1. Introduction
Clouds (mainly the cloud water and cloud ice floating in the air) play an essential role in Earth’s radiation budget, water cycle, and climate change (Ramanathan et al. 1989; Li et al. 1995; Rosenfeld and Ulbrich 2003; Voigt and Shaw 2015). Meanwhile, the hydrometeor contents in precipitating clouds (including cloud water, cloud ice, rain, snow, and graupel/hail) are directly related to the surface rainfall, and their microphysical processes are critical to the location and intensity of precipitation (Takahashi and Kawano 1998; Houze 2004; Wang and Georgakakos 2005; Tao and Moncrieff 2009). Precipitation processes can be classified into warm-rain and cold-rain processes based on the cloud microphysical mechanisms (Rauber et al. 2000; Gao et al. 2021). Cloud water and rainwater are vital in warm-rain processes, and supercooled water, cloud ice, snow, and graupel/hail are closely connected with cold-rain processes.
Reanalysis data and satellite retrievals have been widely used in investigating global cloud water and cloud ice in the nonprecipitating clouds. For example, Heng et al. (2014) pointed out that the long-term distributions and seasonal variations of cloud water and cloud ice, which mainly appear in the tropics and midlatitude westerlies, are related to the atmospheric circulation, and the cloud amount in the subtropical high pressure area is markedly lower. Eliasson et al. (2011) analyzed 3-yr CloudSat datasets and concluded that the ice water path (IWP; vertically integrated ice water per unit area) in nonprecipitating clouds is the highest in the tropical warm pool (around the Indonesian Archipelago), followed by the intertropical convergence zone (ITCZ) and midlatitude storm regions, which are mainly associated with the convection activities (Waliser et al. 2009). In China, Geng et al. (2018) found that there is a southwest–northeast-oriented belt from southwestern China to Japan with high liquid water path, and the IWP is large in the southeastern Tibetan Plateau and Jianghuai region (JH) because of the plateau topographic uplift and intense convection by analyzing Moderate Resolution Imaging Spectroradiometer (MODIS), European Centre for Medium-Range Weather Forecasts (ECMWF) interim reanalysis (ERA-Interim), and Climate Forecast System Reanalysis (CFSR) datasets. The height of maximum cloud ice content in nonprecipitating clouds is around 8 km, and that of cloud water is at 1–2 km and even reaches up to 10 km in summer (Yang and Wang 2012; Zhang et al. 2015). In addition, Chen et al. (2017) analyzed the macro- and microphysical properties of clouds over East China and the Tibetan Plateau through the cloud-resolving model simulations and indicated that latent heat of phase change is the most important term in the total heat and moisture budgets.
However, because of the detection limitations of remote sensing platforms, the studies on the long-term characteristics of precipitating water and ice (i.e., rain, snow, and graupel), regardless of their importance to the precipitation formation, are still lacking. In recent two decades, the satellite retrievals of vertical structure of precipitating hydrometeors by Tropical Rainfall Measuring Mission (TRMM), CloudSat, Global Precipitation Measurement (GPM), and others have become available (Abhik et al. 2013; Randel et al. 2020). Petersen et al. (2005) reported that the precipitating ice path over the tropical ocean is obviously less than that over the land, and the distributions of high-value areas are related to the topographic effects. Masunaga et al. (2002) demonstrated that the rainwater profile over land varies more dramatically than that over the ocean, and a significant decrease in rainwater and increase in precipitating ice occurs above 4 km in height using the TRMM datasets. Yin et al. (2013b) analyzed the profiles of cloud radar reflectivity in East Asia from 6-yr CloudSat dataset and indicated that the precipitating clouds are located generally below 8 km and the nonprecipitating clouds are at the height of 4–12 km. Though some satellite retrievals can provide information on precipitating water and ice particles, the shortages of low temporal resolution and inability to continually detect precipitation systems make them insufficient to meet the goals of this study, namely, the long-term characteristics of hydrometeors in precipitating clouds and their relationships with precipitation daily changes over eastern China.
Hydrometeors are the building blocks of the precipitation processes. The most frequent pathway to form rain is that precipitating ice forms above the freezing level and melts to produce rain using the satellite-based measurements and global climate model results (Heymsfield et al. 2020). Bhattacharya et al. (2014) found that cloud liquid water path and cloud ice path (including precipitation ice) have a monotonically increasing trend with the precipitation intensity over India and its surrounding oceans, but this relation becomes complex during the heavy precipitation based on analysis of 7-yr TRMM 2A12 datasets. Masunaga et al. (2002) presented a linear correlation between the near-surface precipitation water content and rainfall rate using the TRMM 2A12 and 2A25 datasets, but the scatterplots are more dispersed over land than over ocean due to the more complex precipitation mechanisms. Halder et al. (2012) indicated that the cloud water peaks ahead of precipitation due to the water vapor convergence in low levels, while the cloud ice evolution lags behind rainfall because of the subsequently triggered deep convection. Although some promising results have been achieved, the climatological relationship between precipitating hydrometeors and rainfall has not been thoroughly studied in the past. Most previous studies on the hydrometeors and precipitation in the East Asia region focused on the hydrometeor distributions (Luo et al. 2009; Kumar and Bhat 2017), stratiform and convective precipitation microphysics (Liu et al. 2018; Chernokulsky et al. 2019), and warm-rain and cold-rain processes (Noda et al. 2015; Gao et al. 2021). These results are based on either nonprecipitating clouds or specific precipitation events and are not statistically representative of the climatology. In addition, the characteristics of precipitation diurnal cycle are important in understanding the physical mechanisms of precipitation formation and local climate (Dai 2001; Xue et al. 2018). The afternoon rainfall peak in southern and central East China in the warm season is more related to the diurnal cycle of deep convective clouds via 1-yr-long simulations with different precipitation intensity (Chen et al. 2022). However, the relationship between the diurnal variations of hydrometeors and precipitation has rarely been analyzed.
The fifth major global reanalysis produced by the European Centre for Medium-Range Weather Forecasts (ERA5) data (Hersbach et al. 2020) provides quantitative information on four hydrometeor species (cloud water, rain, cloud ice, and snow) in a long period for the first time. Binder et al. (2020) showed that ERA5 can capture the location of the cloud system and vertical structure of ice-phase hydrometeor in the extratropics when compared with CloudSat data. It makes investigating hydrometeor vertical structures and cloud microphysical features in precipitating clouds possible. Since the release of ERA5 data, the hydrometeor dataset has been mainly used in studies of the distribution and evolution of hydrometeors (Dou et al. 2020; Yao et al. 2020; Deng et al. 2021) and model hydrometeor evaluations and as the initial conditions in precipitation simulations (Hwang et al. 2019).
