THE THIRD ATMOSPHERIC SCIENTIFIC EXPERIMENT FOR UNDERSTANDING THE EARTH–ATMOSPHERE COUPLED SYSTEM OVER THE TIBETAN PLATEAU AND ITS EFFECTS

in the tropics and midlatitudes of the North Pacific (e.g., Zhao and Chen 2001b; Liu et al. 2007; Zhao et al. 2007; Nan et al. 2009; Zhao et al. 2009; Zhou et al. 2009; Duan et al. 2012). Therefore, global weather and climate research would be incomplete without considering the significant role of the TP. Compared to other land regions in the world, observational data are scarce over the TP, owing to its high elevations, naturally harsh environmental conditions, and less-developed logistics. Thus, a few field experiments have been implemented in the datascarce areas. For instance, the first Qinghai–Xizang Plateau Meteorology Experiment (QXPMEX) was carried out from May to August 1979 (Tao et al. 1986). This experiment promoted, for the first time, T he Tibetan Plateau (TP), known as the “sensible heat pump” and the “atmospheric water tower,” modifies monsoon circulations and regional energy and water cycles over Asia (Wu and Zhang 1998; Zhao and Chen 2001a; Wu et al. 2007; Xu et al. 2008b; Zhou et al. 2009). Strong ascent over the TP may transport lower-tropospheric water vapor and anthropogenic pollutants into the upper troposphere– lower stratosphere (UT–LS), which exerts an influence on the regional ozone valley (Zhou et al. 1995; Liu et al. 2003; Bian et al. 2011) and the aerosol-layer enhancements near the tropopause (Tobo et al. 2007; Vernier et al. 2015). The TP also modulates largescale atmospheric circulations over the Northern Hemisphere and atmosphere–ocean interactions Ping Zhao, Xiangde Xu, Fei Chen, Xueliang guo, Xiangdong Zheng, liPing liu, Yang hong, Yueqing li, Zuo la, hao Peng, linZhi Zhong, Yaoming ma, Shihao Tang, Yimin liu, huiZhi liu, Yaohui li, qiang Zhang, ZeYong hu, Jihua Sun, ShengJun Zhang, liXin dong, heZhen Zhang, Yang Zhao, Xiaolu Yan, an Xiao, Wei Wan, Yu liu, Junming Chen, ge liu, YangZong ZhaXi, and XiuJi Zhou

