2024 Volume 102 Issue 6 Pages 665-676
The study investigates the relationship between climate pattern, rainfall, and internal warm cloud structure in Thailand using ground-based meteorological and satellite data. Situated in the tropical climate zone, Thailand faces drought conditions exacerbated by changes in warm cloud patterns due to climate change. Results indicate a negative correlation between rainfall and Oceanic Niño Index (ONI) and Pacific Decadal Oscillation (PDO), with rainfall decreasing as ONI and PDO values increase, especially in April and October. In contrast, Indian Ocean Dipole (IOD) shows a weaker correlation with the rainfall during the period from 1991 to 2022.
During the period from 2006 to 2014, the internal warm cloud structure over Thailand was analyzed using contoured frequency by optical depth diagrams (CFODDs) derived from CloudSat Cloud Profiling Radar (CPR) and Aqua Moderate Resolution Imaging Spectroradiometer (MODIS). The results show the warm cloud process, indicated by the transfer of cloud droplets to drizzle and rain as cloud particle size increases near the cloud top. The cloud droplet mode occurs from 4 µm to 15 µm, drizzle mode occurs from 15 µm to 21 µm and dominance of rain mode occurs above 21 µm.
Differences in cloud structures are observed under seasonal and El Niño–Southern Oscillation (ENSO) phases, with the warm clouds contain larger cloud droplet effective radius (Re) during the wet season compared to the dry season. Meanwhile, in cloud structure exhibit thicker cloud and faster transfer drizzle mode to rain mode with Re increasing during La Niña phase compared to El Niño phase. These differences in cloud structures are attributed to variations in aerosols, differences in humidity, and temperature influenced by geographical characteristics.
Climate change characterized by soaring temperatures, accelerated snow and glacier melt, and elevated sea levels, exacerbates widespread drought conditions globally, with Southeast Asia bearing a significant impact (Eckstein and Schäfer 2021). These climate dynamics are closely intertwined with cloud behavior, precipitation patterns, and the frequency of drought events.
Thailand, situated between 5°N to 21°N and 97°E to 106°E, grapples with a tropical climate characterized by distinct wet and dry seasons driven by the southwest and northeast monsoons (Chabangborn 2012). The monsoons bring about cycles of rain and dry spells, with droughts worsened by phenomena such as El Niño–Southern Oscillation (ENSO). Between 2007 and 2014, droughts intensified, affecting an average agricultural area of 2,486,382,400 square meters per year with an average damage value of more than 18 million US dollars annually, indicative of an upward trend (Institute of Ombudsman Studies 2016). Monitoring ENSO events involves assessing ocean indices including the Oceanic Niño Index (ONI), Pacific Decadal Oscillation (PDO), and Indian Ocean Dipole (IOD) (Lam et al. 2019; Gale and Saunders 2013).
Thailand can experience rain from both warm and ice clouds, typically with predominantly warm clouds, followed by mixed and cold clouds, with cloud top temperatures ranging from 282.09 K to 185.51 K depending on the season and analysis of MTSAT-TIR imagery from 2006–2007 revealed that liquid clouds dominate over Thailand throughout the year. Even during the wet season, many large cold clouds persist for several hours; however, the most clouds are observed still liquid clouds, with temperatures ranging from 10 °C to −10 °C (Bumrungklang et al. 2009).
Satellites play a pivotal role in climate monitoring, providing comprehensive, continuous, and representative global data. This data is improving numerical simulations of clouds and enhancing our understanding of cloud microphysical properties. Nakajima et al. (2010a, b) introduced contoured frequency by optical depth diagrams (CFODDs) to visualize the evolution of warm clouds, utilizing data from instruments of Aqua’s Moderate Resolution Imaging Spectroradiometer (MODIS) and CloudSat’s Cloud Profiling Radar (CPR). CFODDs depict cloud evolution based on incloud optical depth (COD) and cloud droplet effective radius (Re), as cloud particle radius near the cloud top closely relates to the cloud evolution state (Nakajima 2010b). Similar structures have been observed in CFODDs obtained through statistical methods and time series analysis (Sato et al. 2012). CFODDs effectively evaluate low-level cloud evolution, including condensation, growth, and extinction, assessing cloud microphysical processes validity in global-scale models using CFODDs derived from low-level cloud signals worldwide (Suzuki et al. 2013a; Jing et al. 2017; Jing and Suzuki 2018). Additionally, Suzuki et al. (2013b) examined aerosol effects on low-level cloud processes based on CFODDs.