Eastern China, the main area of heavy precipitation over China in summer (Li et al. 2013), is chosen as the domain of interest in this study. It is separated into three subregions based on different synoptic and precipitation features: South China (SC), the Jianghuai area, and North China (NC). The northward movement of rainband across these regions from May to August each year is associated with the East Asian monsoon and western Pacific anticyclone activity (Ding 1994). To the best of our knowledge, this is the first investigation of the climatological distributions of all four hydrometeor species and the long-term statistical relationships and diurnal variations between hydrometeors and precipitation over China. These results are expected to deepen our understanding of precipitation mechanisms over eastern China from the perspective of hydrometeor climatology. The paper is organized as follows. Section 2 introduces the study area and data. Section 3 evaluates the hydrometeors and precipitation in ERA5. Section 4 analyzes the 42-yr-averaged spatial distribution of hydrometeors, the contributions of water vapor and hydrometeors to precipitation, the vertical structures and occurrence frequencies of hydrometeors under different precipitation intensities, and their associations with diurnal variations of rainfall. Conclusions are given in section 5.
2. Study area and data
a. Study area
The study area is eastern China, located between 21° and 42°N and between 110° and 122°E. The summer precipitation varies dramatically in this region (Ding 1994), and the cloud microphysics and cloud radiative forcing are somewhat different from those in other regions under the influence of the East Asian monsoon (Wang et al. 2004; Luo et al. 2009; Yin et al. 2013a; Michibata et al. 2014). The study region is divided into three subregions following Wang and Ding (2008) as southern China (21°–28°N), the Jianghuai area (28°–35°N), and northern China (35°–42°N) on the basis of their unique precipitation features.
b. ERA5 reanalysis
Schematic of the cloud scheme in IFS Cycle 41r2.
Citation: Journal of Applied Meteorology and Climatology 62, 5; 10.1175/JAMC-D-22-0076.1
c. Hydrometeor and precipitation observations
To verify the ERA5 hydrometeor dataset, the GPM-retrieved level-3 hydrometeor product (3A GPROF GMI) is used. It is the monthly mean dataset of level 2 product, which is generated by relating the observed brightness temperatures to a large number of profiles in a preexisting database that is computed from cloud-resolving models through the Bayesian inversion scheme (Kummerow et al. 1996; Randel et al. 2020). The grid spacing of the level-3 product is 0.25° × 0.25°, and the vertical coordinate has 28 layers, with the top at 18 km. Because of the limitation of the microwave inversion algorithm, only the rainwater path (RWP) and snow water path (SWP) are used in this study. The gauge–satellite-merged hourly precipitation data [China Meteorological Administration Multisource Precipitation Analysis System (CMPAS)] are obtained from the National Meteorological Information Center of China, produced by more than 30 000 automatic rain gauge observations over China, and Climate Prediction Center morphing precipitation product (Shen et al. 2014) with a horizontal resolution of 0.1° × 0.1°. Hourly precipitation data from May to August in 2009–18 are collected to assess the precipitation in ERA5.
d. Precipitation budget equation
3. Verification
The GPM monthly mean hydrometeor dataset from May to August during 2014–20 is used to quantitatively compare with ERA5 (Fig. 2). The horizontal distributions of RWP and SWP in the two datasets are in general agreement, but the values of RWP in ERA5 are markedly smaller than those in GPM, likely due to the different size thresholds separating cloud droplets from raindrops. The maximum SWP in ERA5 (∼150 g m−2) is larger than that in GPM (∼95 g m−2), as the snow in ERA5 represents all of the precipitating solid-phase hydrometeors. In addition, the mean value, root mean square error (RMSE), and correlation coefficient between the two datasets are calculated (Table 1). The correlation coefficients of RWP and SWP are 0.76 and 0.64, respectively, with 95% confidence levels. These biases may be caused by the numerical model, inversion algorithm, and the differences in hydrometeor definition. It should be noted that there are also uncertainties in GPM hydrometeor retrievals (e.g., light rain, falling snow, separating ice-phase particles, and at coastlines) (GPM GPROF algorithm theoretical basis document; https://gpm.nasa.gov/resources/documents/gpm-gprof-algorithm-theoretical-basis-document-atbd), and they just serve as a reference here (Conrick and Mass 2019; Wu et al. 2021).
Monthly averaged hydrometeor distributions from May to August during 2014–20: (a),(b) RWP and (c),(d) SWP (g m−2) from (left) ERA5 and (right) GPM retrievals.
Citation: Journal of Applied Meteorology and Climatology 62, 5; 10.1175/JAMC-D-22-0076.1
Mean value, RMSE, and correlation coefficient of RWP and SWP between ERA5 and GPM retrievals.
Figure 3 shows the temporal evolutions of area-averaged monthly precipitation in the three regions from May to August during 2009–18. With the northward shift of the western North Pacific subtropical high, the rain belts steadily moved from southern China to northern China. The peak value of monthly precipitation shifts from June in SC to July in JH and then to August in NC. The trends of precipitation in ERA5 are almost the same as those in GPM retrievals, and their values are slightly larger than those in CMPAS. The monthly averaged precipitation is the smallest in NC, where the three datasets are the closest (Fig. 3c). In summary, the hydrometeor and precipitation in ERA5 are generally reasonable and can be used for further analyses.
Evolutions of area-averaged monthly precipitation (mm month−1) in (a) SC, (b) JH, and (c) NC from May to August during 2009–18 by CMPAS (green), GPM (blue), and ERA5 (purple).
Citation: Journal of Applied Meteorology and Climatology 62, 5; 10.1175/JAMC-D-22-0076.1
4. Results and discussion
a. Vertically integrated hydrometeor distributions
The 42-yr-averaged spatial distributions of four hydrometeors, supercooled liquid water, and precipitation rate over eastern China from May to August during 1979–2020 are shown in Fig. 4. The area (22°–31°N, 105°–112°E) of maximum cloud water path (CWP) is consistent with the result by Geng et al. (2018) (Fig. 4a), which is mainly attributed to the influence of plateau topography (not in our study area). The spatial distribution of CWP is similar to that of water vapor path, with the maximum value exceeding 180 g m−2 in SC, about 2 times that in NC. This is attributed to the low air temperature and insufficient water vapor in northern China. Small CWP exists in the southwestern corner of SC, where the IWP and SWP are high. This may be due to the frequent occurrence of low-level jets in this area (Du and Chen 2019), which are conducive to the strong updrafts and formation of rainwater and ice-phase particles. The high-value centers of RWP are located in southeastern JH and Taiwan Island (Fig. 4b). Note that the maximum RWP (>45 g m−2) in JH is a little higher than that in SC excluding Taiwan Island; however, the corresponding precipitation intensity (9–10.5 mm day−1) is smaller than that in SC (10.5–12 mm day−1) because of their different precipitation efficiencies (see further discussion in section 4c). For ice-phase hydrometeors, the IWP (>55 g m−2) is concentrated in southwestern SC, southeastern JH, and Taiwan Island; combined with the supercooled liquid water distribution, this results in the maximum SWP (>140 g m−2) in southeastern JH, followed by Taiwan Island and southwestern SC (Fig. 4d). The precipitation rate also gradually decreases from south to north, with three centers in southwestern SC, southeastern JH, and Taiwan Island (Fig. 4f). The distributions of RWP and SWP roughly correspond to the precipitation rate distribution with three high-value centers.
