1. Introduction
Precipitation is an essential climate system component closely associated with human activity. Seasonal precipitation variations over the Asian monsoon region have the unique feature that most intense storms occur in the premonsoon season, while numerous precipitation systems bring abundant precipitation during monsoon but few severe storms (Lau et al. 1988; Matsumoto 1997; Zipser et al. 2006). Therefore, variations in the amount of precipitation and changes in its characteristics can be of great concern in intensifying floods and droughts under climate change (Seneviratne et al. 2021).
Precipitation studies over the Asian monsoon regions have been performed from various viewpoints, especially regarding seasonal variations, including the transition from the premonsoon to the monsoon and postmonsoon seasons. However, previous studies have shown that the precipitation amount and its characteristics differ: the size of the precipitation systems (e.g., Romatschke and Houze 2011), precipitation intensity (e.g., Islam and Uyeda 2008; Takahashi 2016; Zipser et al. 2006), precipitation type such as convective or stratiform (e.g., Romatschke et al. 2010), precipitation/echo-top height (e.g., Kodama et al. 2005), and other environmental factors (e.g., Yamane and Hayashi 2006; Ono and Takahashi 2016).
For example, Zipser et al. (2006) investigated the global distribution of the most intense thunderstorms in a study addressing observed precipitation characteristics. They suggested intense storms were dominant over the Ganges Plain and Bangladesh in the premonsoon season, whereas the mature monsoon season had few intense storms. Kodama et al. (2005) found that premonsoon precipitation over Bangladesh was characterized by less rainfall and deep convection with frequent lightning. Romatschke et al. (2010) investigated the features of deep convective cores, wide convective cores, and broad stratiform regions. They found that deep convective cores changed markedly from India’s east coast in the premonsoon season to the western Himalayan foothills in the monsoon season. The Tropical Rainfall Measuring Mission (TRMM; Kummerow et al. 1998; Kozu et al. 2001) contributed to these observational findings related to seasonal differences in precipitation characteristics at the global scale, enabling the study of the global characterization of precipitation systems (Nakamura 2021).
In addition to studies showing precipitation characteristics from observational facts, some studies have discussed how seasonal differences in atmospheric environmental fields cause seasonal changes in precipitation characteristics. For example, Yamane and Hayashi (2006) suggested that intense thunderstorms during the premonsoon season may be related to high thermal instability and vertical wind shear over Bangladesh and northeastern India. In terms of the changes in the land surface conditions becoming wet from the premonsoon to mature monsoon seasons, observations (Ono and Takahashi 2016; Takahashi 2016) and numerical modeling (Sugimoto and Takahashi 2017) have shown that dry land surface conditions during the premonsoon season can enhance the convection associated with strong surface heating.
Many studies have addressed seasonal changes in precipitation characteristics over various domains of the Asian monsoon region and found that precipitation in the premonsoon season is characterized by less frequent but more intense deep convection with frequent lightning. This characteristic precipitation can be enhanced under high thermal instability and/or dry land surface conditions.
Such seasonal differences in precipitation characteristics relate to the “structure” of the precipitation systems mentioned above. In addition, some studies have addressed seasonal changes in the “microphysical properties” of cloud–precipitation systems over the Asian monsoon region. For example, Prabha et al. (2011) investigated seasonal differences in the microphysical properties of deep convective clouds using in situ measurements, focusing on different thermodynamic and aerosol environments. Kozu et al. (2006) investigated seasonal and diurnal variations in the drop size distribution over India. However, reliable observational data on the microphysical properties, such as geophysical variables related to particles, are lacking. Although in situ observations have been used in some case studies of the microphysical structure of clouds or precipitation, the available data have been spatially and temporally limited. Therefore, the difference between the premonsoon and monsoon seasons is still not fully understood, especially from a climatological and statistical perspective. Additional spatiotemporal observational data are required to strengthen our understanding of seasonal changes in precipitation characteristics from the perspective of precipitation microphysical properties.
Recently, satellite observations have become capable of capturing the precipitation microphysical quantities. The Global Precipitation Measurement (GPM; Hou et al. 2014; Skofronick-Jackson et al. 2017) mission, as the successor to the TRMM, carries a Dual-Frequency Precipitation Radar (DPR; Kojima et al. 2012). The GPM/DPR consists of a Ku-band Precipitation Radar (KuPR), suitable for observing heavy rainfall, and a Ka-band Precipitation Radar (KaPR), capable of capturing light rainfall. GPM/KuPR is similar to the Precipitation Radar (PR) aboard the TRMM satellite. KaPR has been newly added to increase the sensitivity to obtain snowfall over the midlatitudes and more accurately estimate microphysical properties such as particle size. By utilizing the observed information at two different frequencies, the GPM/DPR can provide new types of valuable products not available in the TRMM/PR product estimated by the Ku-band single frequency.
Owing to the availability of new space-based observational information, the mass-weighted mean diameter (Dm), which is a typical drop size distribution (DSD) parameter, can be estimated using the GPM/DPR (Seto et al. 2021). Recent studies have shown large-scale distributions of DSD parameters by analyzing spaceborne DSD products (Radhakrishna et al. 2020; Yamaji et al. 2020; Ryu et al. 2021; Han and Braun 2021). Radhakrishna et al. (2020) characterized the variations in Dm from the GPM/DPR product over different climatic regions of the Indian subcontinent and adjoining seas during the Asian monsoon season (June–September). They found that the Dm values were large for deep systems and small for shallow systems over the continent, indicating that deep continental storms had larger raindrops.
In addition, new observations of the two different frequencies allow for the detection of the existence of heavy ice precipitation, such as graupel or hail (Iguchi et al. 2018; Akiyama et al. 2019; Le and Chandrasekar 2021), which are known to be closely related to lightning activities in convective clouds (Takahashi 2006). As mentioned previously, lightning is known to occur more frequently in the premonsoon season than in the monsoon season over the Asian monsoon region; therefore, new information from GPM/DPR can help examine the characteristics of the existence of heavy ice precipitation inside convective clouds.
Therefore, this study aims to determine the climatological characteristics of seasonal variations in precipitation microphysical properties over the Asian monsoon region using a newly available accumulated satellite dataset that has not been addressed previously. We focused on the DSD parameter and the existence of ice particles as microphysical properties of precipitation, particularly the differences between the premonsoon and monsoon seasons. Furthermore, the association among precipitation characteristics, precipitation microphysical properties, and atmospheric environmental factors were investigated using 8 years of GPM/DPR product.
Section 2 explains the datasets and the methodology used in this study. The results showing seasonal variations in precipitation characteristics over the Asian monsoon region are presented in section 3, including the seasonal march focusing on the Bangladesh Plain. Section 4 discusses the association between the precipitation characteristics and atmospheric environmental factors that differ between the premonsoon and monsoon seasons. Finally, the conclusions are presented in section 5.