Although CFODDs have been employed globally to study cloud evolution (Nakajima 2010b), including diverse regions like East Asia, California, Peru. (Matsumoto et al. 2023), and the Eastern Asian, Tropical Warm Pool, Australian, North Atlantic, and Equatorial Cold Tongue regions (Michibata et al. 2019), there’s been limited focus on warm clouds over the tropical zone associated with seasonal and climate variations. This study aims to fill this gap by investigating the cloud evolution process, with a focus on Thailand areas. The goal is to understand the impact of cloud processes during climate change, particularly their connection to the ENSO climate pattern, which directly influences rainfall over land in this region, this study marks the first attempt to elucidate the internal structure of warm clouds over Thailand using CFODDs.
The study’s location covered land in Thailand region, from 3°N to 23°N and 95°E to 108°E (Fig. 1), the analysis region using the QGIS, version 3.30 (QGIS 2021). In this region, rainfall data were obtained from rain gauges, sourced from the National Hydroinformatics Data Center, a government agency of Thailand. Climate patterns are related to ocean indices, which track changing sea surface temperatures in specific ocean areas. These indices include ONI from a 3-month running mean of ERSST.v5 sea surface temperature anomalies in the Niño 3.4 region, PDO, and IOD that are obtained from the National Oceanic and Atmospheric Administration (NOAA) and the Japan Meteorological Agency (JMA).
Map of the analysis region, the white box indicates the land area in Thailand analyzed in this work [3–23°N, 95–108°E].
This study used satellite data from Aqua/MODIS and CloudSat/CPR, both of which are part of the A-train constellation, provides an estimate of cloud droplet collection efficiency in single-layer warm liquid clouds (SLWCs). Aqua/MODIS sensor provides data including cloud property products, aerosols, and surface temperatures using its 36 channels (Baum and Platnick 2006). CloudSat/CPR sensor emit millimeter waves and observe backscatter from cloud particles (Stephens et al. 2008). The CPR sensor is valuable in collecting vertical profiles of cloud microphysical properties, offering direct information on cloud microphysical processes by reflectivity (Witkowski and Livermore 2018).
2.2 MethodThe rainfall in Thailand was analyzed and correlated with the ENSO climate pattern using the ONI, PDO, and IOD index values from 1991 to 2022. The internal warm cloud structure was investigated by CFODDs between 2006 and 2014. The CFODDs using cloud properties products from the Aqua/MODIS satellite that are obtained from the Comprehensive Analysis Program for Cloud Optical Measurements (CAPCOM) algorithm (Nakajima and Nakajima 1995; Kawamoto et al. 2001; Nakajima et al. 2009), the products are cloud optical thickness (COT), Cloud droplet effective radius (CDR), and cloud top temperature, which are observed at specific wavelengths (0.68, 2.1, and 10.8 µm) respectively. In additional study of Nakajima et al. (2010a), CDR from the 2.1 µm channel of MODIS can distinguish cloud droplets and drizzle particles and offers advantages over the 3.7 µm channel in terms of its sensitivity to drizzle particles. The behavior of CDR retrieved from the 1.6 µm channel was similar to that of CDR retrieved from the 2.1 µm channel. Moreover, warm cloud phase identification was performed using the Cloud and Aerosol Unbiased Decision Intellectual Algorithm (CLAUDIA) algorithm developed by Ishida and Nakajima (2009), the CLAUDIA algorithm was employed to address partial cloud cover along the edges of clouds within a pixel. The COT and CDR values obtained from CAPCOM were used. In this study, the threshold for cloud top temperature was set at 260 K, aligning with ISCCP guidelines. COT, measured from solar radiation reflected by clouds, is examined for both liquid water and ice clouds. A temperature threshold of 260 K is used to characterize warm clouds, with specific differentiation based on temperature (Raschke et al. 2005).