The 42-yr mean spatial distributions of (a) CWP, (b) RWP, (c) supercooled liquid water path (SCWP), (d) SWP, (e) IWP (g m−2), and (f) precipitation (mm day−1) averaged from May to August during 1979–2020. The black-outlined boxes in each panel denote the NC, JH, and SC regions from top to bottom, respectively.
Citation: Journal of Applied Meteorology and Climatology 62, 5; 10.1175/JAMC-D-22-0076.1
To describe the large-scale features associated with the cloud systems, Fig. 5 shows the 42-yr-averaged synoptic situation from May to August during 1979–2020. The specific humidity at 850-hPa layer in SC is the largest (13 g kg−1), and it is the smallest in NC (7 g kg−1) (Fig. 5a). The water vapor in SC reaches saturation and then forms cloud easier than other regions due to the strongest uplifting in the low level (Fig. 5c), resulting in the highest cloud water path (Fig. 4a). The wind rotates clockwise, with the height increasing from 850 to 500 hPa in SC and JH (Fig. 5b), indicating the existence of warm advection in both regions, which facilitates the upward motion. Strong updrafts above the 0°C level are found in both SC and JH (Fig. 5c). The maximum relative humidity at 8–12 km in JH is favorable to the generation of ice crystals, which contributes to the growth of snow below it and causes the maximum snow content in JH (Fig. 4d). The moisture environment and vertical velocity are the weakest in NC, and the hydrometeors are also the least therein. In addition, the negative vertical gradient of equivalent potential temperature below 850 hPa in SC is the largest, indicating the maximum unstable energy in SC. The dense isothermal zone, convergence center in the mid–lower layer, and divergence center at the top of atmosphere in JH (Fig. 5d) reflect the characteristics of the frequent mei-yu fronts over this region.
The 42-yr mean synoptic situation from May to August during 1979–2020: (a) 850-hPa specific humidity (shading; g kg−1), geopotential height (solid lines; dagpm), and horizontal wind (wind bars; m s−1); (b) 500-hPa temperature (shading; °C), geopotential height (solid lines; dagpm), and horizontal wind (wind bars; m s−1); (c) vertical cross section of relative humidity (shading; %), temperature (red lines; °C), and vertical velocity (arrows; m s−1) averaged between the longitudes of 110° and 122°E; and (d) vertical cross section of equivalent potential temperature (K) and divergence (10−6 s−1) averaged between the longitudes of 110° and 122°E.
Citation: Journal of Applied Meteorology and Climatology 62, 5; 10.1175/JAMC-D-22-0076.1
b. Precipitation budgets
The precipitation budget equation is utilized to analyze the moisture-related and hydrometeor-related processes to surface precipitation. Figure 6 displays the time series of area-averaged monthly precipitation rate, the sum of QWV and QH, and the ratio of hydrometeor-related processes to precipitation rate (QH/Ps) by ERA5 hourly data from May to August during the period of 1979–2020. The time evolutions of precipitation rate and the total of QWV and QH trace each other, indicating the general balance between the surface precipitation and the moisture- and hydrometeor-related activities, especially after 2000, owing to the assimilation of more high-quality observation data in the IFS system. It is evident that the moisture-related processes dominate the surface precipitation (|QH/PRE| < 5% most of the time), which is in agreement with the previous research; however, the hydrometeor-related factors cannot be ignored either [as the specific rainfall event study by Wang et al. (2007) and Huang et al. (2016)]. The hydrometeor-related processes, such as the cloud development by consuming water vapor before precipitation and cloud depletion during precipitation, have a certain influence on surface precipitation. Specifically, the contributions of hydrometeors in three regions are mostly between −5% and 5%, with the maximum value of up to 8% (e.g., in May 2013 in NC). The overall slightly negative impacts of hydrometeors on precipitation in JH and NC regions (mainly through the hydrometeor divergence) are likely due to the nearly synchronous reduction of ice-phase hydrometeors in the mid- and late stages of precipitation during their diurnal variations (as shown in Fig. 11, below).
Time series of area-averaged monthly precipitation rate (PRE) (mm h−1; black dashed line), the sum of QWV and QH (kg m−2 h−1; purple dashed line), and the ratio of QH to precipitation (%; red solid line) in (a) SC, (b) JH, and (c) NC from May to August during 1979–2020.
Citation: Journal of Applied Meteorology and Climatology 62, 5; 10.1175/JAMC-D-22-0076.1
As the local change rate (QWVL and QHL) is only 1% or less of the QWVA and QHA in a long-term mean, only the spatial distributions of 42-yr-averaged QWVA, Es, and QHA are illustrated in Fig. 7. It can be seen that the water vapor flux convergence weakens from SC to NC, while the spatial distribution of surface moisture flux in three regions is relatively uniform. The large-scale moisture-related processes play a very important role in the precipitation formation. Though the hydrometeor-related contribution to precipitation is generally less than 5%, there still exist some areas where the QHA can be near or even greater than 10%, particularly for the areas with light precipitation (<0.1 mm h−1). It is found that in the precipitation centers of southeastern JH and southwestern SC, the hydrometeor flux divergence is evident (Fig. 7b), indicating the cloud systems may feed the precipitation during the heavy rainfall period. In other words, the moisture-related factors determine the intensity of heavy precipitation, whereas the hydrometeor-related processes and moisture-related processes may jointly modulate the location of heavy precipitation, since the former can partly reflect the change characteristics of cloud systems. The hydrometeor flux convergence in SC is the strongest, along with the largest areal coverage (Fig. 7b), which facilitates the cloud growth and partially explains the highest CWP in SC (the pattern of CWP >140 g m−2 is similar to that of QHA >0.005 kg m−2 h−1). In addition, in the mountainous and coastal areas, the QHA is commonly positive and clouds tend to merge, while in the mid–lower Yangtze Plain (most of the JH region), QHA is mostly negative. That is, the topographic impacts on hydrometeor-related processes are more complex than those on the moisture-related processes, as the water vapor in three regions generally shows convergence.
The 42-yr mean spatial distributions of (a) QWVA (kg m−2 h−1), (b) QHA (kg m−2 h−1), (c) Es (mm h−1), and (d) Ps (mm h−1) averaged from May to August during 1979–2020.
Citation: Journal of Applied Meteorology and Climatology 62, 5; 10.1175/JAMC-D-22-0076.1
c. Hydrometeor profiles under different precipitation intensities
The long-term characteristics of cloud and hydrometeor vertical structures, including cloud-top height, cloud thickness, cloud water, and cloud ice profiles, in the Asian region have been widely explored (Luo et al. 2009; Yang and Wang 2012; Bhattacharya et al. 2014). However, few of them provided the hydrometeor profiles in precipitating clouds under the different rainfall intensities. In this study, the hourly precipitation data are classified into four groups: 0.1–1.5, 1.5–7, 7–15, and 15–40 mm h−1 (Figs. 8). Only the hydrometeors over land are considered.
The 42-yr mean hydrometeor profiles of (a1)–(a4) cloud water, (b1)–(b4) rain, (c1)–(c4) cloud ice, and (d1)–(d4) snow under different precipitation intensities [(top) 0.1–1.5, (top middle) 1.5–7, (bottom middle ) 7–15, and (bottom) 15–40 mm h−1] averaged from May to August during 1979–2020.