T he Tibetan Plateau (TP), known as the "sensible heat pump" and the "atmospheric water tower," modifies monsoon circulations and regional energy and water cycles over Asia (Wu and Zhang 1998;Zhao and Chen 2001a;Wu et al. 2007;Xu et al. 2008b;Zhou et al. 2009). Strong ascent over the TP may transport lower-tropospheric water vapor and anthropogenic pollutants into the upper tropospherelower stratosphere ( UT-LS), which exerts an influence on the regional ozone valley (Zhou et al. 1995;Liu et al. 2003;Bian et al. 2011) and the aerosol-layer enhancements near the tropopause (Tobo et al. 2007;Vernier et al. 2015). The TP also modulates largescale atmospheric circulations over the Northern Hemisphere and atmosphere-ocean interactions systematic research on the diurnal and seasonal variations and spatial features of the surface heat budget, the structures and evolutions of atmospheric circulation systems over the TP, and their effects on global and Asian general circulations.
In the 1990s, a longer-term field experiment was conducted over the TP with the support of the Japanese Experiment on Asian Monsoon (JEXAM). It estimated the drag coefficient C d of surface momentum and the bulk transfer coefficient C h of surface sensible heat (SH) and revealed seasonal and interannual variations of the surface heat budget over the TP and their relationships with rainy seasons (Chen 1999;Zhao and Chen 2000a,b). Afterward, the Second Tibetan Plateau Atmospheric Scientific Experiment (TIPEX-II) was carried out from May to August 1998. Its results showed an imbalance phenomenon of the surface heat budget, strong mesoscale convection activities, and shear-line characteristics ). The Global Energy and Water Cycle Experiment (GEWEX) Asian Monsoon Experiment (GAME)/Tibet intensive observation conducted a plateau-scale automated weather station experiment and a mesoscale experiment of the land surface and planetary boundary layer (PBL) observations with one X-band Doppler radar at Naqu, China, from May to September 1998 (Wang 1999;Ueno et al. 2001). GAME/Tibet made progress in retrieving the land surface radiative budget, precipitation, and soil moisture from satellite remote sensing products and understanding the PBL structures, the convective rapid development, and the precipitating cloud characteristics (Wang 1999;Ueno et al. 2001;Uyeda et al. 2001;Choi et al. 2004).
Entering the twenty-first century, the Coordinated Enhanced Observing Period (CEOP) Asia-Australia Monsoon Project on the Tibetan Plateau (CAMP/ Tibet), and the Tibetan Observation and Research Platform (TORP) were implemented over the central northern TP during 2002-04 (Ma et al. 2006, 2008. Their research documented regional characteristics of land surface heat and CO 2 fluxes, turbulence, and the PBL (Ma et al. 2009). Under the support of the Japan International Cooperation Agency (JICA) project, a New Integrated Observational System over the Tibetan Plateau (NIOST) project (JICA/Tibet) was carried out during 2005-09 (Xu et al. 2008aZhang et al. 2012;Chen et al. 2011Chen et al. , 2013. It found diurnal variations of rainfall over the TP and effects of latent heat release on TP vortices, provided evidence of strong troposphere-stratosphere exchanges over the TP, improved the National Oceanic and Atmospheric Administration/National Centers for Environmental Prediction (NOAA/NCEP)-Oregon State University-Air Force Research Laboratory-NOAA/Office of Hydrology land surface model (Noah) on the basis of observational characteristics in the land surface energy balance, and revealed the importance of the deep PBL to the troposphere-stratosphere exchange over the TP. In the summer of 2011, an experiment of the TORP ground-based and airborne remote sensing observations was conducted over the central TP as part of the Global Change Program of China (Ma et al. 2014). This experiment found hydrothermal and momentum exchanges and moisture transports over the southeastern TP during the monsoon period, as well as land surface and atmospheric circulation The abstract for this article can be found in this issue, following the table of contents. DOI:10.1175/BAMS-D-16-0050.1 In final form 31 October 2017 ©2018 American Meteorological Society For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy.
This article is licensed under a Creative Commons Attribution 4.0 license.
variations against the background of global change. Moreover, the Tibetan Plateau: Formation- Climate-Ecosystems (TiP) program focused on a longer-term evolution of climate over the TP and its influence (Mosbrugger and Appel 2012).
To quantify uncertainties in satellite and model products of soil moisture and temperature, some regional-scale observation networks were established. For example, during , the CEOP-Asian-Monsoon System with Ground Satellite Image Data and Numerical Simulations (AEGIS) project monitored the land surface characteristics and analyzed their linkages with convection, precipitation, and Asian monsoons by satellites, the existing groundbased flux measurements, and stable isotopes in precipitation over the southeastern TP (www. ceop-aegis .net/doku.php). The Tibetan Plateau observatory of plateau-scale soil moisture and soil temperature (Tibet-Obs) built a Naqu-Maqu-Ngari regional network in cold semiarid, cold humid, cold arid climates ). A regional-scale Soil Moisture and Temperature Monitoring Network (SMTMN) was also built on the central Tibetan Plateau (Yang et al. 2013). These field observation networks increased the understanding of regional land surface hydrological processes and errors of satellite-derived soil moisture products. In addition, an ecohydrological experiment over the Heihe River basin in 2010 provided a test bed to confirm or falsify new ideas on ecohydrology and new hypotheses on scaling .
The aforementioned field experiments have made significant progress in promoting the scientific understanding of the Earth-atmosphere coupled system over the TP. However, the problem of scarce observations in the area is still not well solved, which has further hindered the better understanding of the local land-atmosphere coupled system and its effects. The answers to some key scientific questions are still unclear. For example, owing to the lack of plateauscale soil moisture and PBL observation networks, there is a wide divergence in estimates of C d and C h over the TP (e.g., Ye and Gao 1979;Zhao and Chen 2000b;Choi et al. 2004;Gao et al. 2015;Zhang et al. 2016), which directly results in uncertainties when estimating the heat intensity and its impacts. Because of the lack of direct observations of cloud microphysical and troposphere-stratosphere exchange processes over the TP, the cloud microphysical characteristics in the formation and development of clouds and precipitation, their interactions with atmospheric environments, the UT-LS atmospheric vertical structures, and their effects on ozone variations and aerosol-layer enhancements near the tropopause are not well understood. Moreover, because of the scarce radiosonde data over the western TP, it is not known how the local atmospheric circulation systems (especially synoptic-and mesoscale systems) develop and move from the west to the east. Thus, numerical weather and climate forecast models often have poor reliability when modeling weather and climate features over the TP, including soil moisture, surface heat fluxes, surface air temperature, rainfall, PBL structures, cloud amount, and stratospheric ozone (Wang 2011;Wu and Zhou 2011;Qiu et al. 2013;Hu et al. 2014;Zheng et al. 2014Zheng et al. , 2015aZheng et al. ,b,c, 2016Guo et al. 2015;Zhuo et al. 2016;Wan et al. 2017). These problems may also cause large uncertainties in reanalysis datasets and satellite products (such as air temperature, soil moisture, surface heat fluxes, and radiation) over the TP Wang et al. 2012;Zhu et al. 2012;Su et al. 2013;Zeng et al. 2016).
To promote Tibetan meteorological research, the Third Tibetan Plateau Atmospheric Scientific Experiment (TIPEX-III), to continue for 8-10 years, was initiated jointly by the China Meteorological Administration (CMA), the National Natural Scientific Foundation of China (NSFC), and the Chinese Academy of Sciences (CAS). A preliminary experiment was implemented in 2013, and TIPEX-III began formally in 2014.

