both invalid in this study. Therein, the relative humidity near the surface is decreased by around 2% over this region as shown in Fig. 1, while the ratio of precipitation to total water content is changed as the upward vertical motion is enhanced in the south and reduced in the north. We demonstrated changes of vertical motion by examining the regional-mean vertical velocity w over three subregions (A, B, and C). For region A at the northern coastal area, w is significantly reduced at almost all levels (Fig. 5c), yielding smaller precipitation rate together with the decrease of relative humidity. For region B in the central NEUS, the w profiles are quite similar between historical and future simulations (Fig. 5d), so the precipitation change is mainly determined by the change of relative humidity. For region C in the south, w is significantly increased in the middle and upper troposphere for events with strong large-scale elevation (Fig. 5e), which increases the precipitation rate and overwhelms the effect of reduced relative humidity.
(a) Changes of 93-event averaged precipitation (mm; future minus historical), (b) relative changes of precipitation per degree of surface warming (% K−1), and (c)–(e) comparisons of regional-mean vertical velocity over three subregions (thick lines for 93-event mean and thin lines for 25th and 75th percentiles). The hatching in (a) indicates regions where the difference is statistically significant (Student’s t test; p value < 0.05); dots in (c)–(e) indicates levels that future results are significantly different (Student’s t test; p value < 0.05) from historical results.
Citation: Journal of Climate 34, 6; 10.1175/JCLI-D-20-0197.1
We then examined pairwise difference of surface precipitation in Fig. 6. Here, a daily precipitation increase or decrease exceeding 1 mm (Δ > 1 mm or Δ < −1 mm) is considered to be a nonnegligible change. Using the threshold of other values (e.g., 5 mm) was also tested and yielded the same conclusion. Three features are noticed. First, changes of precipitation/snow in the majority of 93 events are concentrated over the coastal area, but are minor over remote areas such as in Canada (Figs. 6g,h). Second, the snow decreases across the NEUS especially in northern coastal area (Fig. 6e) but increases over the inland area in 20–25 cases (Fig. 6b), while the rain in most cases is increased throughout the domain (Fig. 6c). Third, the total precipitation tends to be increased in the south for both coastal and inland areas and decreased in the northern coastal area (Figs. 6a,d). These indicate that the temperature increase melts snow into rain in many events and yields larger precipitation because of the associated increase in water vapor abundance and the enhancement of upward vertical motion, and that the precipitation/snow belt tends to be shifted southwestward, making the rain increase over the northern coastal area unable to offset the snow decrease.
Number of events when pairwise differences in total precipitation/total snow/rain (future minus historical) are (a)–(c) larger than 1 mm day−1, (d)–(f) smaller than −1 mm day−1, or (g)–(i) between −1 and 1 mm day−1. Numbers in the plot indicate results spatially averaged at four coastal mega-cities (black circles), where data at land grids within a distance of 50 km from city centers were included in the analysis.
Citation: Journal of Climate 34, 6; 10.1175/JCLI-D-20-0197.1
Figure 7 presents changes of precipitation intensity by comparing the number of days with light (0.1–10 mm day−1), moderate (10–20 mm day−1), and heavy (>20 mm day−1) precipitation/snow between WRF historical and future simulations. In the historical simulations, the inland area is dominated by light-precipitation/snow events while the coastal area is dominated by moderate- and heavy-precipitation/snow events. Particularly, the coastal area around Boston has the most heavy-snow events (Fig. 7k), which is consistent with observation results shown in G. Chen et al. (2019). In the future simulations, the changes differ among three types of regions. First, most regions of the NEUS have fewer light-precipitation events, more heavy-precipitation events, and fewer snow events of all intensities. This indicates that precipitation intensity is systematically increased because of the enhanced upward vertical motion and water vapor content associated with the temperature increase (Colle et al. 2013, 2015). Second, the northern coastal area (e.g., Boston) has more light-precipitation/snow events and fewer heavy-precipitation/snow events. Third, the northern inland area (e.g., western New York and northern Pennsylvania) may have more heavy-snow events. Changes over these two regions are caused by the snow melting and the southwestward shift of the snowband shown above.
Number of events with (a)–(d) light (0.1–10 mm day−1), (e)–(h) moderate (10–20 mm day−1), and (i)–(l) heavy (>20 mm day−1) total precipitation/snow in WRF historical simulations and their future changes. Numbers in the plot indicate results spatially averaged at four coastal mega-cities (black circles), where data at land grids within a distance of 50 km from city centers were included in the analysis; the hatching indicates regions where the difference of frequency is significantly different (χ2 test; p value < 0.05) between two sets of simulations.
