2. Key Laboratory for Cloud Physics of China Meteorological Administration, Chinese Academy of Meteorological Sciences, China Meteorological Administration, Beijing 100081;
3. Collaborative Innovation Center on Forecast and Evaluation of Meteorological Disasters, Nanjing University of Information Science & Technology, Nanjing 210044
The Tibetan Plateau (TP) is a huge uplifted heat source in summer. Its special dynamic and thermal effects play an important role in the formation of regional circulation and the outbreak and maintenance of monsoon (Xu and Chen, 2006; Xu et al., 2014). Based on analysis of GMS (Geostationary Meteorological Satellite) data, Fujinami and Yasunari (2001) revealed that convective activities over the TP often rapidly intensify after the noon time and reach their maxima in early evening. Convective clouds in the central TP are columnar clouds characterized by small horizontal scale and large vertical thickness. Under certain conditions, these clouds can break through the “warm cover” inversion layer and form “popcorn-like” clouds (Xu et al., 2002).
Ueno et al. (2001) analyzed radar and precipitation data of GAME (GEWEX Asian Monsoon Experiment)-Tibet (1998) and found that weak precipitation events frequently occur over the TP during the monsoon period (mid June to early September). Precipitation during the daytime [1000–1300 local standard time (LST), equivalent to UTC + 6 h] is weaker than that during the nighttime (2000–0200 LST) in Naqu basin. Daytime precipitation is often related to scattered, small-scale convective monomers, while nighttime precipitation is associated with large-scale stratiform clouds. Based on analysis of TRMM (Tropical Rainfall Measuring Mission) satellite data, Fu et al. (2007, 2008) found that summertime precipitation over the TP demonstrates significant diurnal variation with the peak and bottom values appearing at 1600 and 0500 LST, respectively. They also revealed that TP precipitation generally arises scatteredly associated with small horizontal scale and a vaulted vertical section. Li et al. (2012) analyzed a strong convective event over the TP based on TRMM satellite data. They found that convective clouds are somehow compressed in the vertical; ice crystals in the clouds are concentrated in the layer between 6 and 18 km, while the precipitable ice particles, water droplets, and cloud droplets are mostly distributed below 8 km; and precipitable ice crystals account for the largest part of the clouds. Statistical analyses of satellite data also revealed frequent occurrence of cumulus clouds and precipitation in summer over the TP, whereas the precipitation overall is weak (Pan and Fu, 2015; Li and Zhang, 2016).
Using radar data and precipitation observations collected in the Third Tibetan Plateau Atmospheric Science Experiment (TIPEX-III), Liu et al. (2015) demonstrated that the total clouds, the top, amount and thickness of high clouds all display obvious diurnal variation features. Chang and Guo (2016) indicated that the average top of convective clouds is about 11.5 km (above the sea level), the maximum cloud top can be above 19 km, and the average height of cloud bottom is 6.88 km.
With scarce observations in the TP, various numerical models have been employed to study precipitation processes and characteristics over the TP. Liu et al. (1999) simulated convective clouds over the TP using a three-dimensional cloud model. Their results indicate that ice-phase microphysics plays a critical role in clouds and precipitation processes in the TP. Sato et al. (2008) investigated the impact of horizontal resolution of the WRF model on the simulation of diurnal variation of precipitation in the TP. They found that low resolution (> 14 km) would lead to lagged formation and mature of convective clouds. As a result, erroneous precipitation increases sharply. In contrast, high-resolution model (< 7 km) can realistically simulate features of precipitation over the TP.Maussion et al. (2011) used the WRF model to simulate precipitation events that occurred over the TP in October 2008. Their results indicate that the spatial distribution of simulated precipitation is similar to TRMM observations, but the location of maximum precipitation deviates from the observation. The WRF simulation agrees better with observations over areas of weak precipitation. Xu et al. (2012) found that cloud-resolving models can reproduce the diurnal variation of precipitation over the TP; however, the simulated precipitation is twice that of TRMM observations.