Citation: Journal of Applied Meteorology and Climatology 62, 5; 10.1175/JAMC-D-22-0076.1
The contents of four hydrometeor species increase steadily with the increase of precipitation intensity. Cloud water and rain mainly exist below the 0°C layer (550–600 hPa), and the height of maximum rainwater is higher than that of cloud water. Ice crystals and snow primarily appear above the 0°C layer, and ice crystals exist at a higher altitude than snow. This is consistent with the result of CloudSat observations (Zhang et al. 2015). A bimodal structure of cloud water profile occurs in SC and JH under various precipitation intensities. The peak of cloud water in low layer (>0°C) is mainly generated by the cloud droplet condensation. The cloud water content decreases with height until near the 0°C layer, as abundant cloud droplets are collected by raindrops. The peak of cloud water near above the 0°C layer should be formed by the water vapor condensation and vertical transportation of cloud water from the low layers, which are mostly related to the convective processes. This is similar to the result of Kubota et al. (2012) showing that cloud water presents a clear peak between 10° and 15°C layers in light precipitation and a peak near the freezing layer in heavy rainfall in the tropics. When rainfall rate is less than 7 mm h−1, the cloud water content in the low layer is slightly larger than that in the upper layer. As the precipitation rate increases, the supercooled cloud water rapidly increases in SC and JH, exceeding the maximum cloud water content in the low troposphere. This indicates that the convection in the two regions is stronger, which transports more cloud water upward as precipitation intensity increases. However, the maximum supercooled cloud water content in NC is always larger than the cloud water in low layer, owing to the lowest height of 0°C layer and the easiest formation of condensed supercooled cloud water. It means that more ice-phase particles can be involved in the weak precipitation in NC. The cloud water distribution at different altitudes has a considerable impact on the growth of rain, ice crystals, and snow particles.
Rainwater grows as a result of collision–coalescence of cloud droplets and the melting of snow below the 0°C layer. These two processes together cause the maximum rain content at about 700 hPa, and the rain content decreases when falling out of the clouds via evaporation. Note that the rainwater content in SC is the lowest under the moderate-intensity precipitation (<7 mm h−1), since the gridscale precipitation in SC (generated by the cloud scheme) is the weakest (the total precipitation is the sum of gridscale precipitation and convective precipitation from cumulus parameterization). However, the precipitation efficiency [as in the large-scale precipitation efficiency (LSPE2) in Sui et al. (2007)] in SC is the highest in this category (89.82% in SC, 86.69% in JH, and 82.72% in NC), resulting in the final similar total precipitation. The rain content in SC remarkably increases when the precipitation rate exceeds 7 mm h−1, and the melting of snow below the freezing layer supports the increase of rainwater.
Ice crystals are formed mainly through the homogeneous freezing of supercooled cloud droplets and heterogeneous nucleation of ice-nucleating particles. They are most abundant above 250 hPa (−40°C) in SC because of the lowest temperature and maximum relative humidity at the upper troposphere (Fig. 5c), which is optimal for homogenous freezing of supercooled droplets. The greatest cloud ice content exists in NC when the precipitation rate is less than 15 mm h−1, as the lowest height of 0°C level in NC is favorable for generation of supercooled water and then heterogeneous nucleation of ice crystals. However, when the precipitation rate further increases, the vertical velocity and humidity between 200 and 400 hPa in SC are larger than those in JH and NC, and the cloud ice content becomes the highest.
Snow above 400 hPa is primarily converted from ice crystals. The dramatic increase of snow between 400 and 600 hPa is due to the coexistence of supercooled water and ice crystals, which favors the Bergeron process and riming growth. When precipitation rate is less than 1.5 mm h−1, the snow content in NC is the greatest due to the largest vertical speed (Fig. 9a). The relative humidity in the midtroposphere in JH is larger than that in SC and NC, along with a vertical velocity center (Figs. 9b,c), resulting in the highest snow content in JH when the precipitation rate is between 1.5 and 15 mm h−1. During the heavy precipitation period (15–40 mm h−1), the centers of RH and updraft in the midtroposphere gradually move to SC, and the snow content in SC reaches its maximum (around 1.8 g kg−1), collocated with the maximum cloud ice content [Figs. 8c(4),d(4)]. It reveals that the ice-phase processes play a more important role in the heavy precipitation cases in SC. Wang et al. (2010) and Gao et al. (2021) drew a similar conclusion in simulating specific precipitation events in SC, where more ice-phase particles are involved, and the cold-rain process is critical for the heavy precipitation.
The 42-yr mean vertical cross section of relative humidity (%; shaded) and vertical velocity (Pa s−1; black line) from May to August during 1979–2020 under the different precipitation intensities: (a) 0.1–1.5, (b) 1.5–7, (c) 7–15, and (d) 15–40 mm h−1.
Citation: Journal of Applied Meteorology and Climatology 62, 5; 10.1175/JAMC-D-22-0076.1
d. Occurrence frequency between hydrometeors and precipitation
Cloud microphysical processes are indispensable to the formation and development of precipitation (Houghton 1951; Grabowski et al. 1999; Wang 2013). Relatively few research focuses on the direct relationship between hydrometeors and surface precipitation, which is important for understanding precipitation microphysical mechanisms. Tubul et al. (2017) proposed a quantified power–law relationship between IWP from MODIS and surface rainfall rate from TRMM in deep convective clouds over tropical and midlatitude regions. The occurrence frequencies between four hydrometeor paths and the precipitation intensities during 1979–2020 are shown in Fig. 10. The results indicate that the CWP increases nonlinearly with the precipitation rate, and when the precipitation rate exceeds 15 mm h−1, the CWP approaches a saturated value (no longer increases) in all three regions. In addition, the variation of CWP is the strongest in SC and the weakest in JH. For instance, when the precipitation rate is between 10– and 30 mm h−1, there exist some rainfall events with the CWP of 2.5–3 kg m−2 in SC, while it is mostly less than 2.5 kg m−2 in JH.
Occurrence frequency (%) between hydrometeor paths and precipitation intensities in the three regions [(left) SC, (center) JH, and (right) NC] from May to August during 1979–2020 for (a1)–(a3) CWP, (b1)–(b3) RWP, (c1)–(c3) IWP, and (d1)–(d3) SWP. The x axis is hydrometeor path (kg m−2), the y axis is precipitation rate (mm h−1), and shading represents the occurrence frequency of a different hydrometeor path under a given precipitation rate. The black lines are the fitting lines of maximal occurrence frequency.