The CFODDs are created by combining cloud properties such as COT, selected CDR (Re), and radar reflectivity. COT, serving as the vertical axis, is determined using the adiabatic condensation growth model, which is represented by the in-COD. The 2B-TAU product is used only for identifying cloud top bin. The COT slicing in our CFODDs had been given using the adiabatic growth assumption. The COT from the cloud bottom to a given height (h), denoted as the total optical thickness (Tc), with the geometric thickness of the cloud (H), is represented by the internal COD at each height (h) (Suzuki et al. 2010) followed equation as.
 |
While, CloudSat CPR provide radar reflectivity data (Ze), obtained from the 2B-GEOPROF product that represent in horizontal axis. The CFODD represents cloud properties in terms of COD and Ze. The Ze ranges correspond to different particle modes: less than −15 dBZe indicates cloud droplet mode, −15 dBZe to 0 dBZe indicates drizzle mode, and 0 dBZe to 15 dBZe indicates rain mode. These ranges help identify cloud condensation, collision, and precipitation processes, which involve the internal cloud structure. With the changes in structure associated with increasing Re, CFODDs can provide insights into the growth of clouds from cloud droplets to rain (Nakajima et al. 2010b).
Over 32 years (1991–2022), rainfall data were transformed to normal percentage rainfall (NPR) for a year to study trends in rainfall with ocean indices (Fig. 2). The NPR of rainfall data showed fluctuations, the highest annual rainfall was 1848 mm, which is 23 % above the NPR in 2022. In contrast, the lowest recorded rainfall was 1218 mm, which is 19 % below the NPR in 2019. When comparing NPR with ocean indices such as ONI, PDO, and IOD, a strong correlation is observed between NPR and ONI and PDO, but a weaker correlation is noted with IOD. Typically, when NPR values are above 0, these indices tend to be lower than normal, while NPR tends to exceed the normal level when the indices are below 0 (Fig 2). The association between Thailand’s rainfall and the changing sea surface temperatures of the Pacific and Indian oceans is highlighted by ONI, PDO, and IOD, respectively.
Comparison of annual NPR with ocean indices.
The correlation coefficients between annual rainfall and these indices were explained, revealing a strong negative correlation between ONI and annual rainfall (−0.73), and a negative correlation between PDO and annual rainfall (−0.61). In contrast, IOD shows a weak positive correlation, as illustrated in Table 1.
Analyzing the average annual rainfall in Thailand by season (Fig. 3), which consists of the dry season (November to April) and the wet season (May to October), each spanning 6 months over the period from 1991 to 2022. This analysis is based on months influenced by El Niño phase (ONI values over 0) and La Niña phase (ONI values below 0). The difference in rainfall between the dry and wet seasons amounted to 905 mm and 909 mm during the El Niño and La Niña phases, respectively. Focusing on the comparison between El Niño and La Niña phases in each season, rainfall during El Niño is lower than during La Niña. Rainfall errors vary between 40–70 mm in the dry season and between 90–96 mm in the wet season. On average, rainfall during both seasons was approximately 100 mm lower during El Niño phase compared to La Niña phase. In Table 1, the ONI index exhibits a very strong negative correlation of −0.79 with dry season rainfall and a correlation value of −0.57 in the wet season. Meanwhile, PDO demonstrates a moderate to strong negative correlation with rainfall in the dry and wet seasons, respectively, while IOD exhibits a weak correlation with rainfall in both seasons. The monthly accumulated rainfall data revealed a very strong negative correlation with the ONI at −0.81 in April, moderate negative correlations of −0.46 in March and −0.49 in May, and a strong negative correlation of −0.58 in October. PDO also shows moderate negative correlations with the monthly rainfall of −0.48 in April and −0.39 in October, as well as weak correlations and moderate negative correlations of the IOD with the monthly rainfall of 0.28 in March and −0.41 in November, respectively (Fig. 4). These results show ENSO climate pattern, indicated by ocean indices that affect Thailand’s rainfall depending on different conditions between the wet and dry seasons, including air stability and varying aerosols. These factors will be discussed in further study.
Average rainfall divided into two distinct periods, dry season and wet season, each spanning 6 months. The error bars represent the variation in monthly rainfall within each period, influenced by ENSO phases from 1991 to 2022.
Correlation coefficients between monthly ocean indices and rainfall in Thailand for period of 1991–2022.