Most of the previous studies investigated macro-characteristics of clouds and precipitation over the TP based on weather station observations and radar and satellite remote sensing data. However, few studies focused on cloud microphysical properties over the TP, which is critical to understanding the precipitation formation mechanism and improving the numerical weather forecasting ability in this region. Although there are some cloud-resolving modeling studies on clouds and precipitation processes over the TP (Liu et al., 1999; Sato et al., 2008; Maussion et al., 2011; Xu et al., 2012; Gao et al., 2016, 2018), their conclusions are not fully consistent (e.g., Liu et al., 1999; Gao et al., 2016) regarding the role of cold and warm cloud microphysical processes in precipitation formation. In addition, most of the numerical experiments are conducted for a single case (Gao et al., 2018). In the present study, six cloud and precipitation processes that occurred during the period of 3–25 July (3–4, 9–10, 13–14, 20–21, 21–22, and 24–25 July) 2014 are investigated based on the data obtained in the TIPEX-III and numerical simulations. The purpose of the present study is to reveal the summer cloud microphysical properties and precipitation formation mechanism over the TP.2 Model and observation data
The WRF model version 3.8 (WRFV3.8) is implemented to simulate the six cloud and precipitation processes. Initial conditions are extracted from the NCEP/FNL analysis data on 1° × 1° global grids. The integration covers 48 h for each case, i.e., 0000 LST 3–0000 LST 5 July, 0000 LST 9–0000 LST 11 July, 0000 LST 13–0000 LST 15 July, 0000 LST 20–0000 LST 22 July, 0000 LST 21–0000 LST 23 July, and 0000 LST 24–0000 LST 26 July 2014. A triple-nested domain is designed for all the simulations (Fig. 1). There are 30 levels unevenly distributed in the vertical; grid numbers for the triple-nested domains are 181 × 181, 301 × 301, and 361 × 361 with grid intervals of 9, 3, and 1 km, respectively. Important physical schemes include the Dudhia scheme for shortwave radiation (Dudhia, 1989), the Rapid Radiative Transfer Model (RRTM) for longwave radiation (Mlawer et al., 1997), the Grell–Devenyi scheme for cumulus parameterization (Grell and Devenyi, 2002, only used in the outermost domain), the Mellor–Yamada–Janjic (Eta) scheme for planetary boundary layer (Janjic, 2002), the Noah land surface model for soil physics (Chen and Dudhia, 2001), and the Lin microphysics scheme (Lin et al., 1983). The Lin scheme includes prognostics of cloud water, cloud ice, rainwater, snow, and graupel mixing ratios. In addition, due to the low resolution of the FNL analysis, lake surface temperature over the TP is derived from sea surface temperature over the Bay of Bengal in the original WRF model. However, this method does not consider the high elevation of the TP, and large biases are introduced into the initial condition of lake surface temperature (Li et al., 2009; Maussion et al., 2011). As a result, precipitation is severely overestimated near Nam Co in the WRF simulations. In the present study, lake surface temperature derived from satellite remote sensing retrievals [MODIS (Moderate Resolution Imaging Spectroradiometer)-MOD11C1] is used to replace the temperature produced by WRF. After the replacement, surface temperature over Nam Co is about 6–8°C, which is consistent with the summertime climatic mean value shown in the study of Haginoya et al. (2009). With corrected lake surface temperature, the abnormally large precipitation over the lake area in the WRF simulations no longer exists.
The airborne observations of liquid water content, radar observations, and merged hourly precipitation data based on weather station observations and Climate Prediction Center morphing technique (CMORPH) are compared with model simulations. The merged hourly precipitation data cover the area (15°S–60°N, 70°–140°E) with a spatial resolution of 0.1° × 0.1° (Shen et al., 2014). The C-band weather radar was deployed at Naqu Bureau of Meteorology (31.48°N, 92.07°E), and the C-band continuous wave radar was deployed at Zhongxin Hotel (31.29°N, 92.03°E). The continuous wave radar implements the solid-state transmitter system to continuously detect clouds in the vertical direction.3 Verification of model simulations
Time–height distributions of vertically pointed C-band continuous wave radar reflectivity in the Naqu region are displayed in Fig. 2, which shows that all six cases are featured with two obvious processes. 1) The convections were generally initiated in the afternoon, indicating that the formation of these initial convections was closely related to the strong solar radiative heating (Uyeda et al., 2001; Liu et al., 2002), although the feature was not obvious in some cases (Figs. 2a, d, e) due to the weaker radar echo that was not well detected by C-band continuous wave radar. In addition, the vertically pointed C-band radar could only detect the overhead clouds and might miss some stronger clouds around the station. 2) The further development of subsequent cloud system usually occurred after about 1800 LST, and the clouds were generally transformed from convective to stratiform-like clouds with apparent bright bands near 0°C level. Depending on the conditions of solar heating, atmospheric stratification, and prevailing weather systems, the six cases showed obvious different intensities and evolutions during the formation and development of clouds. The cloud-top heights were generally more than 15 km (except for the case of 24 July) with small horizontal scale and deep vertical height (Xu et al., 2002).