Citation: Journal of Applied Meteorology and Climatology 62, 5; 10.1175/JAMC-D-22-0076.1
A consistent linear relationship between RWP and precipitation rate exists in all regions, with the slope of linear relationship being 5.68 h−1 in SC, 5.35 h−1 in JH, and 5.06 h−1 in NC, respectively. It indicates that SC can produce the largest precipitation with the same RWP; that is, the precipitation efficiency in SC is the maximum among the three regions. This is consistent with the calculated highest precipitation efficiency in SC in section 4c. The slope value here is greater than that in the tropics by 0.3–0.9 (Masunaga et al. 2002). The highest relative humidity in low level in SC (Fig. 5c) leading to the lowest loss of rainwater by evaporation among these regions also supports this conclusion. For ice-phase hydrometeors, the IWP increases with increasing precipitation rate only during weak precipitation conditions, and IWP becomes saturated (no further growth) at rainfall rate greater than 5 mm h−1. Note that when the rainfall rate exceeds 20 mm h−1, the distribution of IWP in SC is obviously wider than that in the other two regions [Figs. 10c(1)–c(3)]. No IWP greater than 1.5 kg m−2 exists in NC when rainfall rate is higher than 20 mm h−1, implying that ice crystals are limited in heavy precipitation due to the less supercooled water over there (Fig. 4c). In addition, the values of the maximum occurrence frequency of SWP lie between those of IWP and RWP, indicating its relationship with precipitation rate also lies between them. The SWP distribution in SC is considerably broader and shifted to the larger values in heavy precipitation than in the other two regions, corresponding to the earlier finding that more ice-phase particles are required in heavy precipitation in SC.
e. Diurnal variations of hydrometeors and precipitation
There have been many studies on the diurnal variation of precipitation in the East Asia region (Hirose and Nakamura 2005; Yuan et al. 2012), but little attention is paid to the diurnal variation characteristics of hydrometeors and how they affect the precipitation. Figure 11 presents the mean diurnal variations of hydrometeor paths and precipitation rate from May to August in 42 years (LST). There exist two precipitation peaks: an early-morning moderate peak (around 0600 LST) and an afternoon maximum peak (around 1500 LST) in all three regions. Cloud water, cloud ice, and snow almost always reach their maximum before the precipitation in the morning. We speculate that this is dominated by the weak large-scale stratiform precipitation (Chen et al. 2022), and cloud microphysical processes occur ahead of precipitation (Steiner et al. 1995). For the afternoon peak, the variations of cloud microphysical processes are relatively complex. The peak of cloud water is evidently ahead of rainfall by 2–3 h in all three regions. The cloud ice and snow lag behind precipitation by 2 h in SC and 1 h in JH, while they are basically in the same phase as rainfall in NC. These are in general agreement with the results by Halder et al. (2012) showing that cloud water precedes rainfall and cloud ice lags behind precipitation in the subtropical Indian monsoon region. It indicates that the ice-phase processes in afternoon precipitation are involved in varying degrees in the three regions. As the result of different temperature regimes, SC likely requires more time for ice-phase hydrometeors to convert to precipitation than JH and NC, since its conditions are warmest.
The 42-yr mean diurnal variations (LST) of precipitation and hydrometeor paths averaged over (a) SC, (b) JH, and (c) NC from May to August during 1979–2020. The leftmost y axis is PRE (mm h−1), the first axis on the right is CWP and SWP (kg m−2), and the rightmost axis is IWP, RWP, and SCWP (kg m−2).
Citation: Journal of Applied Meteorology and Climatology 62, 5; 10.1175/JAMC-D-22-0076.1
To further analyze the mechanisms for hydrometeor diurnal patterns, Fig. 12 shows the 42-yr mean diurnal variations of upward vertical velocity and water vapor flux convergence in all regions. The updraft in early morning in SC primarily exists in the low troposphere (below 600 hPa) (Fig. 12a), and the horizontal moisture transport at 850 hPa is weak, causing the small precipitation and smooth variations of four hydrometeor species (Fig. 11a). Chen et al. (2013) stated that the early-morning peak may be caused by the enhanced nighttime moisture transport under the influence of summer monsoon diurnal cycle. For the afternoon precipitation, the strongest water vapor convergence occurs almost 7 h earlier than the peak of precipitation, indicating that there is a significant humidification of atmosphere during the preprecipitation period. The time of maximum cloud water corresponds well to the beginning of the maximum updraft (around 1200 LT), and the strong vertical velocity extends upward above 350 hPa after 1400 LT (Fig. 12a). Under the influence of dynamic field, water vapor condenses and produces a large amount of cloud water. Meanwhile, the phase-change processes in cloud formation can also enhance the upward motion. The deep convective structure in SC is mainly due to the interaction between the dynamics and thermodynamics in the moist environment. The supercooled cloud water through vapor condensation and the vertical transportation of cloud water from low level (below 0°C layer) promote the growth of ice-phase particles, causing the peak of ice crystals and snow 3–4 h after the maximum cloud water. After 1500 LT, the updraft continues to develop, coupled with the water vapor convergence in the mid- and upper layers, resulting in a continuous increase of ice-phase particles and a peak ∼2 h behind the precipitation peak.
The 42-yr mean diurnal variation (LST) of (a)–(c) upward vertical velocity (Pa s−1; negative values represent updraft) and (d)–(f) water vapor flux convergence (kg m−3 h−1; positive values represent convergence) averaged over (left) SC, (center) JH, and (right) NC from May to August during 1979–2020.
Citation: Journal of Applied Meteorology and Climatology 62, 5; 10.1175/JAMC-D-22-0076.1
Like the early-morning precipitation in SC, the hydrometeors in JH are also ahead of precipitation. However, the low-level vertical velocity is about half of that in SC (Fig. 12b), causing slightly lower precipitation intensity and CWP. The phenomenon of more rainwater but lower precipitation rate in JH than in SC confirms that a lower precipitation efficiency occurs in JH during early-morning weak precipitation, which echoes previous results. The continuous upward motion between 400 and 600 hPa and the largest water vapor convergence between 500 and 800 hPa before 1500 LT in JH (Figs. 12b,e) cause the most cloud ice and snow among the three regions. In afternoon precipitation, the cloud water is still ahead of precipitation. After 1500 LT, the upward velocity in the mid–upper troposphere increases for a short time, and the ice-phase particles lag behind precipitation by about 1 h. This upper-level updraft structure (the elevated convection) should be related to the frequently occluded frontal system in JH during the summer. After the rapidly weakened vertical velocity at 1800 LT, the ice-phase particles decrease significantly.
The lowest CWP (one-third of SC) and SWP (two-thirds of SC and JH) and the smallest precipitation occur in NC as a result of the weakest vertical velocity and water vapor advection among the three regions. No obvious updraft exists above the midtroposphere (Fig. 12c), resulting in only cloud water and cloud ice being ahead of precipitation, and rain and snow are roughly in the same phase as precipitation in the morning. For the afternoon precipitation, the remarkable differences between NC and the other two regions are that the large vertical velocity is below 600 hPa and no strong updraft occurs in the mid–upper troposphere. Thus, the lagging of ice-phase particles behind rainfall does not appear when precipitation reaches its maximum. This near-surface convection in NC is more attributed to its topography and the driest environment.
5. Summary and conclusions
Eastern China (21°–42°N, 110°–122°E) is significantly affected by the East Asian monsoon and is the main area of summer precipitation in China. Many studies on the characteristics of cloud water and cloud ice in the nonprecipitating clouds in this area have been conducted. However, few studies focus on the long-term characteristics of various hydrometeor species (e.g., cloud water, cloud ice, rain, and snow) in precipitating clouds and the relationships between hydrometeors and precipitation due to the limitation of observation technology and data. The latest ERA5 reanalysis dataset offers a new opportunity to investigate these important questions, and the data from May to August during 1979–2020 are used in this study. ERA5 hydrometeor and precipitation data are validated against the GPM-retrieved hydrometeors and CMPAS hourly precipitation product. The results are generally reasonable over eastern China.