2. Data and methodology
a. Satellite observations
This study used 8 years of data (2014–21) from the GPM/DPR level-2 version 06 product (Seto et al. 2021). The GPM/DPR algorithm can be divided into KuPR-only, KaPR-only, and dual-frequency (both KuPR and KaPR) algorithms. This study used the variables of hourly precipitation rate, Dm, and flag of heavy ice precipitation from the DPR level-2 (DPRL2) product with the dual-frequency algorithm. The observation swath of GPM/KuPR is approximately 245 km, corresponding to ±17° electrical beam scanning, which is the same as that of TRMM/PR. In contrast, the GPM/KaPR swath width is approximately 125 km, corresponding to ±8.5° of electrical beam scanning (Kojima et al. 2012). Therefore, the DPRL2 product with the dual-frequency algorithm has a narrower swath of 125 km, leading to fewer samples.
The monthly mean values of the precipitation amount, frequency, intensity, precipitation-top heights, Dm, and frequency of heavy ice precipitation were examined. The values were calculated from the 8-yr monthly mean 1° × 1° averaged box (latitude and longitude, respectively). The spatial resolution was relatively low because the sampling of the GPM/DPR dual-frequency observations was less than that of TRMM/PR, owing to the differences in observation swath and inclination. The target region for this study was set at 10°S–35°N, 60°–150°E.
We calculated the amount, intensity, and frequency of precipitation from the GPM/DPR hourly precipitation rate at the near-surface level, corresponding to the clutter-free bottom. The precipitation amount was defined as the total precipitation amount divided by the total number of observations. In contrast, the precipitation intensity was the same as the precipitation amount but divided by the total number of precipitating pixels at the near-surface level. Precipitation frequency was determined as the ratio of the total number of precipitating pixels at the near-surface level to the total number of observations, including nonprecipitation samples.
Precipitation-top heights were determined as the height from the surface level to the highest altitude with a precipitation rate above 0.3 mm h−1. This threshold was set by considering the sensitivity of the DPR instrument (Kojima et al. 2012; Masaki et al. 2022). The frequency of heavy ice precipitation was defined as the ratio of the flag of heavy ice precipitation to the total number of precipitating pixels at the near-surface level. In addition to using the flag information having no vertical resolution, we used the detection thresholds proposed by Iguchi et al. (2018) to identify heights above −10°C where the heavy ice particles exist. Three criteria were proposed by Iguchi et al. (2018) and this study considered a detection to have occurred when one of the conditions was met: 1) the measured dual-frequency ratio (DFRm) was larger than 7 dB when the measured apparent radar reflectivity factor (Zm) of KuPR was larger than 27 dBZ; 2) Zm of KuPR was larger than 35 dBZ; and 3) Zm of KaPR was larger than 30 dBZ.
To investigate the differences in the large-scale geographical distributions of these variables between the premonsoon and monsoon seasons, the data were averaged over a 1° × 1° box (latitude and longitude) in two steps. First, we averaged the values for April–May (AM) and July–August (JA) as the premonsoon and monsoon seasons, respectively, from an orbit-based level-2 product (2014AM, 2014JA, etc.). Note that the definition of the premonsoon and monsoon seasons depends on the region and differs among the studies. In this study, we used a definition that captures the characteristics of the entire Asian monsoon region in a unified manner. In the second step, a 2-month dataset was used as one sample and the 8-yr mean was calculated. If the number of nonmissing 2-month data was more than six (the maximum value is 8 as the total duration of 8 years), the grid box was plotted. To determine the difference between the premonsoon and monsoon seasons, Welch’s t test (Welch 1938) was conducted to detect statistically significant differences with a 95% confidence level by taking a 2-month mean for a year as one sample.
In the discussion section, the differences in the relationships among the precipitation variables were investigated on a DPRL2 pixel basis to check whether a change in R, a change in the characteristics of the precipitation, or a combination of both caused seasonal changes in Dm. In addition, the occurrence of heavy ice precipitation was examined as a function of altitude at three Dm levels (large, medium, and small). The original GPM/DPR data did not provide the vertical distribution of heavy ice precipitation; therefore, we used the same threshold as that used by Iguchi et al. (2018) to analyze the vertical profile of the existence of heavy ice precipitation.
b. Atmospheric reanalysis dataset
In addition to the GPM/DPR observations, atmospheric reanalysis data were used to discuss the differences in atmospheric environmental conditions. Geopotential height, temperature, and specific humidity from the Japanese 55-year Reanalysis (JRA-55) (Kobayashi et al. 2015; Harada et al. 2016) were used. This dataset enables investigations on the prevailing atmospheric conditions in the presence of precipitation, which can support our discussions on seasonal changes in precipitation microphysical characteristics based on GPM/DPR observations. The resolution of JRA-55 was 1.25° × 1.25° at 6-h intervals. Following a method similar to that of Fujinami and Yasunari (2001), we calculated the vertical gradients of the dry and moist static energies in kJ kg−1 from 950 to 850 hPa to represent the stability of the lower atmosphere. Note that the spatial resolutions of the DPRL2 data and the JRA-55 data were different, but the analysis and discussions in this study were based on statistics from both datasets.
3. Seasonal variations in precipitation characteristics over the Asian monsoon region
a. Differences in climatological precipitation characteristics between premonsoon and monsoon seasons
Before examining the precipitation microphysical characteristics derived from GPM/DPR, conventional precipitation characteristics (precipitation frequency and conditional precipitation intensity) were reconfirmed from 8-yr DPRL2 data. Because the sampling of the dual-frequency GPM/DPR observations was less than that of TRMM/PR, we examined whether the 8-yr DPRL2 data could present climatological patterns of precipitation characteristics (Fig. 1).
Geographical distribution of mean precipitation amount (mm day−1) in the (a) premonsoon, (b) monsoon, and (c) their difference (monsoon minus premonsoon). The grids with black dots and black open squares in Fig. 1c indicate statistically significant differences at a 99% confidence level using Welch’s t test and the domain of the BP region (23°–27°N, 88°–93°E) in the Bangladesh plain, respectively.