Figure 3 presents hourly precipitation rate from observations and simulations averaged over d04. Observations show the surface precipitation rate had two apparent peaks, primarily occurring at 1300–1500 and 2000–2200 LST. For some cases such as those of 3 and 21 July, the maximum peak of precipitation rate occurred in the evening. These cases had similar characteristics in the formation and development of clouds and precipitation. The isolated and weak convective cells were initially formed in the afternoon and then they further developed and merged into a widespread stratiform-like cloud system and produced large precipitation in the evening. This phenomenon was also noted by previous studies based on satellite data (Fujinami and Yasunari, 2001; Kurosaki and Kimura, 2002; Zhu and Chen, 2003; Bhatt and Nakamura, 2005; Fujinami et al., 2005; Fu et al., 2006). The reason for the formation of the large-scale stratiform-like cloud system in the evening over the TP has not been well known. But the eastward expansion of the upper tropospheric anticyclone accompanied with the enhancement of near-surface low pressure might be an important cause for the formation of the large-scale cloud system occurring in the afternoon over the TP (Sugimoto and Ueno, 2010).
However, for other cases such as those of 9, 13, 20, and 24 July, the maximum peak of precipitation rate occurred in the afternoon rather than in the evening. Generally, the deep and strong convections occurred rapidly and generated intense showery precipitation in the afternoon, and in the evening, these convections were evidently weakened and transformed into stratiform-like clouds with an obvious bright band, and produced additional precipitation. This phenomenon can be relatively easier to understand since the convections induced by strong solar radiative heating could be obviously weakened after stronger precipitation and then transformed into stratiform clouds after the convective available potential energy (CAPE) was released. The simulated and observed precipitation trends are well consistent for most cases, whereas large differences of peak precipitation and total amount of precipitation can be found between simulations and observations for a few cases like the one of 09 July. The reasons for the differences are complicated, indicating that further improvements are necessary in many aspects of the model, such as the PBL scheme, the cloud microphysics, topographic forcing, etc. (Yu et al., 2015).
Figures 4 and 5 display the evolutions of simulated strong reflectivity (greater than 5 dBZ) frequency and maximum updraft, respectively, in the main area of operational C-band Doppler radar coverage (d04), for comparison with radar observation. It is shown that the simulated cloud characteristics are basically consistent with what radar has observed. Moreover, the time–height distributions of simulated maximum updraft could reflect better the evolution characteristics of convections in the afternoon (Fig. 5). For the cases of 3 and 21 July (Figs. 4a, e and 5a, e), the frequency of strong reflectivity was relatively low in the afternoon although the updrafts were stronger, with maximum updraft reaching more than 12 m s–1. The high frequency of strong reflectivity occurred in the evening with weak updraft and obvious bright band, indicating that the convections developed in the daytime were transformed into stratiform-like clouds. On the cases of 13, 20, and 24 July (Figs. 4c, d, f and 5c, d, f), the frequency of strong reflectivity, high-frequency reflectivity height, and maximum updraft were apparently larger in the afternoon and weakened in the evening, which were consistent with that observed by radar. For the case of 9 July, the high-frequency reflectivity height and maximum updraft were larger in the afternoon, whereas the frequency of strong reflectivity was larger in the nighttime (Figs. 4b, 5b). The strong reflectivity heights for some cases could exceed 15 km and maximum updraft could approach 40 m s–1, showing that strong convective activity occurred in the studied region in the afternoon. The transformation from convective into stratiform-like cloud system accompanied with obvious bright band in the evening could also be well simulated.