For the 42-yr mean spatial distributions of hydrometeors and precipitation, CWP decreases gradually from south to north due to the large-scale atmospheric circulation and water vapor distribution. The maximum CWP in SC exceeds 180 g m−2, nearly 2 times that in NC. The centers of RWP are in southwestern SC, southeastern JH, and Taiwan Island, respectively, whose distributions are comparable to that of surface precipitation. The highest SWP occurs in JH because of its largest amount of supercooled water as well as the humidity and upward motion at 8–12-km height. The contributions of hydrometeor-related processes to precipitation in three regions are generally between −5% and 5% and can be up to 10% in some areas, especially in the area with light precipitation. The moisture-related processes affect precipitation intensity more strongly, while the hydrometeor-related processes along with the moisture convergence center may jointly modulate the location of heavy precipitation.
The 42-yr-averaged profiles of hydrometeors demonstrate a bimodal structure of cloud water, indicating the different roles of warm- and cold-rain processes. Rainwater in SC is the least among all regions when the precipitation rate is less than 7 mm h−1, but the precipitation efficiency is the highest (89.82%), followed by 86.69% in JH and 82.72% in NC. The contents of ice-phase particles in NC are the greatest in weak precipitation (<1.5 mm h−1). When the precipitation rate gets stronger, the snow content gradually becomes the highest in JH due to the suitable water vapor and vertical velocity in the midtroposphere. During strong precipitation periods (>15mm h−1), the largest contents of ice-phase hydrometeors occur in SC, implying that the cold-rain process is crucial to heavy precipitation.
The occurrence frequency of hydrometeor paths with precipitation rate shows that CWP increases with precipitation intensity when the rainfall rate is less than 15 mm h−1, and cloud water approaches a saturated value thereafter (no longer increases). A clear linear relationship exists between RWP and precipitation rate in all regions, with the highest slope (5.68 h−1) in SC, followed by JH (5.53 h−1), and the lowest in NC (5.06 h−1), indicating that the efficiency in converting rainwater to precipitation is the highest in SC. The diurnal variations of precipitation and hydrometeor paths reveal that cloud water, cloud ice, and snow commonly reach their peaks before precipitation peaks in the early-morning precipitation. This rainfall is likely dominated by large-scale stratiform cloud precipitation, and the cloud microphysical processes occur before precipitation. For the afternoon precipitation, the peak of CWP is 2–3 h ahead of precipitation in all regions. The ice-phase particles lag behind precipitation by about 2 and 1 h in SC and JH, respectively, but are the same phase in NC. This is attributed to the different temperature regimes as well as the structures of upward velocity and water vapor convergence in the mid–upper troposphere.
We acknowledge that only the ERA5 data are used in this study in analyzing the climatological characteristics of hydrometeors in precipitating clouds. More satellite and radar observation data as well as the long-term trends of hydrometeor variables should be considered for the joint analysis in the future.
Acknowledgments.
This work was supported by the National Natural Science Foundation of China (42275082, 41775131, U2142209) and the S&T Development Fund of CAMS (2019KJ020, 2021KJ019). The National Center for Atmospheric Research is sponsored by the National Science Foundation.
Data availability statement.
The data were obtained online as follows: the ERA5 reanalysis data (https://cds.climate.copernicus.eu/#!/search?text=ERA5&type=dataset), the hourly precipitation data (http://data.cma.cn/en), and the GPM satellite data (https://disc.gsfc.nasa.gov/datasets/GPM_3GPROFGPMGMI_07/summary?keywords=GPM_3GPROFGPMGMI).
REFERENCES
Abhik, S., M. Halder, P. Mukhopadhyay, X. Jiang, and B. N. Goswami, 2013: A possible new mechanism for northward propagation of boreal summer intraseasonal oscillations based on TRMM and MERRA reanalysis. Climate Dyn., 40, 1611–1624, https://doi.org/10.1007/s00382-012-1425-x.
Bhattacharya, A., A. Chakraborty, and V. Venugopal, 2014: Variability of cloud liquid water and ice over South Asia from TMI estimates. Climate Dyn., 42, 2411–2421, https://doi.org/10.1007/s00382-013-1978-3.
Binder, H., M. Boettcher, H. Joos, M. Sprenger, and H. Wernli, 2020: Vertical cloud structure of warm conveyor belts–A comparison and evaluation of ERA5 reanalysis, CloudSat and CALIPSO data. Wea. Climate Dyn., 1, 577–595, https://doi.org/10.5194/wcd-1-577-2020.
Chen, G. X., W. M. Sha, M. Sawada, and T. Iwasaki, 2013: Influence of summer monsoon diurnal cycle on moisture transport and precipitation over eastern China. J. Geophys. Res. Atmos., 118, 3163–3177, https://doi.org/10.1002/jgrd.50337.
Chen, J., X. Wu, Y. Yin, Q. Huang, and H. Xiao, 2017: Characteristics of cloud systems over the Tibetan Plateau and East China during boreal summer. J. Climate, 30, 3117–3137, https://doi.org/10.1175/JCLI-D-16-0169.1.
Chen, J., X. Wu, C. Lu, and Y. Yin, 2022: Seasonal and diurnal variations of cloud systems over the eastern Tibetan Plateau and East China: A cloud-resolving model study. Adv. Atmos. Sci., 39, 1034–1049, https://doi.org/10.1007/s00376-021-0391-9.
Chernokulsky, A., F. Kozlov, O. Zolina, O. Bulygina, I. I. Mokhov, and V. A. Semenov, 2019: Observed changes in convective and stratiform precipitation in northern Eurasia over the last five decades. Environ. Res. Lett., 14, 045001, https://doi.org/10.1088/1748-9326/aafb82.
Conrick, R., and C. F. Mass, 2019: Evaluating simulated microphysics during OLYMPEX using GPM satellite observations. J. Atmos. Sci., 76, 1093–1105, https://doi.org/10.1175/JAS-D-18-0271.1.
Dai, A., 2001: Global precipitation and thunderstorm frequencies. Part II: Diurnal variations. J. Climate, 14, 1112–1128, https://doi.org/10.1175/1520-0442(2001)014<1112:GPATFP>2.0.CO;2.
Deng, L., Y. Zhao, W. Gao, L. Xue, and Y. Duan, 2021: Intensity and microphysical properties surrounding the rapid intensification in landfalling super typhoons over China during the summer and autumn seasons. Int. J. Climatol., 41, 6366–6381, https://doi.org/10.1002/joc.7200.
Ding, Y., 1994: The summer monsoon in East Asia. Monsoons over China, Y. Ding, Ed., Atmospheric Sciences Library, Vol. 16, Springer, 1–90.