Citation: Journal of the Atmospheric Sciences 80, 8; 10.1175/JAS-D-22-0198.1
Precipitation amount over the 10°–30°N latitudinal band over the Asian monsoon region was generally larger in the mature monsoon season than in the premonsoon season (Fig. 1), which is well known as the northward progress of the rainy season from spring to summer. We confirmed clear peak values over the western coastal regions of the Indian subcontinent, the Indochina Peninsula, and the Philippines, which the summer monsoon westerlies can explain in the low troposphere. In addition to these western coastal regions, precipitation significantly increased from premonsoon to summer monsoon seasons over the Bay of Bengal, the South China Sea, and the Philippine Sea. In contrast, the amount of precipitation during April–May was higher than that during July–August over the Maritime Continent. These seasonal marches in rainfall amount over the Asian monsoon regions mainly corresponded to previous studies (Matsumoto 1997; Wang and LinHo 2002; Takahashi and Yasunari 2006), which implies that the 8-yr DPRL2 data can capture the climatological seasonal march in rainfall amount using a 2-month average and 1° × 1° horizontal resolution.
Figure 2 shows the precipitation frequency, determined as the ratio of the total number of precipitating pixels at the near-surface level to the total number of observations, including nonprecipitating samples, and the conditional precipitation intensity.
As in Fig. 1, but for (a)–(c) precipitation frequency and (d)–(f) precipitation intensity (mm h−1). Gray-shaded grids in (d)–(f) are masked due to sampling issues.
Citation: Journal of the Atmospheric Sciences 80, 8; 10.1175/JAS-D-22-0198.1
We found that the precipitation frequency in Figs. 2a–c exhibited a similar trend to the precipitation amount in Fig. 1, thereby indicating that the number of precipitating pixels was higher in the monsoon season than in the premonsoon season over most of the land regions in the 10°–30°N latitudinal band. However, the opposite results were observed over the Maritime Continent. Furthermore, regions with statistically significant changes in precipitation frequency were broader than those with statistically significant changes in precipitation amounts, as shown in Fig. 1. Over the boreal subtropical oceans in the 10°–25°N latitudinal band, precipitation was more frequent in the monsoon season than in the premonsoon season, whereas precipitation was less frequent in July–August than in April–May over the 25°–35°N latitudinal band region of the western North Pacific.
The precipitation intensity differences between seasons in Figs. 2d–f were similar to the precipitation amount and frequency differences, although the spatial pattern included noise due to sampling issues; the method to exclude them was mentioned in section 2a. It was noted that the precipitation intensity was higher in the premonsoon season than in the monsoon season in some land areas, particularly over Bangladesh, the southern Indian subcontinent, the central areas of the Indochina Peninsula, and southeastern China. These precipitation characteristics (frequency and intensity) over the Asian monsoon region are consistent with the TRMM/PR results (Takahashi 2016). Thus, a moderate spatial scale, as in our analysis, allows for the examination of precipitation characteristics, even with 8 years of DPRL2 data.
Furthermore, the climatological features of the precipitation-top heights, which generally correspond to the precipitation intensity, were checked, as shown in Fig. 3. The values over land were higher over the northeastern Ghats of the Indian subcontinent, Bangladesh Plain, and Indochina Peninsula during the premonsoon season (Fig. 3a). In contrast to the premonsoon season, the precipitation-top heights were higher in the western Himalayan region during the monsoon season (Fig. 3b). The differences between the two seasons are shown in Fig. 3c.
As in Fig. 1, but for precipitation-top heights (km). Gray-shaded grids are masked out due to sampling issues.
Citation: Journal of the Atmospheric Sciences 80, 8; 10.1175/JAS-D-22-0198.1
Focusing on the Bangladesh Plain, Islam and Uyeda (2008) investigated variations in rainfall intensity and echo-top height in Bangladesh. They found that premonsoon rainfall was characterized by convective rain with strong intensity up to high altitudes and high echo tops. In contrast, strong and high echo tops were less significant during the monsoon than during the premonsoon period. These results were consistent with those obtained using the DPRL2 product.
However, precipitation-top heights over the oceans were generally higher over subtropical oceans at latitudes of 10°–20°N during the monsoon season, and the opposite was confirmed over the Maritime Continent. The differences over the Arabian Sea were not as significant as those over the Pacific Ocean and Philippine Sea.
b. Characteristic seasonal changes in precipitation microphysical properties related to the intensity of precipitation
From the results shown in the previous subsection, we confirmed that precipitation over the Indian subcontinent and Indochina Peninsula in the premonsoon season is less in terms of amount and frequency but is more intense and deeply developed compared to precipitation in the mature monsoon season. Therefore, this subsection focuses on the characteristics of intense and tall precipitation systems in the premonsoon season. In addition, the microphysical properties, such as Dm and frequency of heavy ice precipitation, are investigated. Note that the Dm information in this study describes values near the surface.
Values of Dm, a typical drop size distribution parameter that reflects precipitation characteristics, were generally larger over land than over the oceans in both seasons, as shown in Figs. 4a and 4b. However, the land–ocean contrast was more evident in the premonsoon season. In terms of seasonal changes over land, Dm was generally larger in the premonsoon season than in the monsoon season, particularly over the southern and eastern parts of the Indian subcontinent, the Bangladesh Plain, the Indochina Peninsula, and the Philippine Islands, as indicated by the brown shades in Fig. 4c. The Dm signals showed statistically significant changes over the southern part of the Indian subcontinent, the Bangladesh Plain, and the Indochina Peninsula. In contrast, the Dm values over the Tibetan Plateau and the northwestern Himalayan region were smaller during the premonsoon season than during the monsoon season. These regions and seasons are consistent with the intense and deeply developed precipitation systems, as shown in Figs. 2d–f and 3.
As in Fig. 1, but for Dm near the surface in mm. Gray-shaded grids are masked out due to sampling issues.
Citation: Journal of the Atmospheric Sciences 80, 8; 10.1175/JAS-D-22-0198.1
Over subtropical oceans, Dm was generally smaller in the premonsoon season than in the monsoon season, opposite to the Dm changes observed over land. The changes in Dm over the oceans were similar to those in precipitation intensity (Figs. 2d–f) and precipitation-top heights (Fig. 3). Dm exhibited statistically significant changes over the eastern part of the Arabian Sea. In contrast, the changes were not substantial over the Bay of Bengal, which did not show the same tendency as the precipitation intensity. Some previous studies (e.g., Hirose and Nakamura 2005) have suggested that small precipitation systems are dominant over the Arabian Sea, while widespread precipitation systems prevail over the Bay of Bengal, even though these regions off the western coast of India and the Indochina Peninsula are known for abundant monsoonal rainfall. The differences in the amplitude of the seasonal Dm changes between the Arabian Sea and the Bay of Bengal can be related to the differences in the size of the prevailing precipitation.