CFODDs were used to reveal the overall warm cloud characteristics in Thailand, considering both seasonal variations and the ENSO climate pattern. The analysis used 9 years from 2006 to 2014. Six threshold values of cloud droplet effective radius (Re) were selected to examine the following ranges: 4 < Re < 12, 12 < Re < 15, 15 < Re < 18, 18 < Re < 21, 21 < Re < 27, and 27 µm < Re < 36 µm.
a. Cloud characteristics in ThailandAs Re expands from 4 µm to 12 µm (Fig. 5), the cloud droplet mode becomes dominant, leading to a condensation process that results in COD of approximately 40. With the Re range extending to 15 µm, the cloud droplet mode continues to dominate, and COD increases to above 40. A transition from the cloud droplet mode to the drizzle and rain modes occurs at the cloud base when Re reaches 15–18 µm, with COD remaining above 40. This transition is characterized by collision and coalescence processes, along with condensation from the cloud top. Between Re of 18 µm and 21 µm, the drizzle and rain modes are more observed from the middle to the base of the cloud, resulting in a decrease in COD to 40. Condensation processes are observed at the cloud top. Finally, when Re exceeds 21 µm, the rain modes dominate, with COD at 40. COD decreases to 30 when Re extends to 27 µm.
CFODDs classified by Re with threshold values between 4 µm and 36 µm for Thailand.
During the wet season (Fig. 6), a high frequency of condensation processes occurs when Re ranges between 4 µm and 12 µm from the top to the base of the cloud, with a COD of 40, and from the top to the middle of the cloud when Re ranges between 12 µm and 15 µm. When Re reaches 15–18 µm, the COD is approximately 40. A transition from the cloud droplet mode to the drizzle mode occurs in the middle of the cloud, and a transition from the drizzle mode to the rain mode occurs at the cloud base. Between Re of 18 µm and 21 µm, the drizzle and rain modes are observed from the middle to the base of the cloud, with a COD below 40. Between Re of 21 µm and 27 µm, the rain mode dominates from the middle to the base, with a COD below 40. Finally, when Re exceeds 27 µm, the COD decreases to 30, with rain modes dominating. In contrast, during the dry season (Fig. 7), Re between 4 µm and 12 µm exhibit a COD of 40, which increases to 50 when Re ranges from 12 µm to 15 µm. Between Re of 15 µm and 18 µm, the COD is approximately 50, a transition from the cloud droplet mode to the drizzle and rain modes occurs near the base of the cloud. Eventually, with Re extending from 18 µm to 21 µm, a transition from the drizzle mode to the rain mode is most observed from the middle to the base of the cloud, with COD rapidly decreasing to below 30 and dissipating after that. For the CFODD used in the analysis, it indicates that warm clouds during the dry season had smaller Re, approximately 4–21 µm, compared to wet season clouds, which had Re ranging from 4–36 µm.
CFODDs classified by Re with threshold values between 4 µm and 36 µm for Thailand during wet season.
CFODDs classified by Re with threshold values between 4 µm and 36 µm for Thailand during dry season.
Figures 8 and 9 illustrate warm cloud structure during La Niña and El Niño phase. During the La Niña phase, a high frequency of condensation processes occurs when Re ranges between 4 µm and 15 µm, with a COD above 40. When Re reaches 15–18 µm, the COD is approximately 40, a transition from the cloud droplet mode to the drizzle and rain modes occurs in the middle and at the base of the cloud, respectively, along with condensation from the cloud top. Between Re of 18 µm and 21 µm, the drizzle and rain modes appear from the middle to the base of the cloud, with COD decreasing below 40. Between Re of 21 µm and 27 µm, the rain mode dominates from the middle to the base, with COD below 40. Finally, when Re exceeds 27 µm, the COD decreases to 30, with rain modes dominating. However, during El Niño phase, Re between 4 µm and 12 µm exhibit a COD below 40, which increases to approximately 40 when Re ranges from 12 µm to 15 µm. Between Re of 15 µm and 18 µm, a transition from the cloud droplet mode to drizzle is most observed from the middle to the base cloud, with COD decreasing to above 30. Between Re of 18 µm and 27 µm, a transition from the drizzle mode to the rain mode is most observed from the middle to the base cloud, with COD decreasing to below 30. Eventually, Re exceeds 27 µm, the COD decreases to approximately 20 with the domination of rain modes. For the CFODD analysis, warm clouds under the El Niño phase show a slower transition from the drizzle mode to the rain mode and exhibit thinner clouds compared to warm clouds under the La Niña phase.
CFODDs classified by Re with threshold values between 4 µm and 36 µm for Thailand under La Niña phase.
CFODDs classified by Re with threshold values between 4 µm and 36 µm for Thailand under El Niño phase.
