**3. Results**

### *3.1. SASM Evolution and Synoptic Situations*

The SASM evolution could have grea<sup>t</sup> impacts on the weather and climate in Asia through general circulation changes [1,17,21,23,27,32]. The onset of SASM usually occurs at the end of May or early June in South Asia [33]. To investigate the impacts of the SASM on the land-air heat exchange processes over TP, the SASM evolution during the TIPEX III experiment in 2014 is first analyzed.

To characterize the SASM evolution in the TIPEX III in 2014, a SASM index (SASMI) from Wang et al. [34] is adopted in this study. The SASMI is defined by the standardized difference of averaged zonal wind speeds at 850 hPa from two regions, 5–15◦N, 40–80◦E, and 20–30◦N, 70–90◦E. A large positive SASMI corresponds to a strong monsoonal circulation, while a large negative SASMI corresponds to a weak monsoon. Figure 2 presents the daily variations in SASMI during the observation period from 1 May to 30 September 2014. This figure shows that the SASMI turns positive on 6 June, and then begins with a sudden increase to a maximum on 11 June 2014, with the maximum SASMI value being larger than the average value (5.0 m/s). At this time, a strong cyclonic circulation prevailed over the Arabian Sea at 850 hPa (figure not shown), which represents the SASM onset in 2014. Thereafter, SASM experienced several active periods during 11–16 July, 22 July, 29 July–1 August, 5 August, and 27 August–2 September, with positive SASMI exceeding the standard deviation, and break periods during 21–23 June, 28–30 June, 13–15 August, 24–25 August, with negative SASMI values exceeding the standard deviation. Considering the observation period, 29 July–1 August, and 5 August were selected as the SASM active period, and 13–15 August, and 24–25 August were selected as the SASM break period. In order to minimize the influence of the solar altitude angle, the active and break periods were selected close to each other. In the following studies, the land-air heat transfers, as well as radiation fluxes, will be averaged for the SASM active and break periods, to investigate their responses to the SASM evolution.

**Figure 2.** Variations in the SASM index (SASMI), with the averaged SASMI value (5.0 m/s) shown as a solid line, dashed lines represent one standard deviation (3.8 m/s) above and below the mean value.

To illustrate convection related to the SASM evolution, the outgoing long-wave radiation (OLR) was averaged for the entire observational period and for the SASM active and break periods; their distributions are shown in Figure 3. The low OLR values represent strong convection and vice versa. During the observational period (Figure 3a), there were three strong convective activity centers (with OLR values lower than 190 Wm−2) in the eastern part of the Bay of Bengal, the northeastern part of India and the central portion of

TP. During the SASM active period (Figure 3b), the strong convective activities over the Bay of Bengal and North India both intensified and extended northward, with central OLR values less than 160 Wm−2. The convection over the central TP became more severe and enlarged, with central OLR values lower than 160 Wm−<sup>2</sup> and covering almost the entire TP region. During the break period (Figure 3c), however, all the three convections moved southwards, and the convective activity center over the TP as shown in Figure 3a retreated to the south of TP. Therefore, obvious differences can be found in the OLR distributions between the SASM active and break periods, especially over the TP regions, representing the dominant strong and weak convections there.

**Figure 3.** Outgoing long-wave radiation (OLR) distributions over the South Asia and TP regions, averaged for the (**a**) observation period, (**b**) SASM active period, and (**c**) SASM break period, with a contourintervalof10Wm−2.

Figure 4 presents the averaged wind and specific humidity fields at 500 hPa in the entire observation period, SASM active and break periods. During the observation period (Figure 4a), a cyclone formed over Central India and the Bay of Bengal, with the highest specific humidity greater than 4.5 g/kg. A westerly existed in the west TP, bringing dry air masses there (with the specific humidity less than 3.0 g/kg). A southwesterly prevailed over the south and central plateau regions, with specific humidity values greater than 5.5 g/kg. During the SASM active period (Figure 4b), the cyclone over Central India and the Bay of Bengal intensified, and moved northward and westward, associated with an enhancement of moisture (central specific humidity greater than 5.5 g/kg). An obvious cyclonic circulation appeared over the main body of the plateau. The dry westerly prevailed over the west TP weakened. The southwesterly existed in the south and Central TP became stronger, leading to higher moisture levels over the entire TP, with central values greater than 7.0 g/kg. During the SASM break period (Figure 4c), the cyclone with high water vapor over Central India and the Bay of Bengal as shown in Figure 4b disappeared. The dry westerly over the west and southwest of TP strengthened significantly, with moisture values below 2.0 g/kg. The weakened southwesterly led to the retreat and shrinking of the high moisture center over the south and Central TP.