Citation: Journal of Climate 34, 6; 10.1175/JCLI-D-20-0197.1
For winter storms, in addition to the precipitation amount, the precipitation phase is also important: rain causes floods, snow hinders traffic, and mixed rain and snow or freezing precipitation creates hazardous road and flight conditions, and damages plant life and power lines. Therefore, we also compared the number and intensity of events with different precipitation phases between historical and future simulations.
Figure 8 presents comparisons of events with snow-only, mixed, and rain-only precipitation. Here, days with rain more than 0.1 mm and snow less than 0.1 mm were defined as rain events, days with rain less than 0.1 mm and snow more than 0.1 mm as snow events, and days with rain and snow both more than 0.1 mm as mixed events. In historical simulations, snow events are the most frequent type in the domain, especially for the northern and inland area, while the mixed and rain events are frequent over the southern and coastal areas. All three types of events are the most intense at the coastal area. When temperature increases in the future, snowstorms are decreased in frequency across the NEUS, with reduced intensity over most area and enhanced intensity over northern inland area and the coastal area around Philadelphia and Washington D.C. This indicates that the warming can transform weak snowstorms into rainstorms but has less effect on strong snowstorms in some circumstances, consistent with findings of O’Gorman (2014) and Zarzycki (2018). Rainstorms are increased in frequency and enhanced in precipitation across the domain. The most noteworthy changes are that the mixed storms are increased in frequency in the north, affecting a broader area, and enhanced in intensity over vast areas except the northern coastal area. This is consistent with the expansion of the zone with temperature near 0°C as shown above and implies serious threatens to the NEUS.
Number and intensity (mm day−1) of events with (a)–(d) snow, (e)–(h) rain and snow mixed, and (i)–(l) rain in WRF historical simulations and their future changes. Rainstorms never happen to the northern area in historical simulations in (i), so rain intensity over this region [in (k)] and its future changes [in (l)] could not be calculated. Numbers in the plot indicate results spatially averaged at four coastal mega-cities (black circles), where data at land grids within a distance of 50 km from city centers were included in the analysis; the hatching indicates regions where the difference is significantly different (χ2 test for frequency and Student’ t test for intensity; p value < 0.05) between two sets of simulations.
Citation: Journal of Climate 34, 6; 10.1175/JCLI-D-20-0197.1
Figure 9 shows the number and the mean intensity of events with freezing precipitation in two WRF simulations. Freezing precipitation refers to the phenomenon that occurs when supercooled rain or drizzle droplets fall on the surface and freeze instantly. The supercooled droplets are formed in two ways. In one way, snow or ice from cold clouds gets melted when falling through a warm layer in the atmosphere, gets cooled quickly below in a near-surface layer with temperature below 0°C, and reaches the surface before refreezing. In the other way, there is not a warm melting layer between the precipitation source zones and the ground, and the freezing droplets are formed directly via coalescence without an ice phase in clouds with temperature everywhere lower than 0°C (Huffman and Norman 1988; Stuart and Isaac 1999). The second may be the dominant one, but it is rare over the NEUS and mainly contributes freezing drizzle (Rauber et al. 2000). To be simple, this study defined events with freezing precipitation as days when the situations of surface rainfall more than 0.1 mm, snowfall less than 0.1 mm, and the near-surface temperature within a range from −5° to 0°C (Stuart and Isaac 1999) were simultaneously detected once or more times in the 3-hourly output. The approach may have larger uncertainties than sophisticated algorithms that examine sounding profiles (e.g., Bourgouin 2000; Schuur et al. 2012; Reeves et al. 2014) or high-frequency model output (e.g., Benjamin et al. 2016), but as least the results can indicate the occurrence probability of freezing precipitation.
(a),(b) Number and (c),(d) mean intensity (mm) of events with freezing precipitation in WRF (left) historical and (right) future simulations. Numbers in the plot indicate results spatially averaged at four coastal mega-cities (black circles), where data at land grids within a distance of 50 km from city centers were included in the analysis; the hatching indicates regions where the difference is significantly different (χ2 test for frequency and Student’s t test for intensity; p value < 0.05) between two sets of simulations.
