1. Introduction
In the cold regions located from the mid- to high latitudes, snowfall is a common occurrence and covers the land surface during winter. The solid precipitation particles, constituting the snowfall, exhibit diverse habits generated from the various cloud microphysical processes (Kikuchi et al. 2013; Kobayashi 1961; Mason 1971; Pruppacher and Klett 1997). Key contributors to determining these habits among various cloud microphysical processes are the riming, deposition, and melting. The temperature and supersaturation during depositional growth determine the shape of solid precipitation particles, as indicated by numerous earlier studies (Magono and Lee 1966; Kikuchi et al. 2013; Kobayashi 1961; Hueholt et al. 2022; Fukuta and Takahashi 1999; Pruppacher and Klett 1997; Bailey and Hallett 2009; Mason 1971; Nakaya 1954; Hallett and Mason 1958).
Riming, occurring actively within the temperature range of −5° to −15°C (von Blohn et al. 2009; Magono and Lee 1966), results in rimed particles or graupel with complicated shapes compared to the pure crystals generated through depositional growth. Furthermore, the melting process, triggered when the temperature exceeds 0°C, significantly alters the habit of the solid precipitation particles. The reports of the earlier studies outlined above indicate that the habit of these particles is largely dependent on the temperature and supersaturation in clouds and thus exhibits regional variability (Hueholt et al. 2022; Mizuno 1992).
The characteristics of these particles are closely related to the formation of weak layers in the snowpack, with unrimed solid precipitation particles (Schaerer and McClung 2006) and heavily rimed particles such as graupel (Abe 2004) contributing to snow avalanches after snow accumulation over these weak layers. This underscores the crucial link between the characteristics of the solid precipitation particles and the probability of snow avalanches (Nakamura 2019). Consequently, understanding these characteristics, particularly their habits, is imperative for effective disaster prevention in cold regions.
In Hokkaido, a northern region of Japan (Fig. 1a), where winter snowfall and snowpack coverage extend from late November to April (Japan Meteorological Agency 2023a), the characteristics of the solid precipitation particles, including their habits, differ significantly from other regions in Japan due to the colder temperatures. The anticipated increase in temperature due to climate change is expected to impact the characteristics of the solid precipitation particles in Hokkaido. While previous studies examined future changes in snowfall in Japan, including Hokkaido, and focused on aspects such as snowfall amount (Matsumura and Sato 2011; Kawase et al. 2016, 2021), phase changes (Matsumura and Sato 2011; Kawase et al. 2016, 2021; Kawazoe et al. 2020), heavy snowfall frequency (e.g., Inatsu et al. 2021; Kawazoe et al. 2020), and snowpack amount (Katsuyama et al. 2020), none have focused on the prospective changes in the characteristics of solid precipitation particles.
This study aimed to examine future changes in the characteristics of solid precipitation particles, particularly their habits and growth processes, in Hokkaido. Achieving this goal necessitates employing a meteorological model coupled with a sophisticated cloud microphysical model explicitly predicting the habits of solid precipitation particles (Chen and Lamb 1994a,b; Hashino and Tripoli 2007; Shima et al. 2020, 2009). However, these models require a significant amount of computational resources compared to those of the cloud microphysical models typically used in regional climate models. To mitigate this problem, Hashimoto et al. (2020) developed a framework known as the process tracking model (PTM). PTM predicts the mass of hydrometeors generated by the riming process and depositional growth for each cloud hydrometeor category treated in bulk microphysical models (e.g., cloud ice, snow, and graupel), as summarized in Text S1 in the online supplemental material. The cloud microphysical processes treated in PTM are shown in Table 1. According to Nakaya (1954), the habit of the snowfall particles generated by each process is roughly associated with the categories shown in Table 1. In nature, snowfall particles have various habits that do not necessarily correspond to typical habits but often represent a mixture of habits. PTM represents this mixture by the mass fraction of the depositional growth and riming. Thus, the large mass fraction of each category does not directly correspond to the habit of the solid precipitation particles. To clarify this difference in the meaning of the Habit in Table 1 and habits of the solid precipitation particles, we henceforth capitalize the first letter of each habit in Table 1 as a proper noun.
Prognostic variables of PTM for each solid hydrometeor category (i.e., cloud ice, snow, and graupel) and its corresponding habits (Nakaya 1954). T and Si represent temperature (°C) and supersaturation over ice (%), respectively. Habit corresponds to Plate when −17° < T < −14°C and Si < 7% and when −20° < T < −17°C or −14° < T < 10°C regardless of Si.
Hashimoto et al. (2020) applied PTM to a meteorological model developed by the JMA (Saito et al. 2006) and evaluated the validity of PTM while examining the mass fraction of riming to the total mass of solid precipitation over Hokuriku, located in the central part of Japan (Fig. 1a). This study attempts to comprehend the mass fraction of depositional growth and the riming process to the total mass of the solid precipitation over Hokkaido in the current climate. Additionally, we aim to investigate future changes in the characteristics of solid precipitation particles in Hokkaido using PTM coupled with a meteorological model.
2. Methodology
a. Model configuration
This study used the Scalable Computing for Advanced Library and Environment (SCALE; Nishizawa et al. 2015; Sato et al. 2015) model (version 5.3.6) and also implemented PTM in SCALE (Hashimoto et al. 2020). This model solves the nonhydrostatic three-dimensional fully compressible governing equation. The prognostic variables are total density (ρ), three-dimensional momentum, density-weighted potential temperature, and density-weighted tracer (ρq). The tracer, q, includes specific humidity, mass ratio, number concentration, and mass of hydrometeor in each category, generated by the microphysical processes detailed in Table 1. Spatial discretization utilized a third-order upward and a second-order central difference scheme for advection and other terms, respectively. To prevent negative values, the positive definitive scheme of Koren (1993) was applied for tracer advection. Temporal integration employed the horizontally explicit and vertically implicit (HEVI) scheme.
A two-moment bulk cloud microphysical model by Seiki and Nakajima (2014), coupled with PTM, simulated cloud effects, categorizing cloud hydrometeors into cloud, rain, cloud ice, snow, and graupel. PTM calculated the mass of each solid precipitation category generated by the depositional growth and the riming process shown in Table 1. It is thought that the distribution, mass of the hydrometeor, and mass fraction of each category of the hydrometeor are largely dependent upon the cloud microphysical model (e.g., Kondo et al. 2021; Sato et al. 2015). Therefore, a cloud microphysical scheme that can reproduce the measured microphysical properties of precipitation particles must be selected. The validity of the two-moment bulk cloud microphysical model for simulating winter snow clouds over Hokkaido was confirmed via comparisons with disdrometer data and satellite observations according to Kondo et al. (2021).