In Section 3.1, the results reveal a correlation between the ENSO climate pattern and the amount of rainfall in Thailand, which is located in Southeast Asia. A 2015 Met Office report illustrates reduced rainfall in Southeast Asia during El Niño phase, the historical analysis shows that ENSO is associated with anomalous surface temperatures in both the Pacific and Indian oceans. During El Niño events, weakened surface winds in the tropical Pacific result in warmer ocean temperatures in the central and eastern regions. This induces increased sinking air motion over Southeast Asia, leading to elevated surface pressure and dry conditions. Consequently, below-average rainfall occurs, contrasting with conditions during La Niña phases (Bogale and Temesgen 2021). As the result, the ONI and PDO indices demonstrate a strong negative correlation with heavy rainfall in Thailand, whereas the IOD index exhibits a weaker correlation. According to studies by Ueangsawat et al. (2015), the ENSO phase impacts the variability of temperature and rainfall. During the El Niño phase, there is an increase in maximum surface temperature in Thailand, particularly significant in May and June, and a tendency for decreased rainfall during these months. Consequently, studies by Kirtphaiboon et al. (2014) observed that ENSO phases lead rainfall anomalies in Thailand, with a lag of 4 months, based on historical data spanning from 1971 to 2010.
Thailand’s seasons are categorized into dry and wet seasons based on warm temperatures and variations in humidity levels. The dry season is characterized by warmer temperatures and lower humidity, influenced by factors such as northeast monsoon, cold air masses, upper westerly winds, thermal low pressure, local convection, and the sun’s declination. In contrast, the wet season is dominated by southwest monsoon, monsoon trough, tropical cyclones, and low-pressure areas from the ocean, which introduce marine aerosols and bring lower temperatures and higher humidity (Laonamsai et al. 2021). The ONI exhibited a stronger influence on dry season rainfall more than wet season, showing very strong and strong negative correlations, respectively. The ONI had the most significant impact in April and October, preceding and following the wet season, respectively. Meanwhile, the influence of the PDO remained consistent across both seasons. Positive ONI index values are strongly associated with dry conditions, promoting biomass burning and resulting in increased aerosol optical depth (AOD) over the maritime continent. This induces a decrease in relative humidity and reduced precipitation over Southeast Asia beyond normal levels for the season (Ng et al. 2017). Bridhikitti (2013) found that ENSO significantly affects Thai rainfall in the pre-monsoon and post-monsoon months of the following year, significantly decreased rainfall with increasing AOD during the post-summer monsoon months in inland regions.
The CFODDs in Section 3.2 provide internal warm cloud structure of Thailand that represent clouds grow with increases in Re. The studies conducted in Eastern Asia indicate that the macrophysical structures of clouds (such as liquid water path and cloud thickness) and their dynamical regimes (including lower-tropospheric stability and updraft velocity) exhibit considerable seasonal variability. This variability leads to significant differences in how clouds convert to rain depending on land and ocean (Takahashi et al. 2017; Michibata et al. 2014). The warm clouds over Thailand progress from cloud droplet to drizzle and eventually to rain, and that COD depends on the size of the cloud particles. The behavior of the COD during cloud growth in Thailand changes, which can be explained by the precipitation process. Re increases from 4 µm to 15 µm during the condensation phase, increasing the COD to higher than those of the other phases. When Re reaches 15 µm to 21 µm, the collision process becomes more prevalent, leading to the appearance of drizzle droplets. Eventually, the warm cloud undergoes a dissipation stage, leading to cloud disintegration that information is based on a study conducted by Nakajima et al. (2010b). According to studies by Michibata et al. (2019), warm rain is observed from a small Re range (5 µm < Re < 12 µm) in the Thailand region. Regarding a study by Matsumoto et al. (2023), it demonstrates that the characteristics of clouds over Asia include a high frequency of various cloud formations with Re of 6–27 µm, as indicated by the broad frequency of CFODD results. This is consistent with the warm cloud patterns observed over Thailand, which also exhibit a high frequency of various cloud formations.