OBJECTIVES.
The field observational objective of TIPEX-III is to constitute a three-dimensional (3D) observation system of the land surface, PBL, troposphere, and lower stratosphere over the TP. This system integrates ground-, air-, and space-based platforms based on the meteorological operational networks, the TIPEX-III network, the existing NIOST network, and the part of the TORP observation sites in China. The scientific objectives of TIPEX-III are to understand the surface heat budget, cloud microphysical characteristics, atmospheric water cycles, and troposphere-stratosphere exchange characteristics over the TP; to clarify the impacts of the changing Tibetan land-atmosphere coupled system on severe weather and climate events and atmospheric energy and water cycles; to improve the parameterization schemes of the land surface, PBL, cloud-precipitation, and troposphere-stratosphere exchange processes over the TP; and to enhance the skills of weather and climate forecast operations. The specific intensive field observational and scientific objectives of TIPEX-III are addressed as follows: 1) In the previous field experiments, the soil moisture and PBL networks were mainly located over the central and southeastern TP, not yet constituting a plateau-scale observation network. Meanwhile, the meteorological operational observation sites are still sparse from the western to central TP (Figs. 1a and 1b). To create plateauscale soil moisture and PBL tower observation networks, TIPEX-III will build new sites over the data-scarce central and western areas, which may increase the understanding of temporal and spatial variations of the land surface characteristics over the TP and their interactions with atmospheric circulations. Meanwhile, the regional-scale soil moisture and PBL networks over the TP will be used to understand mesoscale spatial differences in the surface heat budget over the complex topography and landscape and their effects on mesoscale systems. These intensive observations also allow an objective evaluation of numerical models, reanalysis data, and satelliteretrieval products.
2) The previous field experiments only utilized the X-band Doppler radar for probing large precipi-tating cloud particles, lacking direct observations of cloud microphysical features. TIPEX-III will carry out intensive observations over key areas with frequent cloud activities by integrating measurements of cloud radars, aircraft campaigns, and radiosondes, as well as regional-scale land surface and PBL observation networks. These data will help to understand the microphysical characteristics of cloud development and interactions between clouds, surface heating, and atmospheric environments and to improve parameterization schemes of cloud microphysical processes.
3) The previous field experiments also lacked plateau-scale observations of troposphere-stratosphere exchange processes. TIPEX-III will conduct intensive observation tasks for water vapor, aerosol, and ozone in the troposphere and lower stratosphere by balloonborne package instruments and ground-based remote sensing measurements. These data will help to clarify the UT-LS characteristics and the mechanisms for ozone variations and aerosol-layer enhancements near the tropopause; to validate satellite-retrieval profiles of water vapor, aerosol, and ozone, especially in the UT-LS; and to improve parameterization schemes of the lower-stratospheric physical and chemical processes over the TP. 4) Only a few of the field experiments (such as the QXPMEX and TIPEX-II) executed intensive tropospheric radiosonde observations over the data-scarce western TP for about one year, not constituting a longer-term plateau-scale radiosonde observation network. To constitute such an observation network from the western to eastern TP, TIPEX-III will build new radiosonde stations over the western TP (Fig. 1b). These stations will gradually become part of the CMA operational observation system. The intensive observational data can be used to monitor the evolution of atmospheric circulation systems and water cycles from the west to the east and may be applied in services of weather and climate forecasts. 5) The importance of the summer TP sensible heat pump and atmospheric water tower to downstream monsoon rainfall has been documented (Wu and Zhang 1998;Xu et al. 2008b). However, their maintaining mechanisms and effects on the local atmospheric circulation systems, as well as extreme weather and climate events over the downstream area and a larger area of the Northern Hemisphere, are still not clearly understood.
In particular, the dominant control of the South Asian monsoon by thermodynamic functions over the TP is controversial. Boos and Kuang (2010) believed that the uplift of a narrow orography of the Himalayas and adjacent mountain ranges, instead of surface heating over the TP, plays a more important role in the South Asian monsoon climate. TIPEX-III will deeply investigate mechanisms responsible for the effects of the TP thermodynamic functions on the local atmospheric circulation systems and water cycles and extreme weather and climate events in the downstream area, as well as the Northern Hemisphere, by multiple observational datasets and numerical simulations. TIPEX-III will also probe a way to enhance the skills of weather and climate forecasts.
Compared to the previous experiments, the unique feature of TIPEX-III is that it will greatly expand the surface and tropospheric meteorological operational observation networks over the traditionally data-scarce central and western parts of the TP and promote integrated observations of multiple physical processes from the land surface to the lower stratosphere. Ultimately, the implementation of these observation networks could increase the understanding of the Earth-atmosphere coupled system over the TP and its effects on extreme weather events and regional climate variability and could also improve weather and climate forecast operations.