Citation: Journal of Climate 34, 6; 10.1175/JCLI-D-20-0197.1
In historical simulations, freezing precipitation is concentrated over two regions centered at 36° and 40°N, with frequency less than 12 days of the 93 events and intensity weaker than 4 mm over most area. When the temperature increases, the frequency is reduced in southern and coastal areas and expanded significantly northward while the intensity is increased over most areas. This means that more regions are exposed to the risks, and that higher-level protective measures should be taken for possible damages. The patterns are similar to changes of events with mixed precipitation shown above. This is because the freezing precipitation and the mixed precipitation both favor the temperature range near 0°C, which is northward shifted and expanded as shown above.
b. City perspective
Changes of precipitation at four coastal cities were given in Figs. 6–9 with numbers. In most storms, global warming increases total precipitation at New York City, Philadelphia, and Washington, D.C but decreases it at Boston and decreases total snow at all four cities (Fig. 6). As a result, the three southern cities tend to have more and heavier precipitation but less snowfall while Boston has a reduction in frequency and intensity of both total precipitation and snowfall (Fig. 7). All four cities tend to have fewer snow-only events and more rain-only and mixed events, with intensities of all three types of events increased at Philadelphia and Washington, D.C. and decreased at Boston and New York City (Fig. 8). The occurrence frequency of freezing precipitation is decreased at three southern cities and unchanged in Boston, whereas the mean intensity is increased at all four cities (Fig. 9).
Figures 10 and 11 present the subseasonal characteristics of changes in precipitation at four cities. The changes in precipitation intensity and phase are consistent in all five months (November–March) rather than dominated by changes in a single month, the early winter, or the late winter. Thus, more warming in early winter than in late winter seems to have no effects on the simulated precipitation changes. One noteworthy thing is that freezing precipitation favors the early winter in historical simulations but the late winter in future simulations (Fig. 11, right column), causing increased occurrence of freezing precipitation at Boston.
Number of events with light, moderate, and heavy total precipitation/snow in individual months (November, December, January, February, and March) at four coastal mega-cities from WRF historical (blue-edged) and future (red-edged) simulations. In this figure and Fig. 11, land data within 50 km from city centers were first spatially averaged, and then the resulting 3-hourly time series were used to diagnose events at individual cities. Because of the different procedures, data in Figs. 10 and 11 may be slightly different from those in Figs. 6–9. Plus signs in red-edged bars indicate the frequencies of respective precipitation types in future simulations are significantly different (χ2 test; p value < 0.05) from those in historical simulations.
Citation: Journal of Climate 34, 6; 10.1175/JCLI-D-20-0197.1
As Fig. 10, but for (left) different precipitation phases (i.e., snow, rain and snow mixed, and rain) and (right) freezing precipitation from WRF historical (blue-edged) and future (red-edged) simulations.
Citation: Journal of Climate 34, 6; 10.1175/JCLI-D-20-0197.1
5. Summary and discussion
This study investigates effects of global warming on extreme snowstorms along the NEUS coast by conducting and comparing dynamical-downscaling WRF simulations driven with and without the mean climate change signal extracted from HiRAM historical and future simulations. The 93 observed snowstorms in 1980–2015 documented in G. Chen et al. (2019) were adopted as surrogates for studying the potential effects. Results show that the number of events with moderate and heavy daily snowfall (SWE greater than 10 mm day−1) at Boston, New York City, Philadelphia, and Washington, D.C. is decreased from 60, 54, 43, and 30 to 32, 29, 30, and 20, respectively (Fig. 7). In addition, although the rainfall increases in all four cities, total precipitation increases in the south (especially along the coast), where the rainfall is increased due to the warming-caused increase of water vapor content and upward vertical motion, but decreases in the north (especially along the coast), where the rainfall increase is overwhelmed by the snow decrease (Figs. 5–7). As a result, the precipitation distribution exhibits a southwestward shift. Another notable effect is that the frequency of mixed rain and snow and freezing precipitation events is increased in the north of NEUS (Figs. 8 and 9), attributed to the northward retreat of the 0°C zone and the expansion of the zone with temperature near 0°C.
It is worth pointing out that the changes of individual events caused by future warming could differ among events and are not always consistent with the multiple-event mean changes, although the superimposed forcing of mean climate change is very similar and only has slight subseasonal variation. For example, the changes of cyclone tracks, which affect the spatial distribution of surface precipitation (e.g., Changnon et al. 2008), vary significantly among events (shown in Fig. 12). This is similar to results found in hurricanes as shown by Gutmann et al. (2018), indicating that some inherent and subtle features that distinguish the events from one another must exist and be main causes for the different responses. Nevertheless, it is clear that cyclone centers in ~2/3 of the events are shifted westward (54 events) or southward (67 events). These changes are also consistent with the southwest shift of the precipitation distribution. A more detailed investigation of cyclone changes including intensity and center locations as well as the underlying mechanisms is warranted.
