**4. Results**

### *4.1. Diurnal Cycle of Precipitation and Cloud*

During the mei-yu period, the diurnal variation in precipitation from HE satellite estimates shows phase propagation from west to east over the SB (103◦–110◦E), while no such propagation can be seen over the TP (90◦–102◦E) (Figure 2a), which is similar to the results of [23]. The diurnal maximum precipitation rate over the TP is concentrated between 1500 LT and 2100 LT, and the diurnal minimum precipitation rate lies within 0000–1200 LT. Over the SB, meanwhile, the diurnal maximum precipitation rate occurs at night, and then decreases to a minimum in the morning, which was also found in previous studies (e.g., [42]). The result for the PF number percentage is similar to that of the precipitation amount, except the maximum PF number percentage occurs at the boundary of the TP and SB (Figure 2c), which is related to the convergence rising motion over the slopes in the afternoon and rising motion in the center of the SB at night. During the midsummer period, the diurnal variation in the precipitation rate shows no propagation from the TP to SB, which is different from the mei-yu period. The diurnal maximum precipitation rate is concentrated between 1500 LT and 2100 LT, while the diurnal minimum precipitation rate lies within 0000–0600 LT, which is similar to that during the mei-yu period (Figure 2b). The average precipitation rate over the SB is larger than that over the TP. The result for the PF number percentage is similar to that of the precipitation amount (Figure 2d). The

reasons of the diurnal propagation of rainfall are related to the decreases in the zonal wind profile from the mei-yu to midsummer period and the diurnal cycle of low-level winds over the SB [23].

**Figure 2.** The (**<sup>a</sup>**,**b**) diurnal variation in precipitation and (**<sup>c</sup>**,**d**) precipitation feature number percentage during the (**<sup>a</sup>**,**<sup>c</sup>**) mei-yu and (**b**,**d**) midsummer period from the Hydro Estimator satellite rainfall estimates.

The precipitation from HE satellite estimates and ground observations are compared in Figure S1. Over the TP, the maximum precipitation rate from the HE satellite is 3 h ahead of that of the ground observations during the mei-yu and midsummer periods, and the precipitation rate from HE satellite is larger than that observed on the ground. Over the SB, the maximum precipitation rate from the HE satellite is also 3 h ahead of that of the ground observations during the mei-yu period, while it is in-phase with ground observations during the midsummer period. The result for the PF number percentage is comparable to that of the precipitation rate.

The diurnal variations in cloud cover, LWC, and IWC from ERA5 are shown in Figure 3. During the mei-yu period, the diurnal variation in cloud cover (Figure 3a), LWC (Figure 3c), and IWC (Figure 3e) shows propagation from west to east over the SB, while no diurnal propagation is apparent over the TP. The diurnal maximum cloud cover and LWC over the TP is concentrated between 1800 LT and 0300 LT, while over the SB they lie within 1500– 2400 LT. The diurnal minimum cloud cover and LWC lie within 1000–1500 LT over both the TP and SB. The cloud cover and LWC at the border of the SB and TP are smaller than those over the inner parts of the two regions. The cloud cover and LWC in the SB are larger than those over the TP. The diurnal maximum IWC over the TP is concentrated between 1500 LT and 2100 LT, while the range for the diurnal minimum IWC is 0600–1200 LT. Over the SB, the diurnal maximum IWC is in the mid-afternoon to early evening (1500–2100 LT) and then decreases to a minimum during the morning (0600–1200 LT). The cloud IWC over the edge of TP and SB is larger than that over the inner TP and SB.

**Figure 3.** Diurnal variation in (**<sup>a</sup>**,**b**) cloud cover, (**<sup>c</sup>**,**d**) total column cloud liquid water content, and (**<sup>e</sup>**,**f**) cloud ice water content from ERA5 over the Tibetan Plateau and Sichuan Basin during the (**<sup>a</sup>**,**c**,**<sup>e</sup>**) mei-yu and (**b**,**d**,**f**) midsummer periods.