EXPERIMENTAL AREA AND NETWORK CONFIGURATION.
Considering the tremendous heterogeneity in terrain elevations and land-cover types, the distributions of the existing CMA meteorological operational observation stations, and the logistical challenges of maintaining observation sites, TIPEX-III comprises a plateau-scale intensive observation network, as well as regional-scale dense networks (Fig. 1). With a particular focus on understanding the complex interactions between the land surface, PBL, and cloud microphysical processes over the TP, TIPEX-III also integrates intensive observations over a few areas with active convection using the regional-scale land surface and PBL observation networks, ground-based radars, and airborne campaigns.
Intensive observations of land surface and PBL characteristics. TIPEX-III has established a soil observation network consisting of 46 sites in the central and western TP (Fig. 1a). Consistent with the operational observations of the CMA, at each site, the measurement system measures soil water content ( Table 1). All of these sites have been operating since September 2015. Meanwhile, TIPEX-III has also built two regional-scale soil moisture and temperature observation networks over Geji (near Shiquanhe), China, of the western bare soil with few features, and Naqu, of the central alpine steppe (Table 1 and Fig. 1a). The regional network consists of 33 sites over Naqu (Fig. 1c), which began operating in August 2015, and 17 sites over Geji (not shown), which began operating in December 2016.
For the PBL observations over the TP, TIPEX-III includes a plateau-scale network and a regional-scale network. The former consists of 10 multilayer towers (with distance between towers of ~500 km) at Shiquanhe, Gaize, Naqu, Linzhou, Linzhi, Tuotuohe, Maqu, Litang, Dali, and Wenjiang, China ( Fig. 1b and Table 1), and helps to study general patterns of the surface heat fluxes and PBL structures over the TP. The regional-scale network is located within an area of 300 km × 200 km near Naqu, a main source region of mesoscale low pressure and convection systems over the TP. This network consists of six additional sites at Bange, Namucuo, Anduo, Nierong, Jiali, and Biru,

Plateau-scale network for soil moisture
The DZN3 automatic soil water observation instrument (with an accuracy of 2.5% volume water content; made by the China Huayun Group) has one datalogger and five sensors at each site. Five sensors are inserted into soil at depths of 10, 20, 30, 40, and 50 cm. The observational data are recorded every hour. 46 sets Regional-scale network for soil moisture and temperature The 5TM ECH 2 O soil moisture observation instrument (with an accuracy of 2% volume water content and 1°C; made by Decagon, United States) has one datalogger and five sensors at each site. Five sensors are inserted into soil at depths of 2, 5, 10, 20, and 30 cm for soil moisture and temperature measurement. The observational data are recorded every 10 min.