The effects of radiation and turbulence were calculated using the k-distribution-based radiative transfer model, MSTRN-X (Sekiguchi and Nakajima 2008), and the Mellow–Yamada–Nakanishi–Niino level 2.5 scheme (Nakanishi and Niino 2006). Temperature and soil moisture in the land for urban and nonurban areas were calculated using a bucket-type model and a single-layer urban canopy model (Kusaka et al. 2001). The surface flux was calculated based on the bulk coefficients of Beljaars and Holtslag (1991) and Wilson (2001). The ocean and land surface temperatures were obtained from external data, as described in the next section.
b. Experimental setup
Dynamical downscaling simulations (DDSs) were conducted to examine the characteristics of solid precipitation particles over Hokkaido during winter. The simulation period spanned from December 2020 to February 2021 (90 days). The representativeness of the target period will be discussed in section 4 and Text S2. Each daily simulation commenced at 1800 UTC of the previous day, with a duration of 30 h, resulting in a total of 90 simulations. The first 6 h of each simulation were not analyzed. The calculation domain covered the entire Hokkaido area (Fig. 1a), enabling us to simulate snow clouds generated over the Sea of Japan, influenced by the northwest winter monsoon. The horizontal grid spacing was 1 km, and the number of vertical layers was 57, with layer thickness increasing from 40 m at the bottom of the model to 650 m at the model top. Initial and boundary data for DDS were derived from mesoscale analysis (MANL) data provided by the JMA. The horizontal resolution, vertical layer number, and time interval of MANL are 5 km, 50, and 3 h, respectively. The horizontal wind vector, air density, temperature, and relative humidity calculated from specific humidity and air density, soil moisture, soil temperature, and skin temperature, including sea surface temperature (SST) data from MANL, were used. The experimental setup described above used the same methodologies of several previous studies (Inatsu et al. 2020; Kondo et al. 2021; M. Kondo 2024, unpublished material), where the validity of the model with respect to surface precipitation amount and the geographical distribution of surface precipitation were confirmed.
To investigate future changes in solid precipitation particles, a pseudo–global warming experiment (PGWE) was conducted (Schär et al. 1996; Wu and Lynch 2000; Liu et al. 2017; Sato et al. 2007; Adachi and Tomita 2020). In this study, the monthly average temperature difference between the future climate and present-day (PD) conditions was averaged and added to the MANL data without changing the relative humidity. Consequently, water vapor increased in accordance with the Clausius–Clapeyron relationship as the temperature rose in the PGWE. The temperature difference between the future climate and PD was calculated based on the difference between the 4-K experiment and the historical (HIST) experiment using d4PDF data (Mizuta et al. 2017), which were constructed based on the results of the 60-yr integration. This dataset comprises ensemble simulations with a 60-km horizontal grid spacing of the atmospheric general circulation model MRI-AGCM, version 3.2 (Mizuta et al. 2012).
The 4-K experiment assumes that the global mean temperature is 4 K warmer than the nonwarming simulation (Mizuta et al. 2017). This simulation included the same conditions as the HIST experiment except for the greenhouse gas concentration, which was the same as that in 1850. A 4 K increase in global mean temperature corresponds to the global mean temperature in 2090 simulated by the representative concentration pathway 8.5. Details on each experiment are shown in the literature (Ishii and Mori 2020; Mizuta et al. 2017). We selected 12 ensembles from the HIST and 4-K experiments, which included 100 and 90 ensembles, respectively. The temperature differences due to global warming were defined as the difference in the three-dimensional temperature anomaly between the 4-K and HIST experiments. Hereafter, the PGWE will be denoted as 4K, and the control DDS experiment will be denoted as PD.
The difference in the temperature profile averaged over Hokkaido (i.e., 137.5°–147.5°E, 40°–47.5°N) was added to the MANL data, with the difference in 2-m height temperature in each grid added to the surface temperature of MANL. The mean profile of the temperature difference and the horizontal distribution of the difference in 2-m height temperature for each month are shown in Fig. 2.
c. Measurement data
Daily surface precipitation measured by the Automated Meteorological Data Acquisition System (AMeDAS) operated by the JMA was used to confirm the validity of the model. The measured daily precipitation at 272 sites over the Hokkaido area (Fig. 3a) was compared with the precipitation simulated by the model.
3. Results
a. Precipitation amount
The surface precipitation amount is compared between the observations and the model to assess the validity of the model. Figure 3 reveals the daily precipitation amount measured by the AMeDAS averaged throughout the target period and simulated by the model averaged over all 90 simulations. The model reproduced the geographical distribution of the precipitation amount over the Hokkaido area. A large amount of precipitation was observed along the west coast, while a limited amount was observed over the eastern area (area 3 in Fig. 1b), which is leeward of the mountain during the northwest monsoon. These characteristics were captured by the model. The bias of the model was large for some sites although the difference in precipitation amount between the model and observations averaged over 272 sites was 0.69 mm day−1, which is much smaller than the difference between the PD and 4K experiments. The validity of the model was established through its reasonable simulation of the surface precipitation amount under PD conditions. Therefore, the model meets the prerequisite for conducting the PGWE to estimate the impacts of climate change on surface precipitation.
Next, we discuss the distribution of surface precipitation in PD and its anticipated changes in the future to elucidate the fundamental characteristics of precipitation over Hokkaido in PD and the impacts of climate change on them. Figures 4a and 4b illustrate the geographical distribution of daily liquid precipitation amount and daily solid precipitation amount at the surface, averaged over the entire 90 simulations. Solid precipitation is defined as surface precipitation resulting from cloud ice, snow, and graupel, while liquid precipitation is defined as that resulting from clouds and rain. The figures clarify that surface precipitation over the analysis area is mainly produced by solid precipitation. Additionally, they highlight that liquid precipitation mostly occurred over the southern part of the analysis area. The precipitation amount was notably large along the west coast of Hokkaido, and the precipitation amount was larger in the northwest compared to that in the southwest. Contrastingly, the precipitation amount was smaller in the eastern part of Hokkaido. The difference in precipitation among the regions aligned with typical circulation patterns contributed to snowfall. According to reanalysis data from 1960 to 2019, snowfall in Hokkaido was largely derived from snow clouds developing over the Sea of Japan due to the influence of the northwest monsoon (Inatsu et al. 2021). In such cases, substantial snowfall was typically observed on the windward side of the mountain (Fig. 1b), specifically the west to northwest region of Hokkaido. Contrastingly, snowfall over the east of Hokkaido primarily originated from the extratropical cyclones intermittently passing through the eastern part of Hokkaido, although this circulation pattern was less frequent than the northwest monsoon pattern (Inatsu et al. 2021).