During the midsummer period, the diurnal variation in cloud cover (Figure 3b), LWC (Figure 3d), and IWC (Figure 3f) shows no propagation from the TP to SB. Over the TP, the diurnal maximum cloud cover and LWC lie within 1500–2100 LT and 0000–1200 LT, respectively, while the ranges for the diurnal minimum cloud cover and LWC are 1000– 1500 LT and 1800–2100 LT. The cloud cover and LWC at the border of the SB and TP are smaller than those over the inner parts of the TP and SB. The cloud cover and LWC over the SB are larger than over the TP. Over both the TP and SB, the diurnal maximum IWC occurs in the mid-afternoon to early evening (1500–2100 LT) and reaches a diurnal minimum in the morning (2400–1200 LT).

The ratios of cloud LWC and IWC to total cloud water were separately calculated, and the results are presented in Figure 4. The results show that LWC accounts for more than 60% of total water during almost the entire diurnal cycle over the inner TP and SB during the mei-yu period, except for the period 1500–1800 LT at the edge of the SB and TP. The IWC accounts for more than 60% at around 1800 LT at the edge of the TP and SB, which is

related to the convergence rising motion over the slopes at the edge of the two regions. The proportions of LWC and IWC during midsummer are similar to those during the mei-yu period. However, there are two IWC centers during the midsummer period. One is in the period 1500–1800 LT at the edge of the SB and TP, and the other is located in the region 92◦–94◦E during 1700–2000 LT, which is related to development of the westerly jet and southern water vapor transportation.

**Figure 4.** Ratio of cloud (**<sup>a</sup>**,**<sup>c</sup>**) liquid water and (**b**,**d**) ice water to total cloud water during the (**<sup>a</sup>**,**b**) mei-yu and (**<sup>c</sup>**,**d**) midsummer periods.

During the mei-yu period, the cloud LWC mainly distributes between the 900 hPa and 450 hPa level in the daytime and the 850–450 hPa level in the nighttime over the SB, while it mainly distributes between the 600 hPa and 450 hPa level in the daytime and the 600–400 hPa level in the nighttime over the TP (Figure 5a,b). The cloud LWC in the night is larger than that in the daytime over the SB. During the midsummer period, it is mainly similar to that over the SB and TP (Figure 5c,d) except that the cloud LWC during the midsummer period is smaller than that in the mei-yu period.

During the mei-yu period, the cloud IWC mainly distributes between the 450 hPa and 150 hPa level in the daytime and the 500–100 hPa level in the nighttime over the SB, while it mainly distributes between the 500 hPa and 180 hPa level in the daytime and the 450–100 hPa level in the nighttime over the TP (Figure 6a,b). The cloud IWC in the night is smaller than that in the daytime over the SB. During the midsummer period, it is mainly similar to that over the SB and TP except that the cloud IWC during the midsummer period is larger than that in the mei-yu period (Figure 6c,d). The cloud IWC in the daytime during the midsummer period mainly is concentrated over the edge of the SB and TP.

**Figure 5.** The pressure level of cloud liquid water content (shaded color), cloud base (green line), zero degree level (orange line), and the topography (red line) along with longitude during the (**<sup>a</sup>**,**b**) mei-yu and (**<sup>c</sup>**,**d**) midsummer periods during the (**<sup>a</sup>**,**<sup>c</sup>**) daytime and (**b**,**d**) nighttime.

**Figure 6.** Similar with Figure 5, but for cloud ice water content.

The diurnal cycle of CBH AGL in ERA5 is presented in Figure 7. During the mei-yu period, the diurnal variation in CBH AGL shows propagation from west to east over the SB, while no propagation is apparent over the TP. The diurnal maximum CBH AGL over the TP is concentrated between 1500 LT and 2400 LT, while the diurnal minimum CBH AGL lies within 0900–1200 LT. The CBH AGL in the early evening is higher than that in the daytime over the SB, which is consistent with the nocturnal maximum precipitation. The CBH AGL over the SB is higher than that over the TP (Figure 7a). During the midsummer period, the diurnal variation in CBH AGL shows no propagation from the TP to SB. The diurnal maximum CBH AGL over the TP is concentrated between 2100 LT and 0300 LT, while the diurnal minimum lies within 0700–1500 LT. Over the SB, the diurnal maximum CBH AGL lies within the period 2100–0600 LT, while the diurnal minimum is within 1500–2100 LT (Figure 7b). The average CBH AGL observed by the cloud radar at Yushu, Naqu, and Linzhi is close to that in ERA5 (Figure S2). The maximum CBH AGL from ERA5 is in-phase with the observations averaged from the three sites during the mei-yu (Figure S2a) and midsummer (Figure S2b) periods over the TP.