PBL tower measurement systems
One 20-m-high tower has one 020C wind direction sensor (with an accuracy of 3°; made by Met One, United States) at a height of 20 m, five 010C wind speed sensors (with an accuracy of 0.07 m s −1 ; made by Met One), five HMP155A air temperature T and humidity (RH) sensors (with an accuracy of 0.226° − 0.0028°C × T when −80° < T < 20°C and 0.055° + 0.0057°C × T when 20° < T < 60°C, and an accuracy of 1% when 0% < RH < 90% and 1.7% when 90% < RH < 100%; made by Campbell Scientific, United States) at heights of about 1.5, 2, 4, 10, and 20 m, and one Scientific Infrared (SI)-111 surface temperature sensor (with an accuracy of 0.1°C when −10° < T < 65°C and 0.3°C when −40° < T < 70°C; made by Campbell Scientific) at 0 cm. One 10-m-high tower has the same instruments as the 20-m-high tower but with one wind direction sensor at a height of 10 m and four wind speed, air temperature, and humidity sensors at heights of about 1.5, 2, 4, and 10 m.

sets
Surface radiation measurement at PBL site One component net radiometer (CNR)-4 radiation sensor (with an accuracy of 1%; made by Kipp and Zonen, Netherlands) measures downward and upward shortwave and longwave radiation at 1.5 m. The observational data are recorded every 10 s.

sets
Soil heat flux measurement system at PBL site One heat flux probe (HFP) 01 soil heat sensor (with an accuracy of 2.5% volume water content; made by Hukseflux, Netherlands) is inserted at 5 cm. The observational data are recorded every 10 s.

sets
Rainfall measurement at PBL site One TE525MM rain gauge (with an accuracy of 1%-2% h −1 when rainfall is 50 mm h −1 ; made by Campbell Scientific) automatically records rain intensity every 1 min.

sets
China, and contributes to integrated research on the high-resolution land surface and PBL processes over the central TP and their effects on mesoscale systems. These observations have been conducted at Shiquanhe, Namucuo, Naqu, Anduo, Linzhi, Litang, Dali, and Wenjiang from July 2014 to December 2016; at Bange, Biru, Jiali, and Nierong from July 2014 to March 2016; and at Linzhou from August 2015 to December 2016.
Intensive routine radiosonde observations of tropospheric atmospheric profiles. Using the Vaisala portable radiosonde systems (Table 2), TIPEX-III conducted the intensive routine radiosonde observations at Shiquanhe, Gaize, and Shenzha stations in the western TP (Fig. 1b) at 0800, 1400, and 2000 Beijing standard time (BST) each day from 8 July to 31 August 2014 (BST = UTC + 8 h). Meanwhile, routine automatic sounding systems (Table 2) were newly built at these three stations and have carried out intensive observations at 0800 and 2000 BST each day since November 2014. After assessment of their performance, these automatic sounding stations are to become part of the CMA operational observation systems, which will ultimately constitute a long-term plateau-scale sounding observation network. Moreover, TIPEX-III will also include intensive radiosonde observations at Gongshan (27.75°N, 98.67°E), Jinchuan (31.48°N, 102.07°E), and Litang (30°N, 100.27°E) stations on the southeastern slope of the TP (Fig. 1b), a key area for gauging water vapor transports from the Indian Ocean to East Asia.
Intensive observations of tropospheric cloud-precipitation physical characteristics. TIPEX-III combines  Measures echo intensity (1 dB), radial velocity (1 m s −1 ), and velocity spectrum width (0.1 m s −1 ) and has a spatial resolution of 15-30 m and a temporal resolution of 2-3 s. Measures echo intensity (1 dB), radial velocity (1 m s −1 ), velocity spectrum width (0.1 m s −1 ), differential depolarization factor (0.2 dB), specific differential phase (2°), and correlation coefficient (0.01) and has spatial resolutions of 0.3 km and 1° and a temporal resolution of 6 min. Measures brightness temperature (0.2°C) and has a temporal resolution of 2 min, which may be used to retrieve atmospheric liquid water content, atmospheric water vapor, and temperature profiles.
Measures cloud-base and PBL heights (5 m) and has a spatial resolution of 5 m and a temporal resolution of 16 s. 1 set ground-based radars (Table 3), aircraft campaigns (Table 4), and the CMA operational Doppler radar systems for measuring physical properties of cloud and precipitation in the central (Naqu) and southeastern TP (Linzhi and Daocheng) (Fig. 1b). A primary goal of these observations is to explore the cloud microphysical characteristics and their relationships with convective precipitation systems and atmospheric environments. In July and August 2014, a campaign using the ground-based radars and airborne instruments was conducted within a 200 km × 200 km area of Naqu (Fig. 1b). A follow-up field campaign using ground-based radars was carried out at Naqu and Linzhi from 15 July to 31 August 2015 and at Daocheng from 1 July to 31 August 2016.
Intensive observations of troposphere-lower-stratosphere ozone, aerosol, and water vapor profiles. Strong transports of air masses from the troposphere to the lower stratosphere appear over the southeastern TP and weaken toward the north and west (Tao et al. 1986;Cong et al. 2002). TIPEX-III includes plateau-scale intensive measurements for vertical profiles of ozone, aerosol, and water vapor at Shiquanhe, Lhasa, Linzhi, Tuotuohe, Mangya, Golmud, and Xining meteorological stations (Fig. 1b). Using balloonborne package instruments and ground-based remote sensing instruments (Table 5), two observational missions were separately implemented at Linzhi from 6 June to 31 July 2014 and at Shiquanhe, Lhasa, and Golmud during the period from May to September 2016.