Figures 4c and 4d illustrate the differences in liquid and solid precipitation amounts between the PD and 4K experiments. The liquid precipitation amount was consistently higher in 4K than in PD throughout Hokkaido. Contrastingly, the solid precipitation amount in 4K was notably smaller in most areas, especially over the west coast of Hokkaido, where the solid precipitation amount was large in PD (Fig. 4b). These changes directly resulted from the temperature increase in the future. In the 4K experiments, surface air temperatures were 3–8 K higher than those in PD (Figs. 2a–c). Under these warmer conditions, surface temperatures more frequently exceeded 0°C across a wide area, leading to the transition from solid precipitation in PD to liquid precipitation in 4 K. The daily mean liquid, solid, and total precipitation amounts (Fig. 5) support this transition, evident in the positive and negative differences in liquid and solid precipitation. Alongside this transition, the total precipitation amount was significantly larger in 4K compared to that in PD. The large precipitation amount in 4K was mainly attributed to the increase in water vapor, aligning with reports from earlier studies (e.g., Kawase et al. 2021; Matsumura and Sato 2011).
The increase in the total precipitation amounts and the transition from solid-to-liquid precipitation were also observed in areas 1–3, as defined in Fig. 1b (Figs. 5a–c). These changes were more pronounced in area 1 (Fig. 5a), northwest part of Hokkaido, where climatological precipitation amounts were large.
b. Process tracking
1) Present-day experiment
Figure 6 displays the geographical distribution of the daily mean surface precipitation generated by depositional growth, and the riming process is detailed in Table 1. Surface mass precipitation by depositional growth at T < −36°C is not shown in the figure because it was much smaller than that of the other processes (i.e., depositional growth at −36°C < T < 0°C, and Riming). The figure indicates that the mass fraction of the depositional growth of Columnar (generating temperature range from −36° to −20°C) had the largest surface solid precipitation over a wide area of Hokkaido (Fig. 6e). The mass fraction of the depositional growth of Plate (generating temperature range from −20° to −10°C) in the surface solid precipitation was the second largest (Fig. 6c). The mass fraction of the riming process (Riming) and depositional growth of Dendrite (generating temperature range from −17° to −14°C and S > 7%) followed as the third and fourth largest solid surface precipitation, respectively (Figs. 6d,f). The characteristics of the breakdown of the mass fraction are reflected in the geographical distribution of the dominant microphysical processes for the surface precipitation. The Columnar-dominated area was the most widely distributed, and the Plate-dominated area was the second most distributed over the analysis domain (Fig. 7).
The breakdown of the mass fraction was closely related to the temperature of the cloud layer. Figure 8 displays the vertical profile of the mixing ratio of the total hydrometeors (qhyd), defined as the sum of the mixing ratio of cloud, rain, cloud ice, snow, and graupel, and the vertical profile of the relative humidity over ice in cloudy grids. The figure also shows the height of domain averaged temperature of T = 0°, −4°, −5°, −10°, −15°, −20°, and −36°C, which are the critical temperatures for the microphysical processes shown in Table 1. The cloud layer was mostly distributed at −36°C < T < −5°C, which is indicated as the height between dotted red and dot–dashed gray lines. Supersaturation over ice of the cloud layer was smaller than 7% in most cloud layers. Temperature T in the upper part of the cloud layers (i.e., above the peak height of qhyd), where the depositional growth mainly occurs, was in the range of −36°C < T < −15°C. In this situation, the depositional growth of Columnar (−36° to −20°C) or Plate (−20° to −10°C) tended to occur frequently. Additionally, the temperature range that is suitable for the riming process, i.e., −15°C < T < −5°C (von Blohn et al. 2009; Magono and Lee 1966), was only distributed in the lower part of clouds. Therefore, the mass fraction of Riming was relatively small compared with that of the depositional growth of Plate (−20° to −10°C) and Columnar (−36° to −20°C) to the surface solid precipitation.
Figure 7 also indicates that Riming showed the largest mass fraction of the surface precipitation around the east of Hidaka Mountain whose location is shown in Fig. 1b. The snowfall in this area is mainly caused by the extratropical cyclone (Inatsu et al. 2021). According to earlier studies (Nakamura 2019; Colle et al. 2014), rimed particles are dominant in the extratropical cyclone. The results of our simulation are consistent with those of the earlier studies.
2) Future changes
The differences in the surface precipitation by each depositional growth and the riming process between 4K and PD are shown in Fig. 9. It is evident that the mass fraction of the depositional growth of Columnar (−36° to −20°C) in the surface precipitation, which represented the largest fraction of surface solid precipitation in PD, largely decreased in 4K over the west coast of Hokkaido. The mass fraction of the depositional growth by Plate (−20° to −10°C) in the surface precipitation, which represented the second largest fraction of surface precipitation in PD, also decreased in 4K in the same area, although the difference was smaller than that for Columnar. In contrast, surface precipitation by the riming process increased around these areas. The mass fraction of the depositional growth of Needle/Columnar (−10° to −4°C) in surface precipitation also increased over northeastern Hokkaido. However, the increase was smaller compared with that for the riming process.
The daily mean surface precipitation generated by each process averaged over the land of Hokkaido is shown in Fig. 10. The figure supports the increase in the mass fraction of the surface precipitation by Riming and depositional growth of Needle/Columnar (−20°C < T < −10°C) and the decrease in the mass fraction of the surface precipitation by the depositional growth of Columnar (−36° to −20°C) and Plate (−20° to −10°C) in 4K. Additionally, the decreased amount of the mass fraction in the depositional growth of Plate (−20° to −10°C) was smaller than that in the depositional growth of Columnar (−36° to −20°C) (Fig. 10). As a result of the variations in surface precipitation by each process, the mass fraction of Riming and the depositional growth of Plate (−20° to −10°C) became comparably dominant in surface solid precipitation in 4K (Fig. 10d). These variations are consistent across different areas (Figs. 10a–c). The cumulative decrease in the mass fraction in surface precipitation by the depositional growth of Columnar (−36° to −20°C) and Plate (−20° to −10°C) did not offset the increase in the mass fraction in the surface precipitation by Riming and the depositional growth of Needle/Columnar (−10° to −4°C) (Fig. 10). This was attributed to the transition from solid precipitation to liquid precipitation, as discussed in section 3a.
The geographical distribution of the dominant microphysical processes for solid precipitation in 4K reflects the variations outlined above. The riming processes became dominant over the plain area of the west and northwest coasts of Hokkaido in 4K, where solid precipitation was predominantly observed in PD (Fig. 11). The mass fraction of the depositional growth of Plate (−20° to −10°C), another dominant process in 4K, was mainly dominant over the mountain area and area 3 where the riming process occurs less frequently, as discussed later.
From these results, we can conclude that the dominant microphysical process will shift from depositional growth −36°C < T < −10°C (i.e., Columnar and Plate) to the riming process (i.e., Riming) over the plain area of the west and northwest coast of Hokkaido in the future (Fig. 11). Depositional growth with a warmer environment (i.e., −20°C < T < −10°C, Plate) will become dominant in other areas, including the mountainous region (Fig. 11). Additionally, solid precipitation will partly change into liquid precipitation in the future (Figs. 4 and 5).