Position changes of cyclone center between WRF historical and future simulations. The starting and ending points of arrows indicate cyclone centers from historical and future simulations, respectively; blue and red plus signs indicate the 93-event averaged cyclone center. The cyclone center showed a westward (southward) shift caused by global warming in 54 (67) events. The procedure for diagnosing cyclone centers is given in the online supplemental material.
Citation: Journal of Climate 34, 6; 10.1175/JCLI-D-20-0197.1
Acknowledgments
GC and WCW acknowledge the support by the U.S. National Science Foundation (1545917) in support of the Partnership for International Research and Education project at the University at Albany. CTC acknowledges the support by the Ministry of Sciences and Technology (Taiwan) under MOST 108-2621-M-865-001. The HiRAM simulations and the effort of HHH were supported by the Ministry of Sciences and Technology (Taiwan) under MOST 107-2119-M-001-010 and MOST 108-2119-M-001-014. We would also like to thank the National Center for High-Performance Computing in Taiwan for providing computing resources. We appreciate very much the constructive comments and suggestions from the reviewers, which greatly improved the presentation. Simulation data used in this study will be available upon request to Dr. Wei-Chyung Wang (wcwang@albany.edu).
REFERENCES
Ashley, W. S., A. M. Haberlie, and V. A. Gensini, 2020: Reduced frequency and size of late-twenty-first-century snowstorms over North America. Nat. Climate Change, 10, 539–544, https://doi.org/10.1038/s41558-020-0774-4.
Bengtsson, L., M. Botzet, and M. Esch, 1996: Will greenhouse gas-induced warming over the next 50 years lead to higher frequency and greater intensity of hurricanes? Tellus, 48A, 57–73, https://doi.org/10.3402/tellusa.v48i1.11632.
Benjamin, S. G., J. M. Brown, and T. G. Smirnova, 2016: Explicit precipitation-type diagnosis from a model using a mixed-phase bulk cloud–precipitation microphysics parameterization. Wea. Forecasting, 31, 609–619, https://doi.org/10.1175/WAF-D-15-0136.1.
Boer, G. J., 1993: Climate change and the regulation of the surface moisture and energy budgets. Climate Dyn., 8, 225–239, https://doi.org/10.1007/BF00198617.
Bourgouin, P., 2000: A method to determine precipitation types. Wea. Forecasting, 15, 583–592, https://doi.org/10.1175/1520-0434(2000)015<0583:AMTDPT>2.0.CO;2.
Burakowski, E. A., C. P. Wake, B. Braswell, and D. P. Brown, 2008: Trends in wintertime climate in the northeastern United States: 1965–2005. J. Geophys. Res., 113, D20114, https://doi.org/10.1029/2008JD009870.
Cedilnik, J., D. Carrer, J.-F. Mahfouf, and J.-L. Roujean, 2012: Impact assessment of daily satellite-derived surface albedo in a limited-area NWP model. J. Appl. Meteor. Climatol., 51, 1835–1854, https://doi.org/10.1175/JAMC-D-11-0163.1.
Changnon, D., C. Merinsky, and M. Lawson, 2008: Climatology of surface cyclone tracks associated with large central and eastern U.S. snowstorms, 1950–2000. Mon. Wea. Rev., 136, 3193–3202, https://doi.org/10.1175/2008MWR2324.1.
Changnon, S. A., and D. Changnon, 2005: Snowstorm catastrophes in the United States. Global Environ. Change, 6B, 158–166, https://doi.org/10.1016/j.hazards.2006.06.001.
Changnon, S. A., and D. Changnon, 2006: A spatial and temporal analysis of damaging snowstorms in the United States. Nat. Hazards, 37, 373–389, https://doi.org/10.1007/s11069-005-6581-4.
Changnon, S. A., D. Changnon, and T. R. Karl, 2006: Temporal and spatial characteristics of snowstorms in the contiguous United States. J. Appl. Meteor. Climatol., 45, 1141–1155, https://doi.org/10.1175/JAM2395.1.