In order to further verify the WRF simulations of clouds over the TP, temporal evolutions of maximum reflectivity from the C-band weather radar observations and WRF simulations over d04 are displayed in Fig. 6 (note that due to the malfunction of radar on 20 July, observations are not used), which shows that temporal evolutions of maximum reflectivity from observations and simulations are basically consistent, and both exhibit obvious diurnal variation. Overall, the simulated convective clouds are more abundant than observations (Zhu et al., 2011; Gao et al., 2018).
In summary, the WRF simulations of clouds and precipitation processes over the TP are basically consistent with observations. Both indicate that the summer clouds over the TP were initially born as isolated convections induced by strong solar radiative heating in the daytime. The isolated convections produced strong or weak precipitation depending on their intensities, and they were generally weakened and transformed into stratiform-like cloud system with obvious bright bands in the evening and produced more or less precipitation depending on their subsequent development. The mechanism responsible for the further development of the widespread stratiform-like cloud system that produced more precipitation for some cases in the evening has not been well understood yet, and further in-depth studies are necessary.4 Characteristics of vertical structure of cloud microphysics
Figure 7 shows the airborne observations of liquid water content on 10 July. Figures 8–12 present the vertical distributions of area-averaged hydrometeors (cloud water, cloud ice, rainwater, snow, and graupel) over d04 for the six cases. Generally, they demonstrate significant vertical distribution characteristics. Figure 7 indicates that when the flight altitude was about 6.3–7 km and the corresponding temperature was between –4 and 10°C, supercooled cloud water content could reach up to 0.2 g m–3, indicating that supercooled cloud water was abundant in Naqu region at that time. As shown in Fig. 8, supercooled cloud water was abundant below 12 km (–40°C) in all the six cases. Large values of cloud water were concentrated in the layer of 0 to –20°C, which greatly contributed to the rapid riming growth of ice particles, suggesting that the model simulation agreed well with observations. Ice crystals were largely distributed above 9 km (–20°C) and the high values even appeared above –40°C layers in the strong convective clouds (Fig. 9), suggesting that the formation of the initial cloud ice principally depended on the deposition process of water vapor. The rainwater was basically concentrated below the melting layer (Fig. 10), suggesting that the formation of rainwater mainly relied on the melting process of precipitating ice particles. Snow and graupel particles had the characteristics of high content and deep vertical distribution, implying active ice phase processes inside the clouds (Figs. 11, 12).5 Microphysical transformation and precipitation formation mechanism
The conversion processes of hydrometeors in the clouds are the critical microphysical processes for cloud development and precipitation formation. The latent heat release caused by the phase transition changes the thermal structure of the atmosphere, and the drag of the precipitating particles strengthens the downdraft, which in turn changes the dynamic and thermal fields of the atmospheric environment and affects the structure of the cloud (Shi et al., 2008). Therefore, it is essential to further analyze main conversion processes of cloud hydrometeors to form precipitation over the plateau.
In the Lin cloud microphysical scheme (Lin et al., 1983), four microphysical processes of rainwater source terms are included: autoconversion of cloud water to rainwater (Acr), accretion of cloud water by rainwater (Ccr), melting of graupel to form rainwater (Mgr), as well as melting of snow to form rainwater (Msr). Five microphysical processes of snow source terms included: autoconversion of cloud ice to snow (Ais), accretion of cloud ice by snow (Cis), the Bergeron process of cloud ice to form snow (Bis), accretion of supercooled cloud water by snow (Ccs), and depositional growth of snow (Svs), respectively. Nine microphysical processes of graupel source terms are included. The formation of graupel particles includes autoconversion of snow to form graupel (Asg), accretion of supercooled rainwater by ice crystal to form graupel (Cri), accretion of snow by supercooled rainwater to form graupel (Csr), and probabilistic freezing of supercooled rainwater to form graupel (Frg). The growth processes of graupel particles include accretion of supercooled cloud water by graupel (Ccg), accretion of supercooled rainwater by graupel (Crg), accretion of ice crystal by graupel (Cig), accretion of snow by graupel (Csg), and depositional growth of graupel (Svg).