PRELIMINARY ACHIEVEMENTS OF TIPEX-III.
The implementation of TIPEX-III has enhanced the monitoring capability for the land surface, PBL, troposphere, and lower stratosphere over the TP. It has also promoted the understanding of their features, physical processes, and effects and improved the capability of weather and climate models. Noticeable progress has been achieved in research and data applications.

The new TIPEX-III data revealed a possible overestimation of the bulk transfer coefficient of sensible heat over the TP in previous studies, a larger plateau-scale heterogeneity in latent heat flux than in sensible heat flux, and the linkages of surface heat fluxes to Asian monsoon activities.
The plateau-scale heating contrasts exert a significant effect on the development of mesoscale convection and circulation systems over the eastern TP (Sugimoto and Ueno 2010). However, previous estimates of C d , C h , and surface heat fluxes show large uncertainties over the TP. The new data allowed the plateau-scale differences in C d , C h , and surface heat fluxes to be reexamined. It was found that a new estimate of SH is 18 W m −2 in the central TP (with a range between 5 and 40 W m −2 ) and 56 W m −2 at the western TP (with a variation between 40 and 70 W m −2 ) (Fig. 2a). The spatial difference in SH is larger on the plateau scale (between the west and central parts) than on the regional scale (in the Naqu regional network), and surface latent heat flux (LH) has a larger plateau-scale difference than SH (Fig. 2b).
The new estimate of C h (from 2 × 10 −3 to 4 × 10 −3 ) shows the smaller plateau-scale heterogeneity relative to C d (from 3 × 10 −3 to 11 × 10 −3 ) ( Table 6) and is smaller than previous estimates (>4 × 10 −3 ) (e.g., Ye and Gao 1979;Yang et al. 2011). Based on the larger values of C h , the estimated July-August-mean intensity of SH is 60-80 (150-190) W m −2 over the central (western) TP by Ye and Gao (1979) and 50-60 (75-90) W m −2 by Yang and Guo (2011), remarkably larger compared to the new estimate. This result indicates that SH has been possibly overestimated by the previous studies when calculating SH using the bulk transfer method, which could lead to an incorrect understanding of the TP SH intensity. Therefore, the role of summer SH over the TP in local thermal convective formation and development, and the uncertainties in model land surface processes, must be reestimated. Moreover, it was also found that in the numerical forecast models, the overestimated effect of summer SH over the TP may be reduced by improving the parameterization scheme of land surface processes for bare soil . The data diagnosis of TIPEX-III further revealed a plateau-scale difference in diurnal variations of surface heat fluxes over the TP and their linkages with the South Asian monsoon. SH shows a larger diurnal variation in the west than in the middle and east, but LH does not show remarkable diurnal variations . When the strong warm moist southerly wind in front of the South Asian monsoon trough prevails over the southeastern TP, the diurnal variations of local SH and downward shortwave radiation are weaker (Li et al. 2016), which suggests the feedback of the Asian monsoon on the TP heat source. Previous studies paid more attention to the impacts of the TP heating on the Asian monsoon (e.g., Wu and Zhang 1998;Zhao and Chen 2001a,b;Liu et al. 2007;Zhou et al. 2009;Duan et al. 2012) or the effects of extratropical atmospheric planetaryscale waves over the Northern Hemisphere on the TP heating Cui et al. 2015). Thus, this feedback of the Asian monsoon provides a new insight for understanding reasons for the TP heating variations. Consists of a CPI with a resolution of 3.2 µm and 2DS probe with a measurement range between 10 and 1,280 µm and an imaging frequency of 400 frames per second.