Chen, C.-A., H.-H. Hsu, C.-C. Hong, P.-G. Chiu, C.-Y. Tu, S.-J. Lin, and A. Kitoh, 2019: Seasonal precipitation change in the western North Pacific and East Asia under global warming in two high-resolution AGCMs. Climate Dyn., 53, 5583–5605, https://doi.org/10.1007/s00382-019-04883-1.
Chen, F., and J. Dudhia, 2001: Coupling an advanced land surface–hydrology model with the Penn State–NCAR MM5 modeling system. Part I: Model implementation and sensitivity. Mon. Wea. Rev., 129, 569–585, https://doi.org/10.1175/1520-0493(2001)129<0569:CAALSH>2.0.CO;2.
Chen, G., J. Yang, Q. Bao, and W.-C. Wang, 2017: Intraseasonal responses of the East Asia summer rainfall to anthropogenic aerosol climate forcing. Climate Dyn., 51, 3985–3998, https://doi.org/10.1007/s00382-017-3691-0.
Chen, G., W.-C. Wang, L. Tao, H.-H. Hsu, C.-Y. Tu, and C.-T. Cheng, 2019: Extreme snow events along the coast of the northeast United States: Analysis of observations and HiRAM simulations. J. Climate, 32, 7561–7574, https://doi.org/10.1175/JCLI-D-18-0874.1.
Chen, J.-H., and S.-J. Lin, 2011: The remarkable predictability of inter-annual variability of Atlantic hurricanes during the past decade. Geophys. Res. Lett., 38, L11804, https://doi.org/10.1029/2011GL047629.
Chen, J.-H., and S.-J. Lin, 2013: Seasonal predictions of tropical cyclones using a 25-km-resolution general circulation model. J. Climate, 26, 380–398, https://doi.org/10.1175/JCLI-D-12-00061.1.
Colle, B. A., Z. Zhang, K. A. Lombardo, E. Chang, P. Liu, and M. Zhang, 2013: Historical evaluation and future prediction of eastern North American and western Atlantic extratropical cyclones in the CMIP5 models during the cool season. J. Climate, 26, 6882–6903, https://doi.org/10.1175/JCLI-D-12-00498.1.
Colle, B. A., J. F. Booth, and E. K. M. Chang, 2015: A review of historical and future changes of extratropical cyclones and associated impacts along the US east coast. Curr. Climate Change Rep., 1, 125–143, https://doi.org/10.1007/s40641-015-0013-7.
Collins, M., and Coauthors, 2013: Long-term climate change: Projections, commitments and irreversibility. Climate Change 2013: The Physical Science Basis, T. F. Stocker et al., Eds., Cambridge University Press, 1029–1136.
Dee, D. P., 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.
Frei, C., C. Schär, D. Lüthi, and H. C. Davies, 1998: Heavy precipitation processes in a warmer climate. Geophys. Res. Lett., 25, 1431–1434, https://doi.org/10.1029/98GL51099.
Freychet, N., and Coauthors, 2017a: Variability of hydrological extreme events in East Asia and their dynamical control: A comparison between observations and two high-resolution global climate models. Climate Dyn., 48, 745–766, https://doi.org/10.1007/s00382-016-3108-5.
Freychet, N., H.-H. Hsu, A. Duchez, and C.-Y. Tu, 2017b: Projection in snowfall characteristics over the European Alps and its sensitivity to the SST changes: Results from a 50 km resolution AGCM. Atmos. Sci. Lett., 18, 261–267, https://doi.org/10.1002/asl.751.
Groisman, P. Ya., O. N. Bulygina, X. Yin, R. S. Vose, S. K. Gulev, I. Hanssen-Bauer, and E. Førland, 2016: Recent changes in the frequency of freezing precipitation in North America and northern Eurasia. Environ. Res. Lett., 11, 045007, https://doi.org/10.1088/1748-9326/11/4/045007.
Gutmann, E. D., and Coauthors, 2018: Changes in hurricanes from a 13-yr convection-permitting pseudo–global warming simulation. J. Climate, 31, 3643–3657, https://doi.org/10.1175/JCLI-D-17-0391.1.
Hara, M., T. Yoshikane, H. Kawase, and F. Kimura, 2008: Estimation of the impact of global warming on snow depth in Japan by the pseudo-global-warming method. Hydrol. Res. Lett., 2, 61–64, https://doi.org/10.3178/hrl.2.61.
Hsu, H.-H., and Y.-T. Chen, 2020: Simulation and projection of circulations associated with atmospheric rivers along the North American northeast coast. J. Climate, 33, 5673–5695, https://doi.org/10.1175/JCLI-D-19-0104.1.