Dou, T., C. Xiao, Y. Huang, H. Yue, and W. Han, 2020: Estimation of the atmospheric ice content mass, spatial distribution, and long-term changes based on the ERA5 reanalysis. Geophys. Res. Lett., 47, e2020GL088186, https://doi.org/10.1029/2020GL088186.
Du, Y., and G. Chen, 2019: Climatology of low-level jets and their impact on rainfall over southern China during early-summer rainy season. J. Climate, 32, 8813–8833, https://doi.org/10.1175/JCLI-D-19-0306.1.
Eliasson, S., S. A. Buehler, M. Milz, P. Eriksson, and V. O. John, 2011: Assessing observed and modelled spatial distributions of ice water path using satellite data. Atmos. Chem. Phys., 11, 375–391, https://doi.org/10.5194/acp-11-375-2011.
Gao, S., X. Cui, Y. Zhou, and X. Li, 2005: Surface rainfall processes as simulated in a cloud-resolving model. J. Geophys. Res., 110, D10202, https://doi.org/10.1029/2004JD005467.
Gao, W., L. Xue, L. Liu, C. Lu, Y. Yun, and W. Zhou, 2021: A study of the fraction of warm rain in a pre-summer rainfall event over South China. Atmos. Res., 262, 105792, https://doi.org/10.1016/j.atmosres.2021.105792.
Geng, R., Y. Wang, Y. Fu, R. Li, and G. Liu, 2018: Comparison of climatological characteristics of several hydrometeors in China and the surrounding region (in Chinese). Acta Meteor. Sin., 76, 134–147, https://doi.org/10.11676/qxxb2017.075.
Grabowski, W. W., X. Wu, and M. W. Moncrieff, 1999: Cloud-resolving modeling of tropical cloud systems during phase III of GATE. Part III: Effects of cloud microphysics. J. Atmos. Sci., 56, 2384–2402, https://doi.org/10.1175/1520-0469(1999)056<2384:CRMOTC>2.0.CO;2.
Halder, M., P. Mukhopadhyay, and S. Halder, 2012: Study of the microphysical properties associated with the monsoon intraseasonal oscillation as seen from the TRMM observations. Ann. Geophys., 30, 897–910, https://doi.org/10.5194/angeo-30-897-2012.
Heng, Z., Y. Fu, G. Liu, R. J. Zhou, Y. Wang, R. Yuan, J. Cuo, and D. Xue, 2014: A study of the distribution and variability of cloud water using ISCCP, SSM/I cloud product, and reanalysis datasets. J. Climate, 27, 3114–3128, https://doi.org/10.1175/JCLI-D-13-00031.1.
Hersbach, H., and Coauthors, 2020: The ERA5 global reanalysis. Quart. J. Roy. Meteor. Soc., 146, 1999–2049, https://doi.org/10.1002/qj.3803.
Heymsfield, A. J., C. Schmitt, C.-C.-J. Chen, A. Bansemer, A. Gettelman, P. R. Field, and C. Liu, 2020: Contributions of the liquid and ice phases to global surface precipitation: Observations and global climate modeling. J. Atmos. Sci., 77, 2629–2648, https://doi.org/10.1175/JAS-D-19-0352.1.
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.
Houghton, H. G., 1951: On the physics of clouds and precipitation. Compendium of Meteorology, T. F. Malone, Ed., Amer. Meteor. Soc., 165–181.
Houze, R. A., 2004: Mesoscale convective systems. Rev. Geophys., 42, RG4003, https://doi.org/10.1029/2004RG000150.
Huang, Y., X. Cui, and X. Li, 2016: A three-dimensional WRF-based precipitation equation and its application in the analysis of roles of surface evaporation in a torrential rainfall event. Atmos. Res., 169, 54–64, https://doi.org/10.1016/j.atmosres.2015.09.026.
Hwang, S.-O., J. Par, and H. M. Kim, 2019: Effect of hydrometeor species on very-short-range simulations of precipitation using ERA5. Atmos. Res., 218, 245–256, https://doi.org/10.1016/j.atmosres.2018.12.008.
Kubota, T., M. Satoh, T. Nasuno, S. Seto, T. Iguchi, and R. Oki, 2012: Development of cloud liquid water database using global cloud-system resolving model for GPM/DPR algorithm. 2012 IEEE Int. Geoscience and Remote Sensing Symp., Munich, Germany, Institute of Electrical and Electronics Engineers, 350–353, https://doi.org/10.1109/IGARSS.2012.6351566.
Kumar, S., and G. S. Bhat, 2017: Vertical structure of orographic precipitating clouds observed over South Asia during summer monsoon season. J. Earth Syst. Sci., 126, 114, https://doi.org/10.1007/s12040-017-0897-9.
Kummerow, C., W. S. Olson, and L. Giglio, 1996: A simplified scheme for obtaining precipitation and vertical hydrometeor profiles from passive microwave sensors. IEEE Trans. Geosci. Remote Sens., 34, 1213–1232, https://doi.org/10.1109/36.536538.
Li, J., R. Yu, and W. Sun, 2013: Duration and seasonality of hourly extreme rainfall in the central eastern China. Acta Meteor. Sin., 27, 799–807, https://doi.org/10.1007/s13351-013-0604-y.
Li, Z., H. W. Barker, and L. Moreau, 1995: The variable effect of clouds on atmospheric absorption of solar radiation. Nature, 376, 486–490, https://doi.org/10.1038/376486a0.
Liu, D., B. Yang, Y. Zhang, Y. Qian, A. Huang, Y. Zhou, and L. Zhang, 2018: Combined impacts of convection and microphysics parameterizations on the simulations of precipitation and cloud properties over Asia. Atmos. Res., 212, 172–185, https://doi.org/10.1016/j.atmosres.2018.05.017.
Luo, Y., R. Zhang, and H. Wang, 2009: Comparing occurrences and vertical structures of hydrometeors between eastern China and the Indian monsoon region using CloudSat/CALIPSO data. J. Climate, 22, 1052–1064, https://doi.org/10.1175/2008JCLI2606.1.
Masunaga, H., T. Iguchi, R. Oki, and M. Kachi, 2002: Comparison of rainfall products derived from TRMM microwave imager and Precipitation Radar. J. Appl. Meteor. Climatol., 41, 849–862, https://doi.org/10.1175/1520-0450(2002)041<0849:CORPDF>2.0.CO;2.
Michibata, T., K. Kawamoto, and T. Takemura, 2014: The effects of aerosols on water cloud microphysics and macrophysics based on satellite-retrieved data over East Asia and the North Pacific. Atmos. Chem. Phys., 14, 11 935–11 948, https://doi.org/10.5194/acp-14-11935-2014.
Noda, A. T., M. Satoh, Y. Yamada, C. Kodama, T. Miyakawa, and T. Seiki, 2015: Cold and warm rain simulated using a global nonhydrostatic model without cumulus parameterization, and their responses to global warming. J. Meteor. Soc. Japan, 93, 181–197, https://doi.org/10.2151/jmsj.2015-010.
Petersen, W. A., H. J. Christian, and S. A. Rutledge, 2005: TRMM observations of the global relationship between ice water content and lightning. Geophys. Res. Lett., 32, L14819, https://doi.org/10.1029/2005GL023236.