In terms of the relationship between Dm and deep convection characterized by factors such as precipitation-top heights, Houze et al. (2007) investigated the overall variability of summer monsoon convection in the Himalayan region. They suggested that deep convection occurs where the advance edge of the low-level moist southwesterly monsoon flow from the Arabian Sea meets dry air descending from the highland region (i.e., Afghan or Tibetan Plateau). The large value of Dm over the western Himalayan region during the monsoon season (Fig. 4b) also agrees with their results. In addition, Romatschke et al. (2010) investigated the geographical distribution of the probability of finding a deep convective core over the Indian subcontinent during both the premonsoon and monsoon seasons. They showed that deep convective cores dominate the northern Eastern Ghats and Bangladesh Plain during the premonsoon season and over the western Himalayan region during the monsoon season. These results are consistent with the results shown in Fig. 4 of this study, which show that the regions and seasons with deep convective cores correspond well to those with large Dm.
However, the relationship between deep convection and large Dm is probably not a general result on a global scale or for different precipitation system types. For example, Yamaji et al. (2020) showed seasonal differences in Dm and precipitation-top heights. They suggested that extratropical frontal systems dominant in the winter season with moderate precipitation-top heights have larger Dm than the organized precipitation systems prevalent in the summer season, for highly developed precipitation systems over the midlatitude oceans.
We also investigated seasonal changes in precipitation microphysical characteristics from different viewpoints, focusing on lightning activity. Higher precipitation intensity and precipitation-top heights with larger Dm values may be associated with lighting activity. GPM/DPR products have provided increased detection of heavy ice precipitation, which is known to be closely related to lightning activity. Lightning over Asia is more active during the premonsoon season than during the monsoon season. For example, Kodama et al. (2005) demonstrated the seasonal transition between rainfall and lightning activity. Their results indicated that premonsoon rainfall was characterized by more convective rain, consistent with active lightning. In contrast, stratiform and shallow rainfall became more common after the monsoon onset.
We calculated the frequency of heavy ice precipitation using the DPRL2 data, as shown in Fig. 5, indicating whether heavy ice precipitation particles were detected anywhere in the observation column of the DPR. The values were higher over the Indian subcontinent and Indochina Peninsula during the premonsoon season (Fig. 5a). In China, the frequency of heavy ice precipitation is not as high as that of the Indian subcontinent and Indochina Peninsula. Contrary to the premonsoon season, Fig. 5b did not show a clear land–ocean contrast in the monsoon season, whereas the values were relatively higher in the western Himalayan and Tibetan Plateau regions. The seasonal differences in heavy ice precipitation (Fig. 5c) showed a higher frequency of heavy ice precipitation over the Indian subcontinent and the Indochina Peninsula in the premonsoon season. In contrast, seasonal changes were not significant in China. Over the oceans, the frequency of heavy ice precipitation and its seasonal changes were lower than those over land.
As in Fig. 1, but for frequency of heavy ice precipitation. Gray-shaded grids are masked out due to sampling issues.
Citation: Journal of the Atmospheric Sciences 80, 8; 10.1175/JAS-D-22-0198.1
The results in Fig. 5 indicate that heavy ice precipitation particles exist anywhere in the upper layer of the atmosphere more frequently in the premonsoon season than in the monsoon season over the Indian subcontinent and the Indochina Peninsula. These results are consistent with previous results showing that precipitation-top heights were relatively higher in the premonsoon season as well as active lightning activities (Barros et al. 2004; Kodama et al. 2005). Furthermore, this association between high precipitation-top heights, frequent occurrence of ice particles, and active lightning was consistent with the results over the western Himalayan region during the monsoon season.
Thus, the microphysical precipitation characteristics over the entire Asian monsoon region varied seasonally from the premonsoon to the monsoon season. This subsection reveals that microphysical precipitation characteristics of large Dm and frequent heavy ice precipitation were observed over the Indian subcontinent and the Indochina Peninsula in the premonsoon season and over the western Himalayan region during the monsoon season, which can be related to the intense and well-developed precipitation systems.
c. Seasonal transition of the precipitation characteristics over the Bangladesh Plain
In this subsection, we examine the seasonal marches in precipitation microphysical properties, as well as characteristics in the structure of climatological precipitation systems, not only by comparing the two premonsoon and monsoon seasons shown in previous subsections, but also by revealing the seasonal transitions throughout the year, including the postmonsoon season. We focused on the Bangladesh Plain, where the differences between the premonsoon and monsoon seasons are apparent. This study extracted the BP region (23°–27°N, 88°–93°E) as a representative domain.
As shown by the blue bars in all panels in Fig. 6, the precipitation amount had a single peak around the monsoon season, with smaller values in the premonsoon and postmonsoon seasons. The precipitation frequency (solid black line in Fig. 6a) and precipitation amount showed features of similar behavior. These results represent the well-known seasonal changes in precipitation over the Asian monsoon region.
Seasonal march of monthly mean values with 1° × 1° averaged box (latitude and longitude) of (a) precipitation frequency, (b) precipitation intensity (mm h−1), (c) precipitation-top heights (km), (d) Dm (mm), and (e) frequency of heavy ice precipitation over the BP region, all indicated by the solid black line. Blue bars in all panels show precipitation amounts (mm day−1).
Citation: Journal of the Atmospheric Sciences 80, 8; 10.1175/JAS-D-22-0198.1
In contrast, the precipitation intensity and top heights shown in Figs. 6b and 6c had double peaks around May and September, corresponding to pre- and postmonsoon seasons. This result was consistent with that of Takahashi (2016), who used 15 years of TRMM/PR observations from 1998 to 2012. Moreover, Dm near the surface in Fig. 6d shows a double peak.
Notably, Dm increased significantly from January to February, and higher values were maintained during the premonsoon season. However, this feature did not appear in the cases of precipitation intensity and top heights, despite sampling these parameters under the same conditions. These results suggest that the microphysical characteristics exhibited by the large February–March Dm anomaly may not be solely due to changes in the properties of the tall precipitation system with intense precipitation and active lightning, as reported in previous studies.
It should also be emphasized that the frequency of heavy ice precipitation showed a unique seasonal transition that differed from the other variables. For example, Fig. 6e shows a significant peak in April and a small peak in September. This result implies that there may be different precipitation characteristics related to the appearance of heavy ice precipitation in the upper levels between the pre- and postmonsoon seasons, even though the precipitation intensity, precipitation-top heights, and Dm showed similar peak magnitudes in both seasons. The unique seasonal changes in the frequency of heavy ice precipitation also supported the possibility that microphysical characteristics could not vary depending on the known seasonal changes in precipitation characteristics.
Kodama et al. (2005) suggested that maximum lightning counts occurred between 1 and 2 months before monsoon onset, and lightning decreased in the latter half of the monsoon in the Bangladesh region. Their results on the seasonal transition of lightning activity are consistent with our results, showing a noticeable peak in the frequency of heavy ice precipitation in the premonsoon season but not during the monsoon and postmonsoon seasons.