Huang, W.-R., Y.-H. Chang, H.-H. Hsu, C.-T. Cheng, and C.-Y. Tu, 2016a: Dynamical downscaling simulation and future projection of summer rainfall in Taiwan: Contributions from different types of rain events. J. Geophys. Res. Atmos., 121, 13 973–13 988, https://doi.org/10.1002/2016JD025643.
Huang, W.-R., Y.-H. Chang, C.-T. Cheng, H.-H. Hsu, and C.-Y. Tu, 2016b: Summer convective afternoon rainfall simulation and projection using WRF driven by global climate model. Part I: Over Taiwan. Terr. Atmos. Oceanic Sci., 27, 659–671, https://doi.org/10.3319/TAO.2016.05.02.01.
Huang, W.-R., P.-H. Huang, Y.-H. Chang, C.-T. Cheng, H.-H. Hsu, C.-Y. Tu, and A. Kitoh, 2019: Dynamical downscaling simulation and future projection of extreme precipitation activities in Taiwan during the mei-yu seasons. J. Meteor. Soc. Japan, 97, 481–499, https://doi.org/10.2151/jmsj.2019-028.
Huffman, G. J., and G. A. J. Norman, 1988: The supercooled warm rain process and the specification of freezing precipitation. Mon. Wea. Rev., 116, 2172–2182, https://doi.org/10.1175/1520-0493(1988)116<2172:TSWRPA>2.0.CO;2.
Iacono, M. J., J. S. Delamere, E. J. Mlawer, M. W. Shephard, S. A. Clough, and W. D. Collins, 2008: Radiative forcing by long-lived greenhouse gases: Calculations with the AER radiative transfer models. J. Geophys. Res., 113, D13103, https://doi.org/10.1029/2008JD009944.
Janjić, Z. I., 1994: The step-mountain Eta coordinate model: Further developments of the convection, viscous sublayer, and turbulence closure schemes. Mon. Wea. Rev., 122, 927–945, https://doi.org/10.1175/1520-0493(1994)122<0927:TSMECM>2.0.CO;2.
Janjić, Z. I., 2001: Nonsingular implementation of the Mellor-Yamada level 2.5 scheme in the NCEP Meso model. NCEP Office Note 437, 61 pp.
Jeong, D. I., and L. Sushama, 2018: Rain-on-snow events over North America based on two Canadian regional climate models. Climate Dyn., 50, 303–316, https://doi.org/10.1007/s00382-017-3609-x.
Jung, C.-Y., and G. M. Lackmann, 2019: Extratropical transition of Hurricane Irene (2011) in a changing climate. J. Climate, 32, 4847–4871, https://doi.org/10.1175/JCLI-D-18-0558.1.
Jung, C.-Y., H.-J. Shin, C. J. Jang, and H.-J. Kim, 2015: Projected change in East Asian summer monsoon by dynamic downscaling: Moisture budget analysis. Asia-Pac. J. Atmos. Sci., 51, 77–89, https://doi.org/10.1007/s13143-015-0061-x.
Kawase, H., T. Yoshikane, M. Hara, F. Kimura, T. Yasunari, B. Ailikun, H. Ueda, and T. Inoue, 2009: Intermodel variability of future changes in the Baiu rainband estimated by the pseudo global warming downscaling method. J. Geophys. Res., 114, D24110, https://doi.org/10.1029/2009JD011803.
Kimura, F., and A. Kitoh, 2007: Downscaling by pseudo global warming method. Final Report of the ICCAP, Research Institute for Humanity and Nature (RIHN), 4 pp.
Kocin, P. J., and L. W. Uccellini, 2004: A snowfall impact scale derived from northeast storm snowfall distributions. Bull. Amer. Meteor. Soc., 85, 177–194, https://doi.org/10.1175/BAMS-85-2-Kocin.
Kumar, P., B. K. Bhattacharya, R. Nigam, C. M. Kishtawal, and P. K. Pal, 2014: Impact of Kalpana-1 derived land surface albedo on short-range weather forecasting over the Indian subcontinent. J. Geophys. Res. Atmos., 119, 2764–2780, https://doi.org/10.1002/2013JD020534.
Kusunoki, S., 2018: Future changes in precipitation over East Asia projected by the global atmospheric model MRI-AGCM3.2. Climate Dyn., 51, 4601–4617, https://doi.org/10.1007/s00382-016-3499-3.