set
The new TIPEX-III observations uncovered the characteristics of cloud, its radiative effect, and raindrop size distribution and the importance of warm rain processes in the formation and development of cloud and precipitation. Cloud diurnal variaTion and Warm rain ProCeSS. The lack of direct observations for cloud physical processes hinders the understanding of the cloud microphysical characteristics over the TP and their roles in local cloud development. The TIPEX-III intensive observations revealed the diurnal variation of cloud over the TP. It was found that convective cloud and precipitation exhibits a distinct diurnal variation over the central TP. The strong convective precipitation begins at noon (Chang and Guo 2016), corresponding well to the peak of the local sensible heat flux , which implies an influence of surface heating on the diurnal variation of convection and precipitation. Convective precipitation then gradually turns to stratiform precipitation. For the stratiform cloud, the dominant cloud particles are raindrop-sized supercooled water with fewer ice particles (Fig. 3a), which indicate a warm rain process. This process can generate heavier precipitation over the precipitation centers of weak convection systems than the cold rain processes .
CharaCTeriSTiCS oF raindroP SiZe diSTribuTion. The new observational results indicate that the raindrop size distribution (RSD) over the TP is wider during the day than at night, with the widest RSD in the late afternoon (Chang and Guo 2016). Moreover, the RSD is wider over the TP compared to heavy rainfall over the downstream plains, and the concentration of cloud droplets is much lower compared to over clean Measures profiles of pressure, temperature, relative humidity, wind, and ozone concentration (with an accuracy of 5%-10% up to 30-km altitude). The observational data are recorded every 2 s.

Disposable consumables
Cryogenic frost hydrometer (CFH) (Droplet Measurement Technologies) Measures profiles of water vapor from the surface to about 28 km, with an uncertainty of <4% in the troposphere and <9% in the stratosphere. The iMet-1-RSB radiosonde is used as a platform of data transmission, and the observational data are recorded every 1.2 s. Measures total column ozone and solar spectral UVB, with an uncertainty of <1% for ozone and an accuracy of 5%-10% for spectral UVB irradiance. The observational data are recorded every 5 min for ozone and every 8 min for solar spectral UVB (with a given solar zenith angle). oceans. The RSD varies between 10 2 and 10 3 m −3 mm −3 over the TP when the precipitation particle radius is <1 mm (Fig. 3b) and shows a Γ distribution when the radius is <2 mm. The larger raindrops (with a size of 10 µm) in the shallow convection over the TP (Fig. 3c) may enhance collision-coalescence processes, producing light rain, even though the concentration of larger raindrops is relatively low. This phenomenon is quite different from the plain of China where light rain is often suppressed by high aerosol loading. Because the atmospheric environment is relatively clean over the TP, the study of the formation and development of local cloud and precipitation could help to improve understanding of the effects of aerosols in a polluted atmospheric environment.

sets
verTiCal STruCTureS oF Cloud radiaTive eFFeCT. The TIPEX-III analysis with the CloudSat-CALIPSO products indicated a salient difference in the vertical structures of both shortwave and longwave cloud radiative effects (CREs) over the TP from the adjacent regions. This difference is characterized by the deeper shortwave CRE heating and longwave CRE cooling layers in the troposphere and the maximum values of