#### *3.2. The Impacts of SASM Evolution on the Radiation Heat Transfers*

From the above results, large differences between the SASM active and break periods were observed from the synoptic situations, including the convection, wind, and moisture fields over the South Asian and TP regions. In the following study, the impacts of SASM evolution on land-air heat exchange processes, as well as the radiation conditions will be covered.

**Figure 4.** Horizontal distributions of wind (arrows, units: m/s) and specific humidity. (shadings, units: g/kg) at 500 hPa averaged for (**a**) the observation period, (**b**) the SASM active period, and (**c**) the SASM break period.

Figure 5 shows the diurnal variations of the downward shortwave radiation flux (DR), averaged for the observations period, SASM active, and break periods. The diurnal variations of DR are similar for all stations, showing increases at approximately 06:00 LST (local standard time) near sunrise, reaching a maximum at noon, and decreasing to almost zero at approximately 18:00 LST near sunset. However, large DR amplitude differences were found among the 8 plateau stations. The strongest DR occurred over Ali station (northwest plateau), with daily average and maximum values of 319.2 and 1007.8 Wm−2, respectively, which is mainly due to the low precipitation and less moisture in the air. Ali station is located in the northwest of the Qinghai-Tibet Plateau, and Ali is a very dry area with very little rainfall. According to the observation data, Ali station had no precipitation from July to September in 2014, resulting in low water vapor content in the air. The second strong DR is found over Namco station (southeast plateau), with daily averaged and maximum values of 222.9 and 846.3 Wm−2, respectively. For the other stations, the DR differences are noticeably smaller, and the difference of the daily averaged values was less than 20 Wm−2, with a variation between 187.3 Wm−<sup>2</sup> and 206.1 Wm−2. During the SASM active/break periods, the DR is greatly weakened/strengthened at most plateau stations. For example, the daily averaged value of DR at the Baingoin station was 194.2 Wm−<sup>2</sup> during the observation period, and the DR varies from 133.2 Wm−<sup>2</sup> (weakened by 31.4%) during the SASM active period and 234.5 Wm−<sup>2</sup> (strengthened by 20.8%) during the break period, respectively. The weakened/strengthened DR is closely related to the strong/weak convections over the plateau region during the SASM active/break periods (see Figure 3). The strong/weak convections can result in more/less cloudiness, which further affects the solar radiation by blocking/enhancing effects [17,27,32].

Figure 6 presents the net radiation fluxes (NR) for the 8 stations over the TP during the observations, SASM active, and break periods. The NR patterns exhibit similar diurnal variations as those of the DR, with positive values (net heating) during daytime from approximately 06:00 LST to approximately 18:00 LST and negative values (net cooling) for the other times of the day over most of the plateau stations. NR amplitude differences are also seen among the 8 plateau stations. The strongest net radiation also occurs at Namco station (southeast plateau), with daily averaged and maximum values of 133.2 and 609.2 Wm−2, respectively. The second strongest NR is found over Lhari station (central plateau), with a diurnally averaged value of 120.0 Wm−<sup>2</sup> and a maximum of 511.6 Wm−2. Over the other stations, the NR differences are quite small; the difference in the daily averaged value was about 10 Wm−2, with a range from 102.8 to 112.2 Wm−2. The NR is also weakened/strengthened during the SASM active/break periods at most of the stations compared with the observational mean. As with DR, the largest effects of SASM on the NR occurred at Baingoin station, the daily averaged value of NR was 102.8 Wm−<sup>2</sup> during the observation period, while the NR was 66.9 Wm−<sup>2</sup> during the SASM active period, which

weakened by 34.9%, and the NR was 128.8 Wm−<sup>2</sup> during the SASM break period, which strengthened by 25.3%.