Lackmann, G. M., 2013: The south-central U.S. flood of May 2010: Present and future. J. Climate, 26, 4688–4709, https://doi.org/10.1175/JCLI-D-12-00392.1.
Lackmann, G. M., 2015: Hurricane Sandy before 1900 and after 2100. Bull. Amer. Meteor. Soc., 96, 547–560, https://doi.org/10.1175/BAMS-D-14-00123.1.
Laîné, A., H. Nakamura, K. Nishii, and T. Miyasaka, 2014: A diagnostic study of future evaporation changes projected in CMIP5 climate models. Climate Dyn., 42, 2745–2761, https://doi.org/10.1007/s00382-014-2087-7.
Lambert, S. J., and B. K. Hansen, 2011: Simulated changes in the freezing rain climatology of North America under global warming using a coupled climate model. Atmos.–Ocean, 49, 289–295, https://doi.org/10.1080/07055900.2011.607492.
Liu, C., and Coauthors, 2016: Continental-scale convection-permitting modeling of the current and future climate of North America. Climate Dyn., 49, 71–95, https://doi.org/10.1007/s00382-016-3327-9.
Lynn, B. H., R. Healy, and L. M. Druyan, 2009a: Quantifying the sensitivity of simulated climate change to model configuration. Climatic Change, 92, 275–298, https://doi.org/10.1007/s10584-008-9494-x.
Lynn, B. H., R. Healy, and L. M. Druyan, 2009b: Investigation of Hurricane Katrina characteristics for future, warmer climates. Climate Res., 39, 75–86, https://doi.org/10.3354/cr00801.
Mallard, M. S., G. M. Lackmann, A. Aiyyer, and K. Hill, 2013: Atlantic hurricanes and climate change. Part I: Experimental design and isolation of thermodynamic effects. J. Climate, 26, 4876–4893, https://doi.org/10.1175/JCLI-D-12-00182.1.
Maloney, E. D., and Coauthors, 2014: North American climate in CMIP5 experiments: Part III: Assessment of twenty-first-century projections. J. Climate, 27, 2230–2270, https://doi.org/10.1175/JCLI-D-13-00273.1.
Marciano, C. G., G. M. Lackmann, and W. A. Robinson, 2015: Changes in U.S. east coast cyclone dynamics with climate change. J. Climate, 28, 468–484, https://doi.org/10.1175/JCLI-D-14-00418.1.
Menne, M. J., I. Durre, R. S. Vose, B. E. Gleason, and T. G. Houston, 2012a: An overview of the Global Historical Climatology Network–Daily database. J. Atmos. Oceanic Technol., 29, 897–910, https://doi.org/10.1175/JTECH-D-11-00103.1.
Menne, M. J., and Coauthors, 2012b: Global Historical Climatology Network–Daily (GHCN-Daily), version 3.22. NOAA National Climatic Data Center, accessed 19 September 2017, https://doi.org/10.7289/V5D21VHZ.
Michaelis, A. C., J. Willison, G. M. Lackmann, and W. A. Robinson, 2017: Changes in winter North Atlantic extratropical cyclones in high-resolution regional pseudo–global warming simulations. J. Climate, 30, 6905–6925, https://doi.org/10.1175/JCLI-D-16-0697.1.
Mizuta, R., Y. Adachi, S. Yukimoto, and S. Kusunoki, 2008: Estimation of the future distribution of sea surface temperature and sea ice using the CMIP3 multi-model ensemble mean. Meteorological Research Institute, Tech. Rep. 56, 28 pp., https://www.mri-jma.go.jp/Publish/Technical/DATA/VOL_56/56.html.
Mizuta, R., O. Arakawa, T. Ose, S. Kusunoki, H. Endo, and A. Kitoh, 2014: Classification of CMIP5 future climate responses by the tropical sea surface temperature changes. SOLA, 10, 167–171, https://doi.org/10.2151/sola.2014-035.
Musselman, K. N., F. Lehner, K. Ikeda, M. P. Clark, A. F. Prein, C. Liu, M. Barlage, and R. Rasmussen, 2018: Projected increases and shifts in rain-on-snow flood risk over western North America. Nat. Climate Change, 8, 808–812, https://doi.org/10.1038/s41558-018-0236-4.
O’Gorman, P. A., 2014: Contrasting responses of mean and extreme snowfall to climate change. Nature, 512, 416–418, https://doi.org/10.1038/nature13625.
Rasmussen, R., and Coauthors, 2011: High-resolution coupled climate runoff simulations of seasonal snowfall over Colorado: A process study of current and warmer climate. J. Climate, 24, 3015–3048, https://doi.org/10.1175/2010JCLI3985.1.