Domain-averaged (over d04) mean values over the periods of 1200–1400 LST 3, 9, 13, 20, 21, and 24 July are computed to analyze the vertical distributions of microphysical conversion rates of precipitating particles source terms in Fig. 13. In general, despite the difference in the intensity of the six cases, the conversion processes between various hydrometeors and the major microphysi-cal processes involved in the formation of clouds and precipitation have common features in the six cases.
The vertical distributions of microphysical conversion rates of rainwater source terms (Fig. 13a) show that the melting process of graupel particles (Mgr) was the main source of surface precipitation, indicating that ice process had a critical role in the formation of precipitation in the summer season over the TP. In contrast, the melting process of snow particles (Msr) contributed much less to rainwater. However, the warm rain processes such as the autoconversion process of cloud water to rainwater (Acr) and accretion process of cloud water by rainwater (Ccr) were also active in producing rainwater in the clouds. Moreover, the higher value of Acr appeared at about 7–8 km; note that the large amount of supercooled raindrops were also formed at these levels, which was important to generate a lot of graupel embryos (Fig. 13c). Therefore, although the direct contribution of warm rain process to surface precipitation was smaller, it was vital for the formation of graupel embryos in the summer convective clouds over the plateau.
The vertical distributions of conversion rates of snow source terms (Fig. 13b) indicate that the initial snow particles above 12 km (–40°C) were generated by the autoconversion process of ice crystals, and those below 12 km (–40°C) were mainly formed by the deposition growth of ice crystal (Bergeron process). After the formation of snow particle embryos, their growth above 12 km was predominately caused by the aggregation process with ice crystals, and their growth in the middle and low layers of 6–12 km primarily depended on the riming process with supercooled cloud water. The contributions from deposition and aggregation processes to snow particle growth were relatively small.
The formation and growth of graupel embryos were interesting and critical processes in the summer clouds over the TP since the surface precipitation was directly melted from graupel particles as shown above. The vertical distributions of conversion rates of graupel embryos in Fig. 13c indicate that the heterogeneous accretion freezing processes of supercooled raindrops by cloud ice (Cri) and snow particles (Csr) at levels of 6–8 km had critical roles in forming graupel embryos in the clouds. In contrast, the homogenous freezing process of supercooled raindrops to form graupel particles (Frg) between 9 and 12 km contributed less in producing the graupel embryos.
The growth of graupel particles was also a critical process in the clouds over the TP. Figure 13d indicates that the microphysical processes involved in the growth of graupel particles were different at different altitudes. At higher altitude (above 12 km), the growth of graupel particles mainly depended on aggregation of snow particles (Csg) and cloud ice (Cig) by graupel. Between 6- and 12-km altitudes, the riming of supercooled cloud water (Ccg) by graupel was the most important process for graupel growth. Other processes such as the aggregation of snow particles by graupel and deposition process of graupel (Svg) had secondary importance. The accretion of supercooled rainwater (Crg) by graupel was relatively small.
In order to better display the microphysical structure and conversion characteristics of convective clouds, vertical distribution and microphysical conversion rates of hydrometeors at 1700 LST 9 July 2014 related to the heavy precipitation at site A (31.37°N, 92.35°E) are analyzed. Figure 14 indicates that large contents of supercooled water and graupel particles were found in the clouds with active ice phase processes. The melting of graupel particles provided the major source for surface rainfall. Although the warm cloud microphysical process had small direct contribution to the formation of surface precipitation, it had an important contribution in the formation of supercooled raindrops that were essential in the production of graupel embryos through heterogeneous freezing processes. The growth of graupel particles primarily depended on the riming process with supercooled cloud water. The above analyses show that the dominant microphysical conversion processes at the specific single site were consistent with those of the regional average (Fig. 13).
In all, the transformation of hydrometeors in the clouds and the formation of precipitation over the plateau had some obvious characteristics. The melting process of graupel particles was the main source of surface precipitation, and direct contribution of warm rain process to precipitation was much smaller. However, it was very important for the formation of graupel embryos by generating supercooled raindrops. The growth of graupel particles primarily depended on processes of riming with supercooled cloud water and aggregation with snow particles.6 Conclusions and discussion
Numerical studies have been performed by using the WRF model to simulate six cloud and precipitation processes that occurred during 3–25 July 2014 over the TP, when the TIPEX-III was conducted. The conclusions are as follows.