**Figure 7.** Diurnal variation in cloud base height from ERA5 over the Tibetan Plateau and Sichuan Basin during the (**a**) mei-yu and (**b**) midsummer periods.

The CBH AGL is compared with the zero degree level AGL in Figure 5. The zero degree level AGL is much higher than the CBH AGL over the TP and SB during the mei-yu period, while CBH AGL nears the zero degree level AGL over the TP during midsummer. The latter is related to the air temperature lapse rate, which is larger over the TP than the SB. The zero degree level MSL over the TP is higher than that over the SB. The thickness of cloud LWC over the SB is larger than that near the edge of the TP during the mei-yu and midsummer periods. The zero degree level MSL increases while the CBH remains stable with the increase in distance from the TP. However, super-cooled liquid water can exist down to −40 ◦C.

#### *4.2. Factors Influencing the Formation of Cloud over the TP and SB*

During the mei-yu period, the diurnal variation in dewpoint spread shows no propagation from morning to night over the SB and TP (Figure 8a,b). As is well known, the dewpoint spread is mainly influenced by the solar radiation and surface heating. The diurnal maximum dewpoint spread is concentrated between 1500 LT and 2100 LT, and the diurnal minimum dewpoint spread lies within 0000–0900 LT in the region 90◦–100◦E during mei-yu period (Figure 8a). During the midsummer period, the diurnal variation in dewpoint spread shows no propagation from the TP to SB. The diurnal maximum dewpoint spread is concentrated between 1500 LT and 2100 LT, while the range for the diurnal minimum is 0600–1000 LT (Figure 8b). The dewpoint spread over the SB is larger than that over the TP.

**Figure 8.** Diurnal variation in (**<sup>a</sup>**,**b**) dewpoint spread and (**<sup>c</sup>**,**d**) convective available potential energy (CAPE) from ERA5 over the Tibetan Plateau and Sichuan Basin during the (**<sup>a</sup>**,**<sup>c</sup>**) mei-yu and (**b**,**d**) midsummer periods.

A comparison of the dewpoint spread between the observation and ERA5 is given in Figure S3. Over both the TP and SB, the timing of the diurnal maximum dewpoint spread is similar to that in the observations during both the mei-yu and midsummer period. The dewpoint spread from ERA5 is larger than that in the observations, especially during the early evening. Some studies show that the dewpoint spread makes a profound contribution to the LCL (e.g., [43]). In this study, the CBH is close to the LCL during the summer period with large amounts of LWC during that same season (Table 2).

**Table 2.** Comparison of cloud base height (CBH) from ERA5 dataset and lifting condensation level (LCL) above ground level (AGL) during summer over the Tibetan Plateau (TP) and Sichuan Basin (SB).


To investigate the effect of dewpoint spread on the CBH AGL, a correlation analysis between the two at each longitudinal grid point in ERA5 was carried out (Figure 9a). Over the TP, the correlation coefficient reaches up to 0.8 at the 5% significance level. However, it decreases at the edge of the TP and SB and becomes gradually negative, reaching a maximum negative value over the inner part of the SB. According to the above-mentioned results, CBH AGL propagates diurnally from the TP to SB, but the dewpoint spread does not. Therefore, the dewpoint spread makes a profound contribution to the CBH over the TP, but little contribution to the CBH AGL over the SB because of the stronger turbulence and lower air density over the TP than the SB [44].

**Figure 9.** Correlation between (**a**) dewpoint spread and CBH and (**b**) water vapor flux and ΔV during the mei-yu (blue line) and midsummer (red line) periods. (**<sup>c</sup>**,**d**) Correlation of ΔV, CAPE, cloud LWC and IWC with precipitation rate during the (**c**) mei-yu and (**d**) midsummer periods.

From the spatial distribution of the water vapor flux, the transportation of water vapor can be seen to be mainly from the west (Figure 10), which is similar to the results of [45]. The water vapor flux decreases from the mei-yu period to the midsummer period, especially over the SB. This is related to the decreases in the mean zonal wind profile and low-level winds from the mei-yu to midsummer period.