The singular seasonal march in the frequency of heavy ice precipitation can be associated with the differences in surface conditions between the pre- and postmonsoon seasons. For example, Ono and Takahashi (2016) investigated seasonal variations in soil moisture over the Bangladesh region. They found that surface conditions were drier in the premonsoon season than in the postmonsoon season. Related discussions of seasonal changes in the atmospheric environment are presented in section 4b.
Hence, it was suggested that the precipitation characteristics differed among the premonsoon, monsoon, and postmonsoon seasons, focusing on the Bangladesh Plain (BP region). Precipitation amount and frequency had a single peak during the monsoon season, whereas precipitation intensity and top heights showed a double peak during pre- and postmonsoon seasons. These findings on precipitation amount, frequency, intensity, and top heights are consistent with previous studies’ results (Romatschke et al. 2010; Romatschke and Houze 2011; Takahashi 2016). It should be noted that the double peak in Dm, similar to the precipitation intensity and top heights, was observed for the first time in this study, suggesting that a large Dm can be associated with highly developed convection in the premonsoon and postmonsoon seasons.
4. Discussion on the association among precipitation characteristics varying between premonsoon and monsoon seasons
a. Relationship of Dm to R and heavy ice precipitation
The previous section showed that microphysical precipitation characteristics, such as Dm and the frequency of heavy ice precipitation over the Asian monsoon region, varied seasonally from the premonsoon, mature-monsoon, and postmonsoon seasons. Therefore, we performed further analyses to discuss the relationship between the precipitation variables, such as the association of Dm with R [defined in Eq. (2)] and the frequency of heavy ice precipitation.
First, histograms of R and Dm near the surface in the BP region were created to verify the relationship between R and Dm on a DPRL2 pixel basis, as shown in Fig. 7. The samples were limited to R near the surface in the range of 10–100 mm h−1. The value of Dm near the surface was less than 3 mm because the upper limit of Dm was set to 3 mm in the DPRL2 retrieval algorithm (e.g., Seto et al. 2021).
Histogram of (a) R and (b) Dm at the near-surface level only when R at the near-surface level is in the range of 10–100 mm h−1. Values on the y axis are normalized by total precipitating (R > 0.1 mm h−1) samples. Gray solid and black dotted lines show the premonsoon and monsoon seasons, respectively.
Citation: Journal of the Atmospheric Sciences 80, 8; 10.1175/JAS-D-22-0198.1
The R histogram shows that the shapes and mode values were almost the same for the premonsoon and monsoon seasons, as shown in Fig. 7a, although the absolute values of precipitation frequency were different. However, the modes of the Dm differed between the two seasons. For example, the mode value of Dm was approximately 1.7 mm in the monsoon season, whereas there was no significant mode value in the premonsoon season (Fig. 7b). In the premonsoon season, the mode shifted in the larger Dm direction to approximately 2–3 mm. These results suggest that Dm can change even when R is constant. In other words, the relationship between R and Dm can differ among seasons, and precipitation changes do not simply cause seasonal Dm changes. Thus, changes in Dm were probably induced by changes in precipitation characteristics.
Furthermore, we investigated whether the differences in microphysical properties were determined only by the “structure” of precipitation systems because it is well-known that taller convection has higher rainfall intensities. Therefore, we only focused on deep convection for premonsoon and monsoon precipitations to exclude the effects of developing convective systems using the same threshold value of precipitation-top heights. Figure 8 shows the same histogram as Fig. 7; however, the samples were constrained to systems with precipitation-top heights greater than 14 km (approximately corresponding to the top 3% of the total samples in Fig. 7).
As in Fig. 7, but the samples are limited to systems with precipitation-top heights larger than 14 km (highest 3% of total samples of Fig. 7). Values on the y axis are normalized by the total samples with precipitation-top heights above 14 km.
Citation: Journal of the Atmospheric Sciences 80, 8; 10.1175/JAS-D-22-0198.1
The apparent seasonal differences in the shape of the Dm histogram shown in Fig. 7b are reduced in Fig. 8b. This is probably because the effect of the higher frequency of deep convection in the premonsoon season than in the monsoon season was excluded from Fig. 8. However, the shift of the mode toward the larger Dm direction to approximately 2.5–3 mm in the premonsoon season still exists, as shown in Fig. 8b. This result implies a possible change in “microphysical” characteristics and could not be inferred only from changes in the “structure” of precipitation systems known in previous studies (i.e., Romatschke et al. 2010; Romatschke and Houze 2011; Takahashi 2016) although further investigations including the effects in algorithmic issues [e.g., the upper limit of Dm retrievals described in Seto et al. (2021)] should be performed to better address this point.
In addition to the relationship between R and Dm, we investigated the occurrence of heavy ice precipitation as a function of altitude, classified by Dm near the surface. Figure 9 shows the frequency of detected heavy ice precipitation at each altitude for different near-surface Dm ranges of small (0.5 < Dm < 1.0 mm), medium (1.5 < Dm < 2.0 mm), and large (2.5 < Dm < 3.0 mm). Heavy ice precipitation was detected using the same method as Iguchi et al. (2018), as described in section 2, but considering the altitudes because the originally published products had no vertical resolution.
Frequency of detected heavy ice precipitation at each altitude in (a) premonsoon and (b) monsoon seasons. Light gray thick line, medium gray thin line, and thick black line with circles indicate cases of 0.5 < Dm < 1.0 mm, 1.5 < Dm < 2.0 mm, 2.5 < Dm < 3.0 mm, respectively.
Citation: Journal of the Atmospheric Sciences 80, 8; 10.1175/JAS-D-22-0198.1
Heavy ice precipitation in the upper atmosphere above the melting layer was observed more frequently as the Dm increased in both seasons. However, the absolute number of samples was generally almost 10 times larger in the premonsoon season (Fig. 9a) than in the monsoon season (Fig. 9b), which is consistent with the results shown in the previous sections. These showed that heavy ice precipitation particles exist in the upper layer of the atmosphere more frequently in the premonsoon season than in the monsoon season.
Notably, the peak altitudes with frequent heavy ice particles were just above the melting layer at approximately 5 km. Heavy ice particles were observed more frequently in the upper atmosphere (>10 km) as the Dm increased. This tendency is more apparent in the premonsoon season in Fig. 9a.This suggests that deeply developed convection with heavy ice particles is likely to exist more frequently when Dm near the surface is large.
b. Differences in lower-atmospheric stability
In addition to the analyses using GPM/DPR observations, atmospheric conditions over the BP region were confirmed using the JRA-55 dataset, as explained in section 2. The GPM/DPR analysis results in the previous sections did not consider diurnal cycles to ensure sufficient observational sampling for robust statistics. Previous studies (e.g., Ono and Takahashi 2016; Takahashi 2016) have reported that precipitation is dominant in the evening and early morning in the Bangladesh region. In accordance with these results, we used JRA-55 data only at 1800 UTC, corresponding to 0000 LT over the BP region, to extract the representative atmospheric conditions with precipitation. Figure 10 shows the seasonal march of the vertical gradients of dry and moist static energies at 0000 LT (1800 UTC) over the BP region. The vertical gradient represents the differences against pressure, which means that larger values indicate relatively unstable conditions, as dry static energy or moist static energy increases downward from 850 hPa to 950 hPa.