The shift in dominant microphysical processes affecting surface solid precipitation was attributed to the warmer environment in 4K. Figures 8c and 8d illustrate the vertical profile of qhyd and relative humidity over ice averaged across the whole calculation domain during the entire calculation period in 4K. As the relative humidity did not exhibit a significant difference between PD and 4K, the difference in the mass fraction of the depositional growth of Dendrite (−17° to −14°C, S > 7%) was also minor. For the other microphysical processes, temperature served as a key indicator to discern differences between PD and 4K. In 4K, the altitudes of −5° and −15°C were higher than those in PD, encompassing the upper part of the clouds (i.e., above the peak height of qhyd) within the temperature range (Fig. 8c), which is recognized as a suitable temperatures range for the riming process (von Blohn et al. 2009; Magono and Lee 1966). These results indicate that the cloud hydrometeor was distributed at the altitude where the riming process can easily occur compared with that where depositional growth can easily occur. Thus, the riming process prevailed over the solid precipitation in the plain area in the west and northwest of Hokkaido in 4K.
Over the mountainous area, which is relatively more leeward of the winter monsoon compared to that of the coastal areas, depositional growth was dominant in the 4K experiment. Despite the cloud layer being distributed in the temperature range suitable for the riming process in 4K, the amount of supercooled water was limited over the mountainous area compared to that in the coastal area (Harimaya and Kanemura 1995). This small amount of supercooled water inhibits the riming process (Magono et al. 1966), even within a suitable temperature range for riming. Consequently, depositional growth dominated in surface precipitation over the mountainous area in both the PD and 4K experiments, with the warmer temperature (the depositional growth of Plate; −20° to −10°C) dominant in 4K.
c. Intraseasonal variability
This section explores the seasonal variability in future changes in surface solid precipitation. Figure 12 indicates the difference in the daily liquid and solid precipitation amounts between 4K and PD for each month. Generally, an increase in liquid precipitation and a decrease in solid precipitation were observed in almost all analysis areas for all months except the north and northeast parts of Hokkaido in January (Fig. 12d). In January, the solid precipitation amount was large in 4K compared to that in PD, and the liquid precipitation amount did not increase over these areas. This distinct trend in January resulted from increased water vapor and low temperatures compared to that in other months. In 4K, atmospheric water vapor was higher than in PD. In midwinter (January), temperatures were lower than in other months (not shown), causing surface temperatures to remain below 0°C even in the warmer conditions in 4K, especially over the north and northeast parts of Hokkaido. Due to these low temperatures, the increase in water vapor resulted in an increase in solid precipitation over these areas.
The difference in the mass fraction to surface solid precipitation by depositional growth and the riming process in each month is presented in Figs. S3–S5. The general characteristics in the difference of surface precipitation by each process for each month were similar to those observed throughout the entire period (Fig. 9). The mass fraction in the surface precipitation by Riming and the depositional growth of Needle/Columnar (−10° to −4°C) increased in 4K, while that of the depositional growth of Columnar (−36° to −20°C) and Plate (−20° to −10°C) decreased. However, the difference in the mass fraction by the depositional growth of Needle/Columnar (−10° to −4°C) exhibited some intraseasonal variabilities. The difference in the depositional growth of Needle/Columnar (−10° to −4°C) in 4K was more pronounced in January over northeast Hokkaido (Fig. S4). In this area, the solid precipitation amount increased to 4K in January (Fig. 12d). These results indicate that the increase in solid precipitation amount over northeastern Hokkaido in January in 4K originated from the increased mass fraction by the depositional growth of Needle/Columnar (−10° to −4°C).
4. Discussion
In this section, we discuss the representativeness of the target period, the generality of the present findings, and the limitations of the method used in this study for generalizing the knowledge gained in this study.
a. Representativeness of the target period
According to the Japan Meteorological Agency (2023b), compared with climatological averages, the mean air temperature near the surface over Hokkaido during winter 2020/21 was 0.5 K colder, whereas the mean precipitation amount was similar. The mean air temperature over the target area at 850 hPa, which corresponds to the cloud layer, was only 0.17 K colder than the recent 63-yr average (see Text S2). These differences in surface air temperature and cloud layer temperature from the mean were much smaller than the mean temperature difference between PD and 4K shown in Fig. 2d. These results indicate that precipitation and temperature in and below the cloud layer during the target period were similar to the climatological averages, thereby confirming the representativeness of the target period. Thus, our primary finding in terms of the change in the dominant process from the depositional growth of Columnar (−36° to −20°C) to Riming should not be affected if a different target period is selected.
b. Generality of the knowledge obtained from the winter 2020/21 simulation
The main findings of this study: 1) the shift from depositional growth to riming and 2) the transition from solid to liquid precipitation in future climate, were obtained from simulating a single winter, 2020/21. Further to section 4a (on the representativeness of the target period), a discussion about the generality of the findings is also useful. Temperature is a good indicator of generality because the microphysical processes acting on precipitation particles, the main target of this study, are primarily determined by temperature and supersaturation, as discussed in many previous studies (Magono and Lee 1966; Kikuchi et al. 2013; Kobayashi 1961; Hueholt et al. 2022; Fukuta and Takahashi 1999; Pruppacher and Klett 1997; Bailey and Hallett 2009; Mason 1971; Nakaya 1954; Hallett and Mason 1958; Pruppacher and Klett 1997). Based on these previous studies, the shift in the growth process and the transition in precipitation from PD and 4K would display some differences from those during the target period.
In a warm year, temperature exceeds the climatological mean. In such cases, the ratio of liquid precipitation and mass fraction by riming in the present day is expected to be larger than during winter 2020/21. Under such conditions, the ratio of liquid precipitation would increase more compared with 2020/21. During a much warmer year, almost all precipitation would occur as a liquid in both PD and 4K. In contrast, during a cold year, temperatures near the surface and the cloud layer are colder than the climatological mean. In such cases, the ratio of liquid precipitation and mass fraction by riming in the present day is expected to be smaller than that of 2020/21. Under this condition, the extent of the shift from solid precipitation to liquid precipitation would be limited. In addition, future changes would be dependent upon temperature within the cloudy layer in the present day.
If the temperature in most of the cloudy layer is sufficiently colder than the temperature ranges from −5° to −15°C at the present day, the increase of the mass fraction of riming would be limited, and the main changes in cloud microphysics would be only the shift of the mass fraction of the depositional growth at each temperature. In contrast, if the temperature in most of the cloudy layer is colder than usual, but part of the cloud layer is located in the temperature range from −5° to −15°C, the shift from depositional growth to riming would occur but to a lesser extent than during 2020/21.