The difference of warm cloud structure depends on the season (see Section 3.2a) can be explained by a transition behavior, COD and Re. During the wet season, clouds undergo collision and coalescence processes when Re reaches 12–15 µm, while during the dry season, clouds remain in condensation processes, as indicated by the cloud frequency. Between Re of 15–18 µm, dry season clouds experience rainfall at the base of the cloud, after which the cloud dissipates and loses COD. In contrast, warm clouds during the wet season can maintain COD and continue cloud processes as Re increases. These phenomena can be explained by the differences in temperature, humidity, and aerosol conditions between the two seasons. Typically, during dry season, there is a combination of higher surface temperature and lower surface humidity, with temperatures reaching 33.1 °C and humidity at 73 %, which can increase to 44.6 °C. In contrast, during wet season, temperatures average around 32.5 °C with humidity at 80.4 % (Long term climate change monitoring and warning working group 2021). Additionally, dry season is characterized by increased aerosols from biomass burning, competing with available moisture and leading to the formation of clouds with smaller droplet sizes (Lamb and Verlinde 2011). As a result, warm clouds are harder to form during the dry season due to the inhibition of cloud processes. This leads to clouds dissipating as Re increases. This result aligns with a study by Sapphaphab and Ruangrungrote (2019) in Chiang Mai, Thailand, which observed that dry season conditions inhibit cloud formation compared to the wet season. This is indicated by the lower cloud base height and more complicated cloud patterns observed during the wet season compared to dry season. Moreover, when comparing the results with Precipitable Water (PW) during the period (Fig. 10), which indicates the precipitation potential over Thailand, it becomes evident that there is a difference between the wet and dry seasons in 2011.
An overview of precipitable water over Thailand using MYD08_M3 MODIS data from (a) April 2011 in term of dry season and La Niña phase, (b) April 2015 in term of dry season and El Niño phase, (c) October 2011 in term of wet season and La Niña phase, and (d) October 2015 in term of wet season and El Niño phase.
For the warm cloud under ENSO phases (see Section 3.2c), as a result, the warm cloud during the La Niña phase is thicker compared to the El Niño warm cloud as Re increases. Additionally, the La Niña clouds rapidly transform from cloud droplets to rain when Re ranges between 15 µm and 18 µm. The effect of ENSO phases on warm clouds is significant. This climate phenomenon can impact warm clouds through factors such as air speed, the Pacific cold tongue process, and atmospheric moisture conditions, which vary depending on the phase of ENSO (Watanabe et al. 2011). In East Asia, El Niño phase shows negative effects on warm clouds, with opposing effects during the La Niña phase. During the El Niño phase, there is a decrease in precipitation, water content, and water vapor (Zhang and Sumi 2002). Additionally, during the El Niño phase, there is an increase in fire aerosols in equatorial Asia. These fire aerosols can reduce precipitation by cooling the surface, thereby reducing humidity and surface convergence. This contributes to increased static stability and heating of the troposphere, which inhibits convection wind and cloud processes (Tosca et al. 2010), resulting in slower cloud-to-rain transition due to aerosol burdens. Moreover, when comparing the results with PW during the period under ENSO phases (Fig. 10), it becomes evident that there is a difference between the La Niña phase and the El Niño phase.
Therefore, this study suggests that warm cloud process over Thailand expands from droplets to rain, with diverse cloud types that increase Re from 4 µm to 36 µm depending on the season and ENSO phase. Higher temperatures and fire aerosol levels, along with lower humidity during the dry season, inhibit cloud processes, as evidenced by the smaller Re observed in warm clouds compared to the wet season. El Niño phases in Thailand induce drier atmospheric conditions, higher fire aerosol levels, and sinking air conditions, contrasting with La Niña phases. El Niño phases inhibit warm cloud growth processes, resulting in thinner COD and slower rain cloud processes as Re increases.
All rain gauge data used in this study were obtained from the National Hydroinformatics Data Center, which can be accessed online at (https://www.tmd.go.th/en/ClimateChart). The ocean index data in this study can be found online from the NOAA (https://psl.noaa.gov) and the JMA (https://ds.data.jma.go.jp/tcc/tcc/products/elnino/index.html). Eventually, the CloudSat and Aqua data can be requested by contacting Takashi Y. Nakajima (nkjm@yoyogi.ycc.u-tokai.ac.jp).
The authors are grateful to Yu Matsumoto (who graduated from Tokai University in March 2023) for her data analysis of the CloudSat and Aqua data. This research was supported by a Grant-in-Aid for Scientific Research (B) 21H01159 from the Japan Society for the Promotion of Science (JSPS), and the EarthCARE project (JX-PSPC-538977) and GCOM-C project (ER3GCF201) of the Japan Aerospace Exploration Agency (JAXA).