**Figure 10.** Integral of water vapor flux at 1800 LT from ERA5 over the (**a**) mei-yu and (**b**) midsummer periods.

To investigate the relationship between the total water vapor transportation and westerly wind, the difference in horizontal wind speed between 200 hPa and 500 hPa (ΔV) was calculated (Figure 11); that is, ΔV = [u(200) − u(500)]<sup>2</sup> + [v(200) − v(500)]2,where u(200) and v(200) refer to u-wind and v-wind at the 200 hPa level, respectively, and u(500) and v(500) refer to the same but at the 500 hPa level. Correlation analysis was carried out between the water vapor flux and ΔV. The ΔV over both the TP and SB decreases from the mei-yu to midsummer period, and the ΔV over the TP is larger than that over the SB during both periods. The ΔV correlates positively with water vapor transportation at the 5% significance level over the TP and SB, and the correlation coefficient over the TP is smaller than that over the SB during the mei-yu and midsummer periods (Figure 9b). The ΔV has a profound impact on the transportation of water vapor over the SB. The correlation coefficient decreases dramatically from the mei-yu to midsummer period in the region of 90◦–100◦E where the steep slopes and SB are situated.

**Figure 11.** Diurnal variation in ΔV during the (**a**) mei-yu and (**b**) midsummer periods.

#### *4.3. Factors Influencing the Formation of Precipitation over the TP and SB*

To investigate the factors influencing the precipitation, the diurnal cycle of CAPE is shown in Figure 8c,d. During the mei-yu period, CAPE shows no propagation over the TP and SB (Figure 8c). The diurnal maximum CAPE occurs in the late afternoon (1500–1900 LT) and early evening (1900–2100 LT) over the TP and SB during the mei-yu period, and the diurnal minimum occurs in the morning (0600–1000 LT). The CAPE over the TP is larger than that over the SB (Figure 8c), which is different from the precipitation and cloud LWC. During the midsummer period, CAPE shows no diurnal propagation over the TP and SB (Figure 8d). The diurnal maximum and minimum CAPE values during the mei-yu period are similar to those during the midsummer period. CAPE is related to the maximum potential vertical velocity of air within an updraft, so higher values of CAPE are an indicator of precipitation.

To investigate the influence of diurnal variation in CAPE, cloud LWC and IWC on precipitation, correlation analysis was performed. During the mei-yu period, the correlation coefficient between CAPE and precipitation over the TP is larger than that over the SB. The correlation coefficient reaches a maximum (0.8) at the edge of the TP and SB at the 5% significance level (Figure 9c). The correlation coefficient between CAPE and precipitation during the midsummer period is smaller than that during the mei-yu period (Figure 9d). These results indicate that CAPE has a larger impact on precipitation over the TP than over the SB.

The correlation coefficient between cloud IWC and precipitation over the TP and SB is larger than 0.7, with a significance level of 5%, during the mei-yu period. The correlation coefficient between cloud LWC and precipitation over the SB during the midsummer period (Figure 9d) is smaller than that during the mei-yu period, which decreases with longitude from west to east due to the diurnal propagation of LWC and precipitation (Figure 9c), while the opposite is the case for cloud IWC. However, the correlation between the cloud IWC and precipitation over the edge of the TP and SB is better than over the inner parts of the TP and SB due to the convergence rising motion over the slopes. The above results indicate that the cloud LWC makes a large contribution to precipitation over the SB. The precipitation over the edge of the TP and SB (i.e., the steep downstream slopes) is influenced by the cloud IWC.

### **5. Conclusions**

The diurnal variations in precipitation in the HE Satellite Rainfall Estimates and cloud parameters (CBH AGL, cloud cover, cloud LWC and IWC, and dewpoint spread) in the ERA5 dataset were analyzed over the TP and SB. Results show that the precipitation and cloud parameters show diurnal propagation from morning to night over the SB during the mei-yu period, while no such diurnal propagation is apparent over the TP during both the mei-yu and midsummer periods. The precipitation, cloud LWC and IWC, and cloud cover over the TP are smaller than over the SB. The dewpoint spread over the TP is larger than over the SB. The diurnal maximum precipitation and cloud LWC and IWC are concentrated in the early evening, while the diurnal minima occur in the morning.

The precipitation from the HE Satellite is larger than that in the observations. Over the SB, the maximum precipitation from the HE Satellite is 3 h ahead of observations. The result for the PF number percentage is similar to that of the precipitation amount. During both the mei-yu and midsummer period, the cloud LWC from ERA5 at night is larger than that in the daytime. Cloud LWC accounts for more than 60% of total water during almost the entire diurnal cycle over the inner parts of the TP and SB during the mei-yu period except for late afternoon at the edge of the SB and TP. The cloud IWC accounts for more than 60% during early evening at the edge of the TP and SB.