Seasonal march of monthly mean vertical gradients of dry static energy (DSE; solid red line; left axis) and moist static energy (MSE; blue dotted line; right axis) between 850 and 950 hPa at 0000 LT (1800 UTC) over the BP region. Background blue bars show precipitation amount as in Fig. 6 but without a scale.
Citation: Journal of the Atmospheric Sciences 80, 8; 10.1175/JAS-D-22-0198.1
The dry static energy (solid red line in Fig. 10) showed a double peak around March and November, and the values decreased during the summer monsoon period. Dry conditions during the pre- and postmonsoon season can induce lower-atmospheric instabilities, leading to deep convection. In terms of the consistency with changes in microphysical properties, the seasonal changes of dry static energy with double peak was similar to the changes in Dm in Fig. 6d. Moreover, the peak value of dry static energy was larger in March than in November, which is consistent with the result that the frequency of heavy ice precipitation was higher in the premonsoon season than in the postmonsoon season, as shown in Fig. 6e. Ono and Takahashi (2016) showed that the seasonal change in soil moisture followed or coincided with changes in the precipitation amount and frequency, which means that atmospheric surface conditions are drier before the onset of the monsoon than after the monsoon. Their results are in good agreement with the more instable condition at lower atmosphere and deep convection involving lightning and ice particles with a higher probability in premonsoon season compared to the postmonsoon season.
However, the moist static energy (blue dotted line in Fig. 10) exhibits opposite variations to the dry static energy. The values of moist static energy increased from March to April, and relatively high values were maintained during the monsoon season, although the peak values were not during June–August, but in May. It should be noted that the variations in moist static energy corresponded well with the seasonal changes in precipitation amount, as indicated by the blue bars in Fig. 10.
These results suggest that dry convection can be induced in the premonsoon and postmonsoon seasons, whereas moist convection prevails during the monsoon season. Fujinami and Yasunari (2001) presented similar and more precise results, that showed a vertical gradient of potential temperature and equivalent potential temperature over the Tibetan Plateau.
5. Conclusions
This study statistically analyzed 8 years (2014–21) of satellite observation data from the DPRL2 version 06 product to reveal precipitation microphysical characteristics, which are newly available from the dual-frequency radar information derived by GPM/DPR. Before investigating the microphysical characteristics, we confirmed that the 8 years of GPM/DPR data enabled us to capture the climatological seasonal march in rainfall amount using a 2-month average and 1° × 1° horizontal resolution.
To investigate the seasonal variations in precipitation microphysical properties from a climatological perspective, we focused on Dm and the frequency of heavy ice precipitation (i.e., graupel or hail). We observed a large Dm and high frequency of heavy ice precipitation over the Indian subcontinent and Indochina Peninsula during the premonsoon season, when the precipitation amount and frequency were lower over the region, particularly over the northern Eastern Ghats and Bangladesh Plain, and the same was observed over the western Himalayan region during the monsoon season. The regions and seasons corresponded well with the deep convective cores reported by Romatschke et al. (2010), and consistency was confirmed by investigating the seasonal changes in precipitation-top heights. It is noteworthy that the land–ocean contrast showing large Dm and frequent heavy ice precipitation over land was evident over the Indian subcontinent and Indochina Peninsula during the premonsoon season.
A seasonal march of precipitation characteristics, including the postmonsoon season, was examined on the Bangladesh Plain (BP region). Double peaks in Dm appeared during the premonsoon and postmonsoon seasons. Furthermore, the frequency of heavy ice precipitation showed a unique seasonal transition, with a large peak in April and a small peak around September, which was in good agreement with the seasonal variations in lightning activity reported by Kodama et al. (2005).
The relationship between R and Dm was investigated, and the results suggested that Dm can change even if R is within the same range. In other words, the relationship between R and Dm can differ among seasons, and seasonal Dm changes are not simply caused by precipitation changes. However, they are probably induced by changes in precipitation characteristics. The relationship between Dm and heavy ice precipitation was also examined. Notably, the frequency of heavy ice precipitation in the upper atmosphere above the melting layer increased as the Dm near the surface increased. This tendency is more apparent in the premonsoon season and suggests that deep convection with heavy ice particles is likely to occur more frequently when Dm near the surface is large.
Finally, atmospheric conditions over the BP region were confirmed. As Ono and Takahashi (2016) reported, dry surface conditions during the premonsoon season can induce lower-atmospheric instabilities, leading to deep convection involving lightning and ice particles with a high probability. Although the present study focused on analyses mainly using satellite observations, one of the future tasks is to further investigate atmospheric environmental conditions and address these mechanisms.
Acknowledgments.
The authors would like to express their gratitude to Profs. Jun Matsumoto and Hiroshi Matsuyama of Tokyo Metropolitan University for providing valuable comments and suggestions, especially from a geographical perspective. Dr. Takuji Kubota from JAXA offered constructive advice based on satellite data and algorithm insights. The authors also would like to thank three anonymous reviewers for their helpful comments. This study was supported by the Japan Society for the Promotion of Science (JSPS) KAKENHI Grant 22H00037 and the Third Earth Observation Research Announcement (EORA-3) of the Japan Aerospace Exploration Agency (JAXA) Precipitation Measuring Mission Science. The authors declare no conflicts of interest directly relevant to the content of this article.
Data availability statement.
The original GPM/DPR level-2 version 06 product used in this study is openly available from the Japan Aerospace and Exploration Agency and the information is listed at https://www.eorc.jaxa.jp/GPM/en/data_utilization.html. Furthermore, Seto et al. (2021) provide further information regarding the algorithm. The JRA-55 atmospheric dataset is available online from the Japan Meteorological Agency at https://jra.kishou.go.jp/JRA-55/index_en.html (Kobayashi et al. 2015; Harada et al. 2016).
REFERENCES
Akiyama, S., S. Shige, M. K. Yamamoto, and T. Iguchi, 2019: Heavy ice precipitation band in an oceanic extratropical cyclone observed by GPM/DPR: 1. A case study. Geophys. Res. Lett., 46, 7007–7014, https://doi.org/10.1029/2019GL082896.