(1) Cloud and precipitation processes in summer over the TP demonstrate obvious macro features. The six cases in the summer (July) of 2014 were primarily domi-nated by afternoon convective clouds over the plateau, having cloud top heights of more than 15 km and maximum updrafts of 10–40 m s–1. The initial formation of summer clouds and precipitation over the TP showed an obvious diurnal variation, which was closely associated with strong daytime solar radiative heating. The afternoon convections were generally transformed into the stratiform-like clouds with an obvious bright band and usually produced an obvious rainfall in the evening. The above results reveal unique features of convective precipitation in summer over the TP, which is characterized by the conversion from convective to stratiform precipitation. As a result, the duration of precipitation is prolonged and multiple peaks appear during the process. These features are different from the features of thermally induced convective precipitation over low-elevation plains.
(2) The ice processes are very active in the summer clouds over the TP. The summer clouds had high amount of supercooled water content primarily located between 0 and –20°C layers. Ice crystals were mainly formed above –20°C layer and even appeared above –40°C layer in the strong convective clouds. Rainwater mostly appeared under the melting layer, indicating that its formation mainly depended on the melting process of precipitating ice particles. Snow and graupel particles had the characteristics of high content and deep vertical distribution, implying active ice phase processes in the clouds.
(3) The conversion of hydrometeors and mechanisms for the formation of precipitation over the TP in summer demonstrate unique features. The surface precipitation was mainly formed by the melting of graupel particles. Although the warm cloud microphysical process had small direct contribution to the formation of surface precipitation, it had an important contribution in the formation of supercooled raindrops that were essential in the production of graupel embryos through inhomogeneous freezing processes. Above –40°C layer, the formation and growth of snow particles relied on the autoconversion of cloud ice and aggregation process with cloud ice; while below –40°C layer, they depended on the deposition growth of cloud by Bergeron process and riming process with supercooled cloud water. The heterogeneous freezing process of raindrops by the accretion transformation process of snow and ice crystal with supercooled raindrops greatly contributed to the production of graupel embryos; in contrast, the homogenous freezing process of supercooled raindrops had less contribution to the formation of the graupel embryos. The growth of graupel particles mainly relied on the riming process with supercooled cloud water and aggregation of snow particles.
The results above indicate that the summer clouds and precipitation over the TP had some important and unique properties, which should be closely associated with the important topographic effect of the TP. On the one hand, the elevated huge heating source in summer over the TP was favorable to initialize thermal convections due to strong solar radiative process (Chang and Guo, 2016). On the other hand, the huge terrain could also provide a relative cold environmental condition for rapid formation of cloud particles. Therefore, the clouds had some unique properties such as lower cloud-base, higher supercooled water, and more active ice processes over the TP than over the plain regions. Under these conditions, the precipitation could be easily formed, and the weather phenomena such as lightning and icing events could also easily occur in the summertime over the TP (Qiao and Zhang, 1994).