The CBH AGL, LCL AGL, and zero degree level AGL are almost equal over the TP during the summer period. The zero degree level AGL over the SB is higher than that over the TP because the air temperature lapse rate over the TP is larger than that over the SB. The thickness of the LWC over the SB is larger than that over the TP. The dewpoint spread makes a profound contribution to the CBH AGL over the TP but little contribution to the CBH AGL over the SB because of the stronger turbulence and lower air density over the TP. CAPE has a larger impact on precipitation over the TP than over the SB. The cloud LWC makes a large contribution to the precipitation over the SB, which is related to the mean zonal wind profile and diurnal cycle of low-level winds. The precipitation over the edge of the TP and SB (i.e., the steep downstream slopes) is influenced by the cloud IWC owing to the convergence rising motion over the slopes.

Although the distribution of cloud LWC and IWC over the TP and SB was obtained in this study, the distribution of supercooled water and ice level over the edge of TP and SB is still unclear and needs to be investigated. In addition, the results of ERA5 data were uncertainty limited by the resolution of ERA5 data, especially concerning quite an extreme region. The current cloud radar sites are very sparse, making it impossible directly to observe the cloud microphysical parameter data over the entire TP and SB. In the future, the accuracy of the reanalysis dataset and observations over the TP and SB at night should be improved. More cloud radar observations at different locations over the TP and SB are needed in the future. The phase of water within cloud needs to be studied using different kinds of observational and model data.

**Supplementary Materials:** The following supporting information can be downloaded at https: //www.mdpi.com/article/10.3390/rs14112711/s1: Table S1: The information of ground meteorological observation sites of China Meteorological Administration; Table S2: The information of radiosonde station of China Meteorological Administration; Figure S1: The precipitation from Hydro Estimator satellite estimates (blue line) and ground observations (red line) during (a,b) meiyu and (c,d) midsummer period over the (a,c) Tibetan Plateau and (b,d) Sichuan Basin. Blue shaded area is the standard error of precipitation from Hydro Estimator satellite estimates, and red shaded area is the standard error of precipitation from ground observations. The percentage of precipitation feature during (e,f) mei-yu and (g,h) midsummer period over the (e,g) Tibetan Plateau and (f,h) Sichuan Basin and their error bar. Blue shaded area is the standard error of the percentage of precipitation feature from Hydro Estimator satellite estimates, and red shaded area is the standard error of percentage of precipitation feature from ground observations; Figure S2: The comparison of cloud base height from ERA5 dataset (blue line) and cloud radar observations (red line) and their standard error (blue shaded area for ERA5 and red shaded area for observations) during the (a) mei-yu and (b) midsummer period; Figure S3: The comparison of dew point spread from the ERA5 dataset (blue line) and ground observations (red line) and their standard error (blue shaded area for ERA5 dataset, and red shaded area for observations) during the (a,b) mei-yu and (c,d) midsummer period over the (a,c) Tibetan Plateau and (b,d) Sichuan Basin.

**Author Contributions:** Conceptualization, Y.L.; methods, B.C.; software, B.C.; validation, Y.L.; writing—original draft preparation, B.C.; writing—review and editing, X.Y.; visualization, B.L.; supervision, J.W.; project administration, X.Y.; funding acquisition, X.Y. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by the National Science Foundation of China (NSFC) (41975130, 42175174), Strategic Priority Research Program of Chinese Academy of Sciences (XDA20050102), the Second Tibetan Plateau Scientific Expedition and Research (STEP) program (2019QZKK0102), the Research Foundation of Chengdu University of Information Technology (KYTZ201810), and the Project of Science and Technology Plan of Sichuan (2019YJ0408).

**Data Availability Statement:** The ERA5 data employed in this study are available from https: //www.ecmwf.int/en/forecasts/datasets/reanalysis-datasets/era5 (accessed on 1 January 2022). The observational data are available from the third Tibetan Plateau Atmospheric Experiment (TIPEX3; https://data.cma.cn/site/article/id/28986.html (accessed on 1 January 2022)).

**Conflicts of Interest:** The authors declare no conflict of interest.