Even if differences are expected due to the selection of the target period, the findings from this study would be applicable for the other years according to the temperature difference between the PD and 4K (Fig. 2d) and the temperature anomaly in 2020/21 from the climatological value (Fig. S1). Figure 2d indicates that the mean temperature difference between PD and 4K over Hokkaido at 850-hPa level, which is around the cloudy layer, is about 5°C. In contrast, the mean temperature anomaly at 850-hPa level for most of the years (based on the 63-yr average from 1960 to 2022) is smaller than 2°C (see Fig. S1). These results indicate that the temperature difference among years is small compared to that between PD and 4K. Thus, the main finding of this study would be applicable regardless of the target period.
In addition to the temperature itself, discussing the dependency of surface precipitation on weather patterns or synoptic circulation, which is not considered by PGWE, is necessary (as we will discuss in section 4c). The differences in weather patterns during each year partly affect the geographical distribution of surface precipitation, but most of the precipitation over the area where a large amount of snowfall occurs is mainly determined by the orography. Based on Fig. 3b, the areas where the large amounts of precipitation occur are located in western Hokkaido and the eastern side of Hidaka Mountains and Taisetsu Mountains, and these precipitation characteristics are similar to the 80-yr historical simulation by Kawase et al. (2021). The former is the windward side of the northwesterly winter monsoon. The latter is the windward side of the east or southeast wind generated during the passing of extratropical cyclones over eastern Hokkaido or off its east coast. These results indicate that orography is one of the main factors determining which areas receive large amounts of precipitation. The precipitation area is expected to remain unchanged even if the weather pattern changes to some extent due to the variation of the synoptic circulation. Thus, the present findings would be applicable to other years.
c. Limitation of the method used in this study
As we described above, a limitation of the experimental setup used in this study is associated with performing simulations for only a single winter. According to previous studies (e.g., Adachi and Tomita 2020), the effects of internal variability are intrinsically included in the output of the PGWE, and simulations must be run for approximately 30 years to reduce the effects of internal variability. However, even if state-of-the-art supercomputers are used, it is unrealistic to conduct 30 years of simulations using PTM, whose computational cost is higher than that of the usual bulk microphysical models. Thus, in a future study, we will attempt to run the PGWE for several decades.
In addition to the representativeness of the target period, we must consider the limitations of the PGWE. This study focused solely on the effects of temperature and vapor variations on future changes to surface solid precipitation. However, alterations in the circulation patterns around Hokkaido could affect the snowfall in the region (Kawazoe et al. 2020; Inatsu et al. 2021). Inatsu et al. (2021) predicted an increase in the westerly dominated winter monsoon and a decrease in the passage of extratropical cyclones over Hokkaido due to a negative phase of the western Pacific pattern. These shifts may result in increased snowfall over the western side of the mountain and decreased snowfall in most of area 3, as shown in Fig. 1b. Such effects of the snowfall originated from the variation in the circulation pattern that was not reproduced in this study; the snowfall amount was decreased all over Hokkaido in the results of this study (Figs. 4 and 5). Despite not considering the effects of variations in the circulation pattern, the main conclusions of this study are robust because the variations mainly stem from changes in temperature and water vapor. To enhance our understanding of future changes in solid precipitation particles, simulations combining PTM with the global-scale convective permitting simulation (Satoh et al. 2019; Stevens et al. 2019) are powerful tools and could provide valuable insights.
Finally, it is crucial to acknowledge the remaining concerns regarding the validity of PTM. Evaluating the performance of PTM in terms of the breakdown of mass fraction generated by each process is challenging due to limited observational data on the habit of solid precipitation. The habit of particles gradually changes in the atmosphere through multiple cloud microphysical processes, complicating the separation of mass generated by each process. Data from videosonde observations (Suzuki et al. 2018, 2023; Mizuno et al. 1994; Takahashi et al. 2019, 1995; Murakami and Matsuo 1990; Orikasa and Murakami 1997) and aircraft measurements (Heymsfield et al. 2006; Dearden et al. 2012) would enable us to evaluate the performance. Hashimoto et al. (2020) confirmed the validity of PTM through comparisons between the results of PTM and those of videosonde data (Orikasa and Murakami 1997) for the snow clouds over Hokuriku. Thus, we believe that PTM reasonably replicated solid precipitation for the target period. However, to enhance our understanding of future changes in solid precipitation characteristics across Hokkaido, future initiatives should include extensive data collection through aircraft measurements, videosonde observations, and surface measurements focusing on the habit of solid precipitation covering a wide area of Hokkaido. In addition to in situ measurements, employing dual-polarization radar would enable us to estimate the distribution of solid precipitation particle habits. Such observations can also serve as valuable tools for evaluating PTM in the future.
5. Conclusions
In this study, we investigated future changes in the contribution of microphysical processes, such as depositional growth and the riming process, to surface precipitation over Hokkaido. This examination was conducted through a PGWE based on d4PDF (Mizuta et al. 2017) data using PTM (Hashimoto et al. 2020) coupled with a meteorological model, SCALE. These microphysical processes are critical in determining the habits and characteristics of solid precipitation particles. We also analyzed the anticipated change in the fraction of liquid and solid precipitation to total precipitation. Our simulations indicated that, in the future, total precipitation amounts are projected to increase over Hokkaido due to the increase of the water vapor amount in the atmosphere. Additionally, the results suggested that the ratio of solid precipitation will decrease and the liquid precipitation will increase in the future due to warmer future environments, which is consistent with the findings in earlier studies (Matsumura and Sato 2011; Kawase et al. 2021, 2016). The simulations further highlighted an increase in surface precipitation generated by the riming process (Riming) and depositional growth under warmer temperatures from −4° to −10°C (Needle/Columnar). Contrastingly, depositional growth under temperatures ranging from −36° to −10°C (Columnar and Plate), which is dominant in PD for surface solid precipitation, will decrease in the future. These changes stem from the warmer environment of cloud layers in the future, with the cloud layer in PD mostly distributed in temperatures ranging between −36° and −10°C. Contrastingly, in the future, the cloud layer will be located at altitudes with temperature ranges suitable for riming (i.e., from −15° to −5°C) (von Blohn et al. 2009; Magono and Lee 1966).
These anticipated future changes will alter the characteristics of snowfall particles in Hokkaido. Earlier studies reported that the number of days with precipitation by graupel, heavily rimed particles, in Hokkaido is smaller than that in other regions of Japan (Mizuno 1992). Our results also showed that the mass fraction of the riming process to total solid precipitation was small in PD, aligning with those of the earlier findings. However, according to the PGWE results, snowfall constructed by rimed particles and graupel is expected to increase and be dominant in the future. Furthermore, liquid precipitation is projected to increase. This precipitation resembles that in the current Hokuriku (Hashimoto et al. 2020; Mizuno 1992; Steenburgh and Nakai 2020), where snowfall particles are heavier than those in Hokkaido. Consequently, our findings suggest that snowfall particles in Hokkaido will be heavier in the future. This heavy snow would lead to an increased snow-removal workload. Considering these implications, the change in snowfall particles and the ratio of liquid precipitation to total precipitation in the future will significantly impact societies in cold regions, including Hokkaido. Therefore, when contemplating adaptation and mitigation strategies for climate changes in cold regions, emphasis should be placed on understanding the effects of these future changes in precipitation characteristics.