|Bhatt, B. C., and K. Nakamura, 2005: Characteristics of monsoon rainfall around the Himalayas revealed by TRMM precipitation radar. Mon. Wea. Rev., 133, 149–165. DOI:10.1175/MWR-2846.1|
|Chang, Y., and X. L. Guo, 2016: Characteristics of convective cloud and precipitation during summer time at Naqu over Tibetan Plateau. Chinese Sci. Bull., 61, 1706–1720. DOI:10.1360/N972015-01292|
|Chen, F., and J. Dudhia, 2001: Coupling an advanced land-surface/hydrology model with the Penn State/NCAR MM5 modeling system. Part I: Model description and implementation. Mon. Wea. Rev., 129, 569–585. DOI:10.1175/1520-0493(2001)129<0569:CAALSH>2.0.CO;2|
|Dudhia, J., 1989: Numerical study of convection observed during the winter monsoon experiment using a mesoscale two-dimensional model. J. Atmos. Sci., 46, 3077–3107. DOI:10.1175/1520-0469(1989)046<3077:NSOCOD>2.0.CO;2|
|Fu, Y. F., G. S. Liu, G. X. Wu, et al., 2006: Tower mast of precipitation over the central Tibetan Plateau summer. Geophys. Res. Lett., 33, L05802. DOI:10.1029/2005GL024713|
|Fu, Y. F., H. T. Li, and Y. Zi, 2007: Case study of precipitation cloud structure viewed by TRMM satellite in a valley of the Tibetan Plateau. Plateau Meteor., 26, 98–106.|
|Fu, Y. F., Q. Liu, Y. Zi, et al., 2008: Summer precipitation and latent heating over the Tibetan Plateau based on TRMM measurements. Plateau Mountain Meteor. Res., 28, 8–18. DOI:10.1659/mrd.0969|
|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. DOI:10.2151/jmsj.79.1207|
|Fujinami, H., S. Nomura, and T. Yasunari, 2005: Characteristics of diurnal variations in convection and precipitation over the southern Tibetan Plateau during summer. SOLA, 1, 49–52. DOI:10.2151/sola.2005-014|
|Gao, W. H., C. H. Sui, J. W. Fan, et al., 2016: A study of cloud microphysics and precipitation over the Tibetan Plateau by radar observations and cloud-resolving model simulations. J. Geophys. Res. Atmos., 121, 13735–13752. DOI:10.1002/2015JD024196|
|Gao, W. H., L. P. Liu, J. Li, et al., 2018: The microphysical properties of convective precipitation over the Tibetan Plateau by a subkilometer resolution cloud-resolving simulation. J. Geophys. Res. Atmos., 123, 3212–3227. DOI:10.1002/2017JD027812|
|Grell, G., and D. Devenyi, 2002: A generalized approach to parameterizing convection combining ensemble and data assimilation techniques. Geophys. Res. Lett., 29, 1693. DOI:10.1029/2002GL015311|
|Haginoya, S., H. Fujii, T. Kuwagata, et al., 2009: Air–lake interaction features found in heat and water exchanges over Nam Co on the Tibetan Plateau. SOLA, 5, 172–175. DOI:10.2151/sola.2009-044|
|Janjic, Z. I., 2002: Nonsingular implementation of the Mellor-Yamada level 2.5 scheme in the NCEP meso model. NECP Office Note, No. 437, 61 pp.|
|Kurosaki, Y., and F. Kimura, 2002: Relationship between topography and daytime cloud activity around Tibetan Plateau. J. Meteor. Soc. Japan, 80, 1339–1355. DOI:10.2151/jmsj.80.1339|
|Li, M., Y. Ma, Z. Hu, et al., 2009: Snow distribution over the Namco lake area of the Tibetan Plateau. Hydrol. Earth Syst. Sci., 13, 2023–2030. DOI:10.5194/hess-13-2023-2009|
|Li, D., A. J. Bai, and S. J. Huang, 2012: Characteristic analysis of a severe convective weather over Tibetan Plateau based on TRMM data. Plateau Meteor., 31, 304–311.|
|Li, Y. Y., and M. H. Zhang, 2016: Cumulus over the Tibetan Plateau in the summer based on CloudSat–CALIPSO data. J. Climate, 29, 1219–1230. DOI:10.1175/JCLI-D-15-0492.1|
|Lin, Y. L., R. D. Farley, and H. D. Orville, 1983: Bulk parameterization of the snow field in a cloud model. J. Climate Appl. Meteor., 22, 1065–1092. DOI:10.1175/1520-0450(1983)022<1065:BPOTSF>2.0.CO;2|
|Liu, L. P., R. Z. Chu, X. M. Song, et al., 1999: Summary and preliminary results of cloud and precipitation observation in Qinghai–Xizang Plateau in GAME-TIBET. Plateau Meteor., 18, 441–450.|