Awaka, J., M. Le, S. Brodzik, T. Kubota, T. Masaki, V. Chandrasekar, and T. Iguchi, 2021: Development of precipitation type classification algorithms for a full scan mode of GPM Dual-Frequency Precipitation Radar. J. Meteor. Soc. Japan, 99, 1253–1270, https://doi.org/10.2151/jmsj.2021-061.
Barros, A. P., G. Kim, E. Williams, and S. W. Nesbitt, 2004: Probing orographic controls in the Himalayas during the monsoon using satellite imagery. Nat. Hazards Earth Syst. Sci., 4, 29–51, https://doi.org/10.5194/nhess-4-29-2004.
D’Adderio, L. P., G. Vulpiani, F. Porcù, A. Tokay, and R. Meneghini, 2018: Comparison of GPM Core Observatory and ground-based radar retrieval of mass-weighted mean raindrop diameter at midlatitude. J. Hydrometeor., 19, 1583–1598, https://doi.org/10.1175/JHM-D-18-0002.1.
Fujinami, H., and T. Yasunari, 2001: The seasonal and intraseasonal variability of diurnal cloud activity over the Tibetan Plateau. J. Meteor. Soc. Japan, 79, 1207–1227, https://doi.org/10.2151/jmsj.79.1207.
Gatlin, P. N., W. A. Petersen, J. L. Pippitt, T. A. Berendes, D. B. Wolff, and A. Tokay, 2020: The GPM Validation Network and evaluation of satellite-based retrievals of the rain drop size distribution. Atmosphere, 11, 1010, https://doi.org/10.3390/atmos11091010.
Han, M., and S. A. Braun, 2021: Understanding the global three-dimensional distribution of precipitation mean particle size with the Global Precipitation Measurement mission. J. Climate, 34, 9775–9796, https://doi.org/10.1175/JCLI-D-21-0134.1.
Harada, Y., and Coauthors, 2016: The JRA-55 reanalysis: Representation of atmospheric circulation and climate variability. J. Meteor. Soc. Japan, 94, 269–302, https://doi.org/10.2151/jmsj.2016-015.
Hirose, M., and K. Nakamura, 2005: Spatial and diurnal variation of precipitation systems over Asia observed by the TRMM precipitation radar. J. Geophys. Res., 110, D05106, https://doi.org/10.1029/2004JD004815.
Hou, A. Y., and Coauthors, 2014: The Global Precipitation Measurement mission. Bull. Amer. Meteor. Soc., 95, 701–722, https://doi.org/10.1175/BAMS-D-13-00164.1.
Houze, R. A., Jr., D. C. Wilton, and B. F. Smull, 2007: Monsoon convection in the Himalayan region as seen by the TRMM Precipitation Radar. Quart. J. Roy. Meteor. Soc., 133, 1389–1411, https://doi.org/10.1002/qj.106.
Iguchi, T., N. Kawamoto, and R. Oki, 2018: Detection of intense ice precipitation with GPM/DPR. J. Atmos. Oceanic Technol., 35, 491–502, https://doi.org/10.1175/JTECH-D-17-0120.1.
Islam, M. N., and H. Uyeda, 2008: Vertical variations of rain intensity in different rainy periods in and around Bangladesh derived from TRMM observations. Int. J. Climatol., 28, 273–279, https://doi.org/10.1002/joc.1585.
Kobayashi, S., and Coauthors, 2015: The JRA-55 reanalysis: General specifications and basic characteristics. J. Meteor. Soc. Japan, 93, 5–48, https://doi.org/10.2151/jmsj.2015-001.
Kodama, Y.-M., A. Ohta, M. Katsumata, S. Mori, S. Satoh, and H. Ueda, 2005: Seasonal transition of predominant precipitation type and lightning activity over tropical monsoon areas derived from TRMM observations. Geophys. Res. Lett., 32, L14710, https://doi.org/10.1029/2005GL022986.
Kojima, M., and Coauthors, 2012: Dual-Frequency Precipitation Radar (DPR) development on the Global Precipitation Measurement (GPM) Core Observatory. Proc. SPIE, 8528, 85281A, https://doi.org/10.1117/12.976823.
Kozu, T., and Coauthors, 2001: Development of precipitation radar onboard the Tropical Rainfall Measuring Mission (TRMM) satellite. IEEE Trans. Geosci. Remote Sens., 39, 102–116, https://doi.org/10.1109/36.898669.
Kozu, T., K. K. Reddy, S. Mori, M. Thurai, J. T. Ong, D. N. Rao, and T. Shimomai, 2006: Seasonal and diurnal variations of raindrop size distribution in Asian monsoon region. J. Meteor. Soc. Japan, 84A, 195–209, https://doi.org/10.2151/jmsj.84A.195.
Kummerow, C., W. Barnes, T. Kozu, J. Shiue, and J. Simpson, 1998: The Tropical Rainfall Measuring Mission (TRMM) sensor package. J. Atmos. Oceanic Technol., 15, 809–817, https://doi.org/10.1175/1520-0426(1998)015<0809:TTRMMT>2.0.CO;2.
Lau, K.-M., G. J. Yang, and S. H. Shen, 1988: Seasonal and intraseasonal climatology of summer monsoon rainfall over East Asia. Mon. Wea. Rev., 116, 18–37, https://doi.org/10.1175/1520-0493(1988)116<0018:SAICOS>2.0.CO;2.
Le, M., and V. Chandrasekar, 2021: Graupel and hail identification algorithm for the Dual-Frequency Precipitation Radar (DPR) on the GPM Core Satellite. J. Meteor. Soc. Japan, 99, 49–65, https://doi.org/10.2151/jmsj.2021-003.
Masaki, T., T. Iguchi, K. Kanemaru, K. Furukawa, N. Yoshida, T. Kubota, and R. Oki, 2022: Calibration of the Dual-Frequency Precipitation Radar onboard the Global Precipitation Measurement Core Observatory. IEEE Trans. Geosci. Remote Sens., 60, 5100116, https://doi.org/10.1109/TGRS.2020.3039978.
Matsumoto, J., 1997: Seasonal transition of summer rainy season over Indochina and adjacent monsoon region. Adv. Atmos. Sci., 14, 231–245, https://doi.org/10.1007/s00376-997-0022-0.
Meneghini, R., and H. Kim, 2017: Minimizing the standard deviation of spatially averaged surface cross-sectional data from the Dual-Frequency Precipitation Radar. IEEE Trans. Geosci. Remote Sens., 55, 1709–1716, https://doi.org/10.1109/TGRS.2016.2630669.