Ramanathan, V., R. D. Cess, E. F. Harrison, P. Minnis, B. R. Barkstrom, E. Ahmad, and D. Hartmann, 1989: Cloud-radiative forcing and climate: Results from the Earth Radiation Budget Experiment. Science, 243, 57–63, https://doi.org/10.1126/science.243.4887.57.
Randel, D. L., C. D. Kummerow, and S. Ringerud, 2020: The Goddard profiling (GPROF) precipitation retrieval algorithm. Satellite Precipitation Measurement, V. Levizzani, Ed., Advances in Global Change Research, Vol. 67, Springer, 141–152.
Rauber, R. M., L. S. Olthoff, M. K. Ramamurthy, and K. E. Kunkel, 2000: The relative importance of warm rain and melting processes in freezing precipitation events. J. Appl. Meteor. Climatol., 39, 1185–1195, https://doi.org/10.1175/1520-0450(2000)039<1185:TRIOWR>2.0.CO;2.
Rosenfeld, D., and C. W. Ulbrich, 2003: Cloud microphysical properties, processes, and rainfall estimation opportunities. Radar and Atmospheric Science: A Collection of Essays in Honor of David Atlas, Meteor. Monogr., No. 30, Amer. Meteor. Soc., 237–258, https://doi.org/10.1175/0065-9401(2003)030<0237:CMPPAR>2.0.CO;2.
Shen, Y., P. Zhao, Y. Pan, and J. Yu, 2014: A high spatiotemporal gauge–satellite merged precipitation analysis over China. J. Geophys. Res. Atmos., 119, 3063–3075, https://doi.org/10.1002/2013JD020686.
Steiner, M., R. A. Houze, and S. E. Yuter, 1995: Climatological characterization of three-dimensional storm structure from operational radar and rain gauge data. J. Appl. Meteor. Climatol., 34, 1978–2007, https://doi.org/10.1175/1520-0450(1995)034<1978:CCOTDS>2.0.CO;2.
Sui, C.-H., X. Li, and M.-J. Yang, 2007: On the definition of precipitation efficiency. J. Atmos. Sci., 64, 4506–4513, https://doi.org/10.1175/2007JAS2332.1.
Takahashi, T., and T. Kawano, 1998: Numerical sensitivity study of rainband precipitation and evolution. J. Atmos. Sci., 55, 57–87, https://doi.org/10.1175/1520-0469(1998)055<0057:NSSORP>2.0.CO;2.
Tao, W.-K., and M. W. Moncrieff, 2009: Multiscale cloud system modeling. Rev. Geophys., 47, RG4002, https://doi.org/10.1029/2008RG000276.
Tubul, Y., I. Koren, O. Altaratz, and R. H. Heiblum, 2017: On the link between precipitation and the ice water path over tropical and mid-latitude regimes as derived from satellite observations. Atmos. Meas. Tech. Discuss., https://doi.org/10.5194/amt-2017-121.
Voigt, A., and T. A. Shaw, 2015: Circulation response to warming shaped by radiative changes of clouds and water vapour. Nat. Geosci., 8, 102–106, https://doi.org/10.1038/ngeo2345.
Waliser, D. E., and Coauthors, 2009: Cloud ice: A climate model challenge with signs and expectations of progress. J. Geophys. Res., 114, D00A21, https://doi.org/10.1029/2008JD010015.
Wang, J., and K. P. Georgakakos, 2005: Effects of cold microphysical processes on the surface precipitation variability of nonsquall tropical oceanic convection. J. Geophys. Res., 110, D22203, https://doi.org/10.1029/2005JD005787.
Wang, J. J., X. Li, and L. D. Carey, 2007: Evolution, structure, cloud microphysical, and surface rainfall processes of monsoon convection during the South China Sea Monsoon Experiment. J. Atmos. Sci., 64, 360–380, https://doi.org/10.1175/JAS3852.1.
Wang, P. K., 2013: Physics and Dynamics of Clouds and Precipitation. Cambridge University Press, 452 pp.
Wang, W.-C., W. Gong, W.-S. Kau, C.-T. Chen, H.-H. Hsu, and C.-H. Tu, 2004: Characteristics of cloud radiation forcing over East China. J. Climate, 17, 845–853, https://doi.org/10.1175/1520-0442(2004)017<0845:COCRFO>2.0.CO;2.
Wang, Y., X. Shen, and X. Li, 2010: Microphysical and radiative effects of ice clouds on responses of rainfall to the large-scale forcing during pre-summer heavy rainfall over southern China. Atmos. Res., 97, 35–46, https://doi.org/10.1016/j.atmosres.2010.03.005.
Wang, Z., and Y. Ding, 2008: Climatic characteristics of rainy seasons in China (in Chinese). Chin. J. Atmos. Sci., 32, 1–13.
Wu, S.-N., B. J. Soden, Y. Miyamoto, D. S. Nolan, and S. A. Buehler, 2021: Using satellite observations to evaluate the relationships between ice condensate, latent heat release, and tropical cyclone intensification in a mesoscale model. Mon. Wea. Rev., 149, 113–129, https://doi.org/10.1175/MWR-D-19-0348.1.
Xue, M., X. Luo, K. Zhu, Z. Sun, and J. Fei, 2018: The controlling role of boundary layer inertial oscillations in meiyu frontal precipitation and its diurnal cycles over China. J. Geophys. Res. Atmos., 123, 5090–5115, https://doi.org/10.1029/2018JD028368.
Yang, D., and P. Wang, 2012: Characteristics of vertical distributions of cloud water contents over China during summer (in Chinese). Chin. J. Atmos. Sci., 36, 89–101, https://doi.org/10.3878/j.issn.1006-9895.2012.01.08.
Yao, B., S. Teng, R. Lai, X. Xu, Y. Yin, C. Shi, and C. Liu, 2020: Can atmospheric reanalyses (CRA and ERA5) represent cloud spatiotemporal characteristics? Atmos. Res., 244, 105091, https://doi.org/10.1016/j.atmosres.2020.105091.
Yin, J., D. Wang, and G. Zhai, 2013a: A comparative study of cloud-precipitation microphysical properties between East Asia and other regions. J. Meteor. Soc. Japan, 91, 507–526, https://doi.org/10.2151/jmsj.2013-406.
Yin, J., D. Wang, G. Zhai, and Z. Wang, 2013b: Observational characteristics of cloud vertical profiles over the continent of East Asia from the CloudSat data. Acta Meteor. Sin., 27, 26–39, https://doi.org/10.1007/s13351-013-0104-0.
Yuan, W., R. Yu, M. Zhang, W. Lin, H. Chen, and J. Li, 2012: Regimes of diurnal variation of summer rainfall over subtropical East Asia. J. Climate, 25, 3307–3320, https://doi.org/10.1175/JCLI-D-11-00288.1.
Zhang, H., B. Yang, J. Peng, Z. Wang, and X. Jing, 2015: The characteristics of cloud microphysical properties in East Asia with the CloudSat dataset (in Chinese). Chin. J. Atmos. Sci., 39, 235–248, https://doi.org/10.3878/j.issn.1006-9895.1408.13313.