The discussion outlined above is also applicable for other regions in the world. Snow clouds in Hokkaido and Hokuriku are triggered by the outbreak of cold air over the relatively warm sea surface temperatures, which is well known as the lake effect (Steiger et al. 2013; Peace and Sykes 1966; Eichenlaub 1970; Steenburgh and Nakai 2020) and frequently observed over lakes and seas in cold regions, such as the Great Lakes, Lake Baikal, Black Sea, Caspian Sea, and North Sea. Knowledge about future changes in solid precipitation particles and differences in such changes between Hokuriku and Hokkaido will allow us to estimate future changes in lake-effect snow in other areas.
In winter over Hokkaido, which belongs to the subarctic zone based on the Köppen climate classification (Köppen 1936), PD indicated that the temperature is low and solid precipitation is mainly created by depositional growth. In the future, the riming process would become dominant for solid precipitation particles. This is similar to the conditions observed for solid precipitation over Hokuriku, which belongs to the temperate zone (Köppen 1936). According to analyses by Cui et al. (2021), the west coast of Hokkaido would shift to the temperate zone in some future scenarios, which is consistent with the results of this study. In contrast, for winter snow over Hokuriku, PD indicated that the temperature of cloud layers is warmer than that in subarctic regions. Under such conditions, the riming process is already dominant for solid precipitation in PD; moreover, solid precipitation will decrease in the future because the surface temperature will exceed 0°C (Kawase et al. 2016). Therefore, future changes in lake effect snow will likely differ between the temperate zone and subarctic zone. For lake effects snow clouds in the subarctic area, the main microphysical process underlying surface solid precipitation is depositional growth in PD; however, this process would change to riming in the future. In contrast, for lake effect snow clouds in the temperate zone, the main microphysical process underlying surface solid precipitation is riming in PD; however, solid precipitation would change to liquid water precipitation in the future.
Acknowledgments.
This work was supported by the JSPS Kakenhi (Grant 21H04571), Research Field of Hokkaido Weather Forecast and Technology Development (endowed by Hokkaido Weather Technology Center Co., Ltd.), Cooperative Research Activities of Collaborative Use of Computing Facility (JURCAOSCFS23-02) of the Atmosphere and Ocean Research Institute, the University of Tokyo and L-Station, and Creative Research Institution of Hokkaido University based on the “Support System for the Collaborative Research of Next-Generation Researchers (FY2022).” Part of this work was conducted as a joint research program of CEReS, Chiba University (FY2022 and 2023). M. I. is supported by the Environment Research and Technology Development Fund JPMEERF20232003 of the Environmental Restoration and Conservation Agency provided by the Ministry of the Environment of Japan and by JPMXD0722680734 of the Ministry of Education, Sports, Culture, Science and Technology.
Data availability statement.
The MANL data are available via the Meteorological Research Consortium between the JMA and the Meteorological Society of Japan. The data of the AMeDAS are downloadable from the JMA website. The source code and documentation of SCALE except for PTM are available online at https://scale.riken.jp/. Simulated data and the source code of PTM can be accessed by request to the corresponding author.
REFERENCES
Abe, O., 2004: Shear strength and angle of repose of snow layers including graupel. Ann. Glaciol., 38, 305–308, https://doi.org/10.3189/172756404781815149.
Adachi, S. A., and H. Tomita, 2020: Methodology of the constraint condition in dynamical downscaling for regional climate evaluation: A review. J. Geophys. Res. Atmos., 125, e2019JD032166, https://doi.org/10.1029/2019JD032166.
Bailey, M. P., and J. Hallett, 2009: A comprehensive habit diagram for atmospheric ice crystals: Confirmation from the laboratory, AIRS II, and other field studies. J. Atmos. Sci., 66, 2888–2899, https://doi.org/10.1175/2009JAS2883.1.
Beljaars, A. C. M., and A. A. M. Holtslag, 1991: Flux parameterization over land surfaces for atmospheric models. J. Appl. Meteor., 30, 327–341, https://doi.org/10.1175/1520-0450(1991)030<0327:FPOLSF>2.0.CO;2.
Chen, J.-P., and D. Lamb, 1994a: The theoretical basis for the parameterization of ice crystal habits: Growth by vapor deposition. J. Atmos. Sci., 51, 1206–1222, https://doi.org/10.1175/1520-0469(1994)051<1206:TTBFTP>2.0.CO;2.
Chen, J.-P., and D. Lamb, 1994b: Simulation of cloud microphysical and chemical processes using a multicomponent framework. Part I: Description of the microphysical model. J. Atmos. Sci., 51, 2613–2630, https://doi.org/10.1175/1520-0469(1994)051<2613:SOCMAC>2.0.CO;2.
Colle, B. A., D. Stark, and S. E. Yuter, 2014: Surface microphysical observations within East Coast winter storms on long island, New York. Mon. Wea. Rev., 142, 3126–3146, https://doi.org/10.1175/MWR-D-14-00035.1.
Cui, D., S. Liang, D. Wang, and Z. Liu, 2021: A 1 km global dataset of historical (1979–2013) and future (2020–2100) Köppen–Geiger climate classification and bioclimatic variables. Earth Syst. Sci. Data, 13, 5087–5114, https://doi.org/10.5194/essd-13-5087-2021.
Dearden, C., P. J. Connolly, T. Choularton, P. R. Field, and J. Heymsfield, 2012: Factors influencing ice formation and growth in simulations of a mixed-phase wave cloud. J. Adv. Model. Earth Syst., 4, M10001, https://doi.org/10.1029/2012MS000163.
Eichenlaub, V. L., 1970: Lake effect snowfall to the lee of the Great Lakes: Its role in Michigan. Bull. Amer. Meteor. Soc., 51, 403–413, https://doi.org/10.1175/1520-0477(1970)051<0403:LESTTL>2.0.CO;2.
Fukuta, N., and T. Takahashi, 1999: The growth of atmospheric ice crystals: A summary of findings in vertical supercooled cloud tunnel studies. J. Atmos. Sci., 56, 1963–1979, https://doi.org/10.1175/1520-0469(1999)056<1963:TGOAIC>2.0.CO;2.
Hallett, J., and B. J. Mason, 1958: The influence of temperature and supersaturation on the habit of ice crystals grown from the vapour. Proc. Roy Soc. London, 247A, 440–453, https://doi.org/10.1098/rspa.1958.0199.