Meneghini, R., H. Kim, L. Liao, J. A. Jones, and J. M. Kwiatkowski, 2015: An initial assessment of the surface reference technique applied to data from the Dual-Frequency Precipitation Radar (DPR) on the GPM satellite. J. Atmos. Oceanic Technol., 32, 2281–2296, https://doi.org/10.1175/JTECH-D-15-0044.1.
Meneghini, R., L. Liao, J. Kwiatokowki, and T. Iguchi, 2021: Path attenuation estimates for the GPM Dual-Frequency Precipitation Radar (DPR). J. Meteor. Soc. Japan, 99, 181–200, https://doi.org/10.2151/jmsj.2021-010.
Nakamura, K., 2021: Progress from TRMM to GPM. J. Meteor. Soc. Japan, 99, 697–729, https://doi.org/10.2151/jmsj.2021-035.
Ono, M., and H. G. Takahashi, 2016: Seasonal transition of precipitation characteristics associated with land surface conditions in and around Bangladesh. J. Geophys. Res. Atmos., 121, 11 190–11 200, https://doi.org/10.1002/2016JD025218.
Petersen, W. A., 2017: Use of polarimetric radar for evaluating GPM satellite-based retrievals of the rain drop size distribution. 2017 Fall Meeting, New Orleans, LA, Amer. Geophys. Union, Abstract A34A-01, https://agu.confex.com/agu/fm17/meetingapp.cgi/Paper/215702.
Petersen, W. A., P. N. Gatlin, D. B. Wolff, A. Tokay, and M. Grecu, 2018: A radar-based evaluation of GPM retrievals of the rain drop size distribution. 10th European Conf. on Radar in Meteorology and Hydrology, Wageningen, Netherlands, Wageningen University and Research, 10.3, https://doi.org/10.18174/454537.
Prabha, T. V., A. Khain, R. S. Maheshkumar, G. Pandithurai, J. R. Kulkarni, M. Konwar, and B. N. Goswami, 2011: Microphysics of premonsoon and monsoon clouds as seen from in situ measurements during the Cloud Aerosol Interaction and Precipitation Enhancement Experiment (CAIPEEX). J. Atmos. Sci., 68, 1882–1901, https://doi.org/10.1175/2011JAS3707.1.
Radhakrishna, B., S. K. Satheesh, T. N. Rao, K. Saikranthi, and K. Sunilkumar, 2016: Assessment of DSDs of GPM-DPR with ground-based disdrometer at seasonal scale over Gadanki, India. J. Geophys. Res. Atmos., 121, 11 792–11 802, https://doi.org/10.1002/2015JD024628.
Radhakrishna, B., K. Saikranthi, and T. N. Rao, 2020: Regional differences in raindrop size distribution within Indian subcontinent and adjoining seas as inferred from Global Precipitation Measurement Dual-Frequency Precipitation Radar. J. Meteor. Soc. Japan, 98, 573–584, https://doi.org/10.2151/jmsj.2020-030.
Romatschke, U., and R. A. Houze Jr., 2011: Characteristics of precipitating convective systems in the premonsoon season of South Asia. J. Hydrometeor., 12, 157–180, https://doi.org/10.1175/2010JHM1311.1.
Romatschke, U., S. Medina, and R. A. Houze Jr., 2010: Regional, seasonal, and diurnal variations of extreme convection in the South Asian region. J. Climate, 23, 419–439, https://doi.org/10.1175/2009JCLI3140.1.
Ryu, J., H. J. Song, B. J. Sohn, and C. Liu, 2021: Global distribution of three types of drop size distribution representing heavy rainfall from GPM/DPR measurements. Geophys. Res. Lett., 48, e2020GL090871, https://doi.org/10.1029/2020GL090871.
Seneviratne, S. I., and Coauthors, 2021: Weather and climate extreme events in a changing climate. Climate Change 2021: The Physical Science Basis, V. Masson-Delmotte et al., Eds., Cambridge University Press, 1513–1766.
Seto, S., T. Iguchi, R. Meneghini, J. Awaka, T. Kubota, T. Masaki, and N. Takahashi, 2021: The precipitation rate retrieval algorithms for the GPM Dual-Frequency Precipitation Radar. J. Meteor. Soc. Japan, 99, 205–237, https://doi.org/10.2151/jmsj.2021-011.
Skofronick-Jackson, G., and Coauthors, 2017: The Global Precipitation Measurement (GPM) mission for science and society. Bull. Amer. Meteor. Soc., 98, 1679–1695, https://doi.org/10.1175/BAMS-D-15-00306.1.
Sugimoto, S., and H. G. Takahashi, 2017: Seasonal differences in precipitation sensitivity to soil moisture in Bangladesh and surrounding regions. J. Climate, 30, 921–938, https://doi.org/10.1175/JCLI-D-15-0800.1.
Takahashi, H. G., 2016: Seasonal and diurnal variations in rainfall characteristics over the tropical Asian monsoon region using TRMM-PR data. SOLA, 12A, 22–27, https://doi.org/10.2151/sola.12A-005.
Takahashi, H. G., and T. Yasunari, 2006: A climatological monsoon break in rainfall over Indochina—A singularity in the seasonal march of the Asian summer monsoon. J. Climate, 19, 1545–1556, https://doi.org/10.1175/JCLI3724.1.
Takahashi, T., 2006: Precipitation mechanisms in East Asian monsoon: Videosonde study. J. Geophys. Res., 111, D09202, https://doi.org/10.1029/2005JD006268.
Wang, B., and LinHo, 2002: Rainy season of the Asian–Pacific summer monsoon. J. Climate, 15, 386–398, https://doi.org/10.1175/1520-0442(2002)015<0386:RSOTAP>2.0.CO;2.
Welch, B. L., 1938: The significance of the difference between two means when the population variances are unequal. Biometrika, 29, 350–362, https://doi.org/10.1093/biomet/29.3-4.350.
Yamaji, M., H. G. Takahashi, T. Kubota, R. Oki, A. Hamada, and Y. N. Takayabu, 2020: 4-year climatology of global drop size distribution and its seasonal variability observed by spaceborne Dual-Frequency Precipitation Radar. J. Meteor. Soc. Japan, 98, 755–773, https://doi.org/10.2151/jmsj.2020-038.
Yamane, Y., and T. Hayashi, 2006: Evaluation of environmental conditions for the formation of severe local storms across the Indian subcontinent. Geophys. Res. Lett., 33, L17806, https://doi.org/10.1029/2006GL026823.
Zipser, E. J., D. J. Cecil, C. Liu, S. W. Nesbitt, and D. P. Yorty, 2006: Where are the most intense thunderstorms on Earth? Bull. Amer. Meteor. Soc., 87, 1057–1072, https://doi.org/10.1175/BAMS-87-8-1057.