**Figure 5.** Diurnal variation of the downward short-wave radiation flux (DR) (units: Wm−2) from 8 stations, averaged for the observations, SASM active, and break periods.

**Figure 6.** Same as for Figure 5, but for the net radiation flux (NR) (units: Wm−2).

*3.3. The Impacts of SASM Evolution on the Turbulent Heat Transfers*

Figure 7 shows the diurnal variations in sensible heat flux (SH), averaged for the observations, SASM active, and break periods. Driven by the net heating (see Figure 6), the plateau releases heat into the atmosphere during the daytime (positive SH values) and receives heat from the atmosphere at night (negative SH values). Following the diurnal

variation in radiation, the sensible heat flux increases from early morning at approximately 6:00 LST, reaches a maximum at noon, and decreases later in the day. Differences are clearly found in the SH averaged and maximum values among the 8 plateau stations, despite the similar diurnal variations. During the observation period, the largest sensible heat transfer occurs over Ali station (northwest plateau), with daily averaged and maximum values of 60.1 (see Table 2) and 197.4 Wm−2, respectively. The second-largest SH is found over Namco station (southeast plateau), with daily averaged and maximum values of 28.2 and 118.9 Wm−2, respectively. Over the other stations, the diurnally averaged SH varies from 20.0 to 26.5 Wm−2, and the daily maximum values vary from 73.1 to 98.2 Wm−2. The smallest SH occurs over the Biru station (central plateau), with averaged and maximum values of 18.8 and 77.5 Wm−2, respectively. Our results are consistent with recent study results. The recent research reveals that SH in the central TP in August is generally between 5 and 40 Wm−2, with an average of 18 Wm−2, while SH in the western TP is between 40 and 70 Wm−2, with an average of 56 Wm−2, these results are also smaller than that in the past [29,35]. Ye and Gao [1] estimated the July-August-mean intensity of SH is 60–80 Wm−<sup>2</sup> over the central TP and 150–190 Wm−<sup>2</sup> in the western TP, and Yang and Guo [36] estimated SH in July-August is 50–60 Wm−<sup>2</sup> in the central TP and 75–90 Wm−<sup>2</sup> in the western TP, remarkably larger compared to the new findings. This result indicates that SH has been possibly overestimated by the previous studies when calculating SH using the bulk transfer method, which is based on the larger values of the bulk transfer coefficient for sensible heat [28].

During the SASM active period, the response of land-to-atmosphere sensible heat transfer exhibited grea<sup>t</sup> differences among the 8 stations. For example, the daily averaged SH weakened at most stations, with the amplitude of weakening varying from −1.7% to −41.4%. The largest weakening occurred at Namco station (southeast plateau), and the SH was 16.5 Wm−<sup>2</sup> over the SASM active period, which was weakened by 41.4% from its daily averaged value of 28.2 Wm−<sup>2</sup> during the observation period. The smallest weakening happened at Biru station (central plateau), with a weakening of 1.7% from its daily averaged value. At stations Amdo and Nyainrong, the daily averaged SH values strengthened, with the amplitudes increasing by 12.3% and 13.2%, respectively. During the SASM break period, the SH largely strengthens over all stations, and with the strengths varying from 0.8% (at Ali) to 45.3% (at Nyainrong). Therefore, the sensible heat transfers over the plateau region can be affected by the SASM evolution, with the weakened/strengthened amplitudes over most stations during the SASM active/break periods, which are closely related to the weakened/strengthened radiation conditions [9,10,17,23,27,32]. However, these SASM impacts on the sensible heat transfer exhibit large inhomogeneity over the plateau regions. Overall, the more southerly stations received more SASM impacts. The larger SASM impacts on sensible heat transfers occurred at stations Namco, Baingoin, and Lhari, and the SH differences between the SASM active and break periods were greater than 38% of the daily averaged values. The SASM impacts also extended westward and northward to the Ali and Nagqu stations, with the SH differences between the SASM active and break periods reaching 21.4% and 34.6% of their daily averaged values, respectively. However, the SASM impact appeared to be negligible at Amdo station (south of Nagqu station), which complicated our results.