Harimaya, T., and N. Kanemura, 1995: Comparison of the riming growth of snow particles between coastal and inland areas. J. Meteor. Soc. Japan, 73, 25–36, https://doi.org/10.2151/jmsj1965.73.1_25.
Hashimoto, A., H. Motoyoshi, N. Orikasa, and R. Misumi, 2020: Process-tracking scheme based on bulk microphysics to diagnose the features of snow particles. SOLA, 16, 51–56, https://doi.org/10.2151/sola.2020-009.
Hashino, T., and G. J. Tripoli, 2007: The Spectral Ice Habit Prediction System (SHIPS). Part I: Model description and simulation of the vapor deposition process. J. Atmos. Sci., 64, 2210–2237, https://doi.org/10.1175/JAS3963.1.
Heymsfield, A. J., A. Bansemer, S. L. Durden, R. L. Herman, and T. P. Bui, 2006: Ice microphysics observations in Hurricane Humberto: Comparison with non-hurricane-generated ice cloud layers. J. Atmos. Sci., 63, 288–308, https://doi.org/10.1175/JAS3603.1.
Hueholt, D. M., S. E. Yuter, and M. A. Miller, 2022: Revisiting diagrams of ice growth environments. Bull. Amer. Meteor. Soc., 103, E2584–E2603, https://doi.org/10.1175/BAMS-D-21-0271.1.
Inatsu, M., S. Tanji, and Y. Sato, 2020: Toward predicting expressway closures due to blowing snow events. Cold Reg. Sci. Technol., 177, 103123, https://doi.org/10.1016/j.coldregions.2020.103123.
Inatsu, M., S. Kawazoe, and M. Mori, 2021: Trends and projection of heavy snowfall in Hokkaido, Japan as an application of self-organizing map. J. Appl. Meteor. Climatol., 60, 1483–1494, https://doi.org/10.1175/JAMC-D-21-0085.1.
Ishii, M., and N. Mori, 2020: d4PDF: Large-ensemble and high-resolution climate simulations for global warming risk assessment. Prog. Earth Planet. Sci., 7, 58, https://doi.org/10.1186/s40645-020-00367-7.
Japan Meteorological Agency, 2023a: The observation status of the first day and the entire day for snowfall, snow accumulation, and long-term snow cover (in Japanese). Accessed 21 November 2023, https://www.data.jma.go.jp/sapporo/tenki/snow-frost-ice/snow_previous.html.
Japan Meteorological Agency, 2023b: Prospect of the winter climate in Hokkaido (in Japanese). Accessed 19 October 2023, https://www.data.jma.go.jp/cpd/longfcst/kaisetsu/?region=010100&term=P6M.
Katsuyama, Y., M. Inatsu, and T. Shirakawa, 2020: Response of snowpack to +2°C global warming in Hokkaido, Japan. J. Glaciol., 66, 83–96, https://doi.org/10.1017/jog.2019.85.
Kawase, H., A. Murata, R. Mizuta, H. Sasaki, M. Nosaka, M. Ishii, and I. Takayabu, 2016: Enhancement of heavy daily snowfall in central Japan due to global warming as projected by large ensemble of regional climate simulations. Climatic Change, 139, 265–278, https://doi.org/10.1007/s10584-016-1781-3.
Kawase, H., and Coauthors, 2021: Regional characteristics of future changes in snowfall in Japan under RCP2.6 and RCP8.5 scenarios. SOLA, 17, 1–7, https://doi.org/10.2151/sola.2021-001.
Kawazoe, S., M. Inatsu, T. J. Yamada, and T. Hoshino, 2020: Climate change impacts on heavy snowfall in Sapporo using 5-km mesh large ensemble simulations. SOLA, 16, 233–239, https://doi.org/10.2151/sola.2020-039.
Kikuchi, K., T. Kameda, K. Higuchi, and A. Yamashita, 2013: A global classification of snow crystals, ice crystals, and solid precipitation based on observations from middle latitudes to polar regions. Atmos. Res., 132–133, 460–472, https://doi.org/10.1016/j.atmosres.2013.06.006.
Kobayashi, T., 1961: The growth of snow crystals at low supersaturations. Philos. Mag., 6, 1363–1370, https://doi.org/10.1080/14786436108241231.
Kondo, M., Y. Sato, M. Inatsu, and Y. Katsuyama, 2021: Evaluation of cloud microphysical schemes for winter snowfall events in Hokkaido: A case study of snowfall by winter monsoon. SOLA, 17, 74–80, https://doi.org/10.2151/sola.2021-012.
Köppen, W., 1936: Das Geographische System Der Klimate. Borntraeger, 44 pp.
Koren, B., 1993: A Robust Upwind Discretization Method for Advection, Diffusion and Source Terms. Stichting Mathematisch Centrum, 22 pp.
Kusaka, H., H. Kondo, Y. Kikegawa, and F. Kimura, 2001: A simple single-layer urban canopy model for atmospheric models: Comparison with multi-layer and slab models. Bound.-Layer Meteor., 101, 329–358, https://doi.org/10.1023/A:1019207923078.
Liu, C., and Coauthors, 2017: 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.
Magono, C., and C. W. Lee, 1966: Meteorological classification of natural snow crystals. J. Fac. Sci., Hokkaido Univ., Ser. 7, 2, 321–335.
Magono, C., K. Kikuchi, T. Kimura, S. Tazawa, and T. Kasal, 1966: A Study on the snowfall in the winter monsoon season in Hokkaido with special reference to low land snowfall: (Investigation of natural snow crystals 6). J. Fac. Sci., Hokkaido Univ., Ser. 7, 2, 287–308.
Mason, B. J., 1971: The Physics of Clouds. 2nd ed. Oxford University Press, 671 pp.
Matsumura, S., and T. Sato, 2011: Snow/ice and cloud responses to future climate change around Hokkaido. SOLA, 7, 205–208, https://doi.org/10.2151/sola.2011-052.
Mizuno, H., 1992: Statistical characteristics of Graupel precipitation over the Japan Islands. J. Meteor. Soc. Japan, 70, 115–121, https://doi.org/10.2151/jmsj1965.70.1_115.
Mizuno, H., T. Matsuo, M. Murakami, and Y. Yamada, 1994: Microstructure of cirrus clouds observed by HYVIS. Atmos. Res., 32, 115–124, https://doi.org/10.1016/0169-8095(94)90055-8.
Mizuta, R., and Coauthors, 2012: Climate simulations using MRI-AGCM3.2 with 20-km grid. J. Meteor. Soc. Japan, 90A, 233–258, https://doi.org/10.2151/jmsj.2012-A12.
Mizuta, R., and Coauthors, 2017: Over 5,000 years of ensemble future climate simulations by 60-km global and 20-km regional atmospheric models. Bull. Amer. Meteor. Soc., 98, 1383–1398, https://doi.org/10.1175/BAMS-D-16-0099.1.