|Liu, L. P., J. M. Feng, R. Z. Chu, et al., 2002: The diurnal variation of precipitation in monsoon season in the Tibetan Plateau. Adv. Atmos. Sci., 19, 365–378. DOI:10.1007/s00376-002-0028-6|
|Liu, L. P., J. F. Zheng, Z. Ruan, et al., 2015: Comprehensive radar observations of clouds and precipitation over the Tibetan Plateau and preliminary analysis of cloud properties. J. Meteor. Res., 29, 546–561. DOI:10.1007/s13351-015-4208-6|
|Maussion, F., D. Scherer, R. Finkelnburg, et al., 2011: WRF simulation of a precipitation event over the Tibetan Plateau, China—an assessment using remote sensing and ground observations. Hydrol. Earth Syst. Sci., 15, 1795–1817. DOI:10.5194/hess-15-1795-2011|
|Mlawer, E., S. Taubman, P. Brown, et al., 1997: Radiative transfer for inhomogeneous atmospheres: RRTM, a validated correlated-k model for the longwave . J. Geophys. Res. Atmos., 102, 16663. DOI:10.1029/97JD00237|
|Pan, X., and Y. F. Fu, 2015: Analysis on climatological characteristics of deep and shallow precipitation cloud in summer over Qinghai–Xizang Plateau. Plateau Meteor., 34, 1191–1203.|
|Qiao, Q. M., and Y. G. Zhang, 1994: Tibetan Plateau Weather. China Meteorological Press, Beijing, 45–46. (in Chinese)|
|Sato, T., T. Yoshikane, M. Satoh, et al., 2008: Resolution dependency of the diurnal cycle of convective clouds over the Tibetan Plateau in a mesoscale model. J. Meteor. Soc. Japan, 86A, 17–31. DOI:10.2151/jmsj.86A.17|
|Shen, Y., P. Zhao, Y. Pan, et al., 2014: A high spatiotemporal gauge–satellite merged precipitation analysis over China. J. Geophys. Res. Atmos., 119, 3063–3075. DOI:10.1002/2013JD020686|
|Shi, Y. Q., X. F. Lou, X. J. Deng, et al., 2008: Simulations of mesoscale and microphysical characteristics of cold front clouds in South China. Chinese J. Atmos. Sci., 32, 1019–1036.|
|Sugimoto, S., and K. Ueno, 2010: Formation of mesoscale convective systems over the eastern Tibetan Plateau affected by plateau-scale heating contrasts. J. Geophys. Res. Atmos., 115, D16105. DOI:10.1029/2009JD013609|
|Ueno, K., H. Fujii, H. Yamada, et al., 2001: Weak and frequent monsoon precipitation over the Tibetan Plateau. J. Meteor. Soc. Japan, 79, 419–434. DOI:10.2151/jmsj.79.419|
|Uyeda, H., H. Yamada, J. Horikomi, et al., 2001: Characteristics of convective clouds observed by a Doppler radar at Naqu on Tibetan Plateau during the GAME-Tibet IOP. J. Meteor. Soc. Japan, 79, 463–474. DOI:10.2151/jmsj.79.463|
|Xu, J. Y., B. Zhang, M. H. Wang, et al., 2012: Diurnal variation of summer precipitation over the Tibetan Plateau: A cloud-resolving simulation. Ann. Geophys., 30, 1575–1586. DOI:10.5194/angeo-30-1575-2012|
|Xu, X. D., and L. S. Chen, 2006: Advances of the study on Tibetan Plateau experiment of atmospheric sciences. J. Appl. Meteor. Sci., 17, 756–772.|
|Xu, X. D., M. Y. Zhou, J. Y. Chen, et al., 2002: A comprehensive physical pattern of land–air dynamic and thermal structure on the Qinghai–Xizang Plateau. Sci. China Earth Sci., 45, 577–594. DOI:10.1360/02yd9060|
|Xu, X. D., T. L. Zhao, C. G. Lu, et al., 2014: Characteristics of the water cycle in the atmosphere over the Tibetan Plateau. Acta Meteor. Sinica, 72, 1079–1095.|
|Yu, R. C., J. Li, Y. Zhang, et al., 2015: Improvement of rainfall simulation on the steep edge of the Tibetan Plateau by using a finite-difference transport scheme in CAM5. Climate Dyn., 45, 2937–2948. DOI:10.1007/s00382-015-2515-3|
|Zhu, G. F., and S. J. Chen, 2003: Analysis and comparison of mesoscale convective systems over the Qinghai–Xizang (Tibetan) Plateau. Adv. Atmos. Sci, 20, 311–322. DOI:10.1007/BF032690789|
|Zhu, S. C., Y. Yin, L. J. Jin, et al., 2011: A numerical study of the vertical transport of water vapor by intense convection over the Tibetan Plateau. Chinese J. Atmos. Sci., 35, 1057–1068.|