**Figure 7.** Diurnal variations of the sensible heat flux (SH) (units: Wm−2) from 8 stations, averaged for the observations and SASM active and break periods.


**Table 2.** Sensible heat flux (SH) (Wm−2) over the 8 plateau stations, averaged for the observations and SASM active and break periods. The bracketed values denote the SH percentage increase (decrease) of the daily averaged value, in which positive (negative) values mean increasing (decreasing).

Figure 8 presents the diurnal variation of latent heat flux (LH), averaged for the observations, SASM active, and break periods. Differing from the sensible heat flux, the latent heat over the TP is always transferred upwards (positive LH values) during the entire day. In addition, the amplitude of LH was much larger than that of SH over most of the plateau stations and that was consistent with previous results [37]. Obvious differences are seen in the LH daily averaged and maximum values among the 8 plateau stations. During the observation period, the largest latent heat transfer occurred at Nagqu station (central plateau), with daily averaged and maximum values of 74.7 (see Table 3) and 238.6 Wm−2, respectively. The smallest LH occurred at Ali station (northwest plateau) due to small amounts of precipitation there, with daily averaged and maximum values of 10.1 and 27.2 Wm−2, respectively. Over the other stations, the diurnally averaged LH varied from 53.0 to 74.4 Wm−2, and the maximum values varied from 152.8 to 245.7 Wm−2.

**Table 3.** Latent heat flux (LH) (Wm−2) over the 8 plateau stations, averaged for the observations and SASM active and break periods. The bracketed values denote the LH increasing (decreasing) percentage of the daily averaged value, in which the positive (negative) values mean increasing (decreasing).


Compared with the impacts of the SH, the SASM impacts on LH were relatively small and complicated. At stations Namco, Lhari, and Baingoin, while the SASM impacts on the SH were large, but the SASM impacts on LH were smaller, with the LH differences between the SASM active and break periods varying from 12.0% to 27.9% of their daily averaged values. The small impacts over these stations could be closely related to the high moisture conditions there. The SASM impacts could also extend to the north plateau, and Nagqu station with a large LH difference (27.8% of the daily averaged value) between the SASM active and break periods. However, the same as for SH, the SASM impacts on LH seem negligible over station Amdo. It should be noted that the SASM impact over Ali could be ignored due to the small LH value despite having the largest LH response amplitude of the daily averaged values.

The total heat transfer (TH) is defined as the sum of SH and LH. Figure 9 shows the diurnal variation in TH, averaged for the observations, SASM active, and break periods. Clear differences can be seen in the TH magnitudes among the 8 plateau stations. During the observation period, the largest TH occurred at Nagqu station (central plateau), with daily averaged and maximum values of 101.2 (see Table 4) and 336.7 Wm−2, respectively. The smallest TH occurred over Ali station (northwest plateau), with daily averaged and maximum values of 70.2 and 220.3 Wm−2, respectively. Over the other stations, the

diurnally averaged TH varied from 73.0 to 98.8 Wm−2, and the daily maximum values varied from 225.9 to 343.9 Wm−2.

**Figure 8.** Same as Figure 7, but for latent heat flux (LH) (units: Wm−2).

**Figure 9.** Same as Figure 7, but for total heat flux (TH) (units: Wm−2).

**Table 4.** Total heat flux (TH) (Wm−2) over the 8 plateau stations, averaged for the observations and SASM active and break periods. The bracketed values denote the TH increasing (decreasing) percentage of the daily averaged value, in which the positive (negative) values mean increasing (decreasing).


From Table 4, the SASM impacts on the TH can be clearly seen at stations Namco, Baingoin, Lhari, and Nagqu, with a weakened/strengthened magnitude during the SASM active/break period, and the TH differences in daily averaged value between the SASM active and break periods are with a range between 19.6% and 36.6%. The largest impacts occurred at Baingoin station, with the TH difference reaching 36.0% of the daily averaged value between the SASM active and break periods. In comparison, the SASM impacts on TH over the other plateau stations were quite small or even negligible.