Murakami, M., and T. Matsuo, 1990: Development of the hydrometeor videosonde. J. Atmos. Oceanic Technol., 7, 613–620, https://doi.org/10.1175/1520-0426(1990)007<0613:DOTHV>2.0.CO;2.
Nakamura, K., 2019: Implementation and demonstration of a system for the forecasting of surface avalanche potential caused by snowfall from a cyclone. J. Disaster Res., 14, 1201–1226, https://doi.org/10.20965/jdr.2019.p1201.
Nakanishi, M., and H. Niino, 2006: An improved Mellor–Yamada level-3 model: Its numerical stability and application to a regional prediction of advection fog. Bound.-Layer Meteor., 119, 397–407, https://doi.org/10.1007/s10546-005-9030-8.
Nakaya, U., 1954: Snow Crystals, Natural and Artificial. Harvard University Press, 510 pp.
Nishizawa, S., H. Yashiro, Y. Sato, Y. Miyamoto, and H. Tomita, 2015: Influence of grid aspect ratio on planetary boundary layer turbulence in large-eddy simulations. Geosci. Model Dev., 8, 3393–3419, https://doi.org/10.5194/gmd-8-3393-2015.
Orikasa, N., and M. Murakami, 1997: A new version of hydrometeor videosonde for cirrus cloud observations. J. Meteor. Soc. Japan, 75, 1033–1039, https://doi.org/10.2151/jmsj1965.75.6_1033.
Peace, R. L., Jr., and R. B. Sykes Jr., 1966: Mesoscale study of a lake effect snow storm. Mon. Wea. Rev., 94, 495–507, https://doi.org/10.1175/1520-0493(1966)094<0495:MSOALE>2.3.CO;2.
Pruppacher, H. R., and J. D. Klett, 1997: Microphysics of Clouds and Precipitation. 2nd ed. Kluwer Academic Publisher, 955 pp.
Saito, K., and Coauthors, 2006: The operational JMA nonhydrostatic mesoscale model. Mon. Wea. Rev., 134, 1266–1298, https://doi.org/10.1175/MWR3120.1.
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.
Sato, Y., S. Nishizawa, H. Yashiro, Y. Miyamoto, Y. Kajikawa, and H. Tomita, 2015: Impacts of cloud microphysics on trade wind cumulus: Which cloud microphysics processes contribute to the diversity in a large eddy simulation? Prog. Earth Planet. Sci., 2, 23, https://doi.org/10.1186/s40645-015-0053-6.
Satoh, M., B. Stevens, F. Judt, M. Khairoutdinov, S.-J. Lin, W. M. Putman, and P. Düben, 2019: Global cloud-resolving models. Curr. Climate Change Rep., 5, 172–184, https://doi.org/10.1007/s40641-019-00131-0.
Schaerer, P., and D. McClung, 2006: The Avalanche Handbook. 3rd ed. Mountaineers Books, 342 pp.
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.
Seiki, T., and T. Nakajima, 2014: Aerosol effects of the condensation process on a convective cloud simulation. J. Atmos. Sci., 71, 833–853, https://doi.org/10.1175/JAS-D-12-0195.1.
Sekiguchi, M., and T. Nakajima, 2008: A k-distribution-based radiation code and its computational optimization for an atmospheric general circulation model. J. Quant. Spectrosc. Radiat. Transfer, 109, 2779–2793, https://doi.org/10.1016/j.jqsrt.2008.07.013.
Shima, S., K. Kusano, A. Kawano, T. Sugiyama, and S. Kawahara, 2009: The super-droplet method for the numerical simulation of clouds and precipitation: A particle-based and probabilistic microphysics model coupled with a non-hydrostatic model. Quart. J. Roy. Meteor. Soc., 135, 1307–1320, https://doi.org/10.1002/qj.441.
Shima, S.-I., Y. Sato, A. Hashimoto, and R. Misumi, 2020: Predicting the morphology of ice particles in deep convection using the super-droplet method: Development and evaluation of SCALE-SDM 0.2.5-2.2.0, -2.2.1, and -2.2.2. Geosci. Model Dev., 13, 4107–4157, https://doi.org/10.5194/gmd-13-4107-2020.
Steenburgh, W. J., and S. Nakai, 2020: Perspectives on sea- and lake-effect precipitation from Japan’s “Gosetsu Chitai”. Bull. Amer. Meteor. Soc., 101, E58–E72, https://doi.org/10.1175/BAMS-D-18-0335.1.
Steiger, S. M., and Coauthors, 2013: Circulations, bounded weak echo regions, and horizontal vortices observed within long-lake-axis-parallel–lake-effect storms by the Doppler on wheels. Mon. Wea. Rev., 141, 2821–2840, https://doi.org/10.1175/MWR-D-12-00226.1.
Stevens, B., and Coauthors, 2019: DYAMOND: The dynamics of the atmospheric general circulation modeled on non-hydrostatic domains. Prog. Earth Planet. Sci., 6, 61, https://doi.org/10.1186/s40645-019-0304-z.
Suzuki, K., K. Nakagawa, T. Kawano, S. Mori, M. Katsumata, F. Syamsudin, and K. Yoneyama, 2018: Videosonde-observed graupel in different rain systems during Pre-YMC project. SOLA, 14, 148–152, https://doi.org/10.2151/sola.2018-026.
Suzuki, K., Y. Hara, T. Sugidachi, K. Shimizu, and M. Fujiwara, 2023: Development of a new particle imaging radiosonde with particle fall velocity measurements in clouds. SOLA, 19, 261–268, https://doi.org/10.2151/sola.2023-034.
Takahashi, T., Y. Nagao, and Y. Kushiyama, 1995: Possible high ice particle production during Graupel–Graupel collisions. J. Atmos. Sci., 52, 4523–4527, https://doi.org/10.1175/1520-0469(1995)052<4523:PHIPPD>2.0.CO;2.
Takahashi, T., S. Sugimoto, T. Kawano, and K. Suzuki, 2019: Microphysical structure and lightning initiation in Hokuriku winter clouds. J. Geophys. Res. Atmos., 124, 13 156–13 181, https://doi.org/10.1029/2018JD030227.
von Blohn, N., K. Diehl, S. K. Mitra, and S. Borrmann, 2009: Riming of Graupel: Wind tunnel investigations of collection kernels and growth regimes. J. Atmos. Sci., 66, 2359–2366, https://doi.org/10.1175/2009JAS2969.1.
Wilson, D. K., 2001: An alternative function for the wind and temperature gradients in unstable surface layers. Bound.-Layer Meteor., 99, 151–158, https://doi.org/10.1023/A:1018718707419.
Wu, W., and A. H. Lynch, 2000: Response of the seasonal carbon cycle in high latitudes to climate anomalies. J. Geophys. Res., 105, 22 897–22 908, https://doi.org/10.1029/2000JD900340.