Rauber, R. M., L. S. Olthoff, M. K. Ramamurthy, and K. E. Kunkel, 2000: The relative importance of warm rain and melting processes in freezing precipitation events. J. Appl. Meteor., 39, 1185–1195, https://doi.org/10.1175/1520-0450(2000)039<1185:TRIOWR>2.0.CO;2.
Rayner, N. A., D. E. Parker, E. B. Horton, C. K. Folland, L. V. Alexander, D. P. Rowell, E. C. Kent, and A. Kaplan, 2003: Global analyses of sea surface temperature, sea ice, and night marine air temperature since the late nineteenth century. J. Geophys. Res., 108, 4407, https://doi.org/10.1029/2002JD002670.
Reeves, H. D., K. L. Elmore, A. Ryzhkov, T. Schuur, and J. Krause, 2014: Sources of uncertainty in precipitation-type forecasting. Wea. Forecasting, 29, 936–953, https://doi.org/10.1175/WAF-D-14-00007.1.
Sato, T., F. Kimura, and A. Kitoh, 2007: Projection of global warming onto regional precipitation over Mongolia using a regional climate model. J. Hydrol., 333, 144–154, https://doi.org/10.1016/j.jhydrol.2006.07.023.
Schär, C., C. Frei, D. Lüthi, and H. C. Davies, 1996: Surrogate climate-change scenarios for regional climate models. Geophys. Res. Lett., 23, 669–672, https://doi.org/10.1029/96GL00265.
Schuur, T. J., H.-S. Park, A. V. Ryzhkov, and H. D. Reeves, 2012: Classification of precipitation types during transitional winter weather using the RUC model and polarimetric radar retrievals. J. Appl. Meteor. Climatol., 51, 763–779, https://doi.org/10.1175/JAMC-D-11-091.1.
Smith, A. B., and J. L. Matthews, 2015: Quantifying uncertainty and variable sensitivity within the US billion-dollar weather and climate disaster cost estimates. Nat. Hazards, 77, 1829–1851, https://doi.org/10.1007/s11069-015-1678-x.
Stuart, R. A., and G. A. Isaac, 1999: Freezing precipitation in Canada. Atmos.–Ocean, 37, 87–102, https://doi.org/10.1080/07055900.1999.9649622.
Thompson, G., P. R. Field, R. M. Rasmussen, and W. D. Hall, 2008: Explicit forecasts of winter precipitation using an improved bulk microphysics scheme. Part II: Implementation of a new snow parameterization. Mon. Wea. Rev., 136, 5095–5115, https://doi.org/10.1175/2008MWR2387.1.
Tiedtke, M., 1989: A comprehensive mass flux scheme for cumulus parameterization in large-scale models. Mon. Wea. Rev., 117, 1779–1800, https://doi.org/10.1175/1520-0493(1989)117<1779:ACMFSF>2.0.CO;2.
Trapp, R. J., and K. A. Hoogewind, 2016: The realization of extreme tornadic storm events under future anthropogenic climate change. J. Climate, 29, 5251–5265, https://doi.org/10.1175/JCLI-D-15-0623.1.
Tsou, C.-H., P.-Y. Huang, C.-Y. Tu, C.-T. Chen, T.-P. Tzeng, and C.-T. Cheng, 2016: Present simulation and future typhoon activity projection over western North Pacific and Taiwan/east coast of China in 20-km HIRAM climate model. Terr. Atmos. Oceanic Sci., 27, 687–703, https://doi.org/10.3319/TAO.2016.06.13.04.
Zarzycki, C. M., 2018: Projecting changes in societally impactful northeastern U.S. snowstorms. Geophys. Res. Lett., 45, 12 067–12 075, https://doi.org/10.1029/2018GL079820.
Zhang, C., Y. Wang, and K. Hamilton, 2011: Improved representation of boundary layer clouds over the southeast Pacific in ARW-WRF using a modified Tiedtke cumulus parameterization scheme. Mon. Wea. Rev., 139, 3489–3513, https://doi.org/10.1175/MWR-D-10-05091.1.
Zhao, M., I. M. Held, S.-J. Lin, and G. A. Vecchi, 2009: Simulations of global hurricane climatology, interannual variability, and response to global warming using a 50-km resolution GCM. J. Climate, 22, 6653–6678, https://doi.org/10.1175/2009JCLI3049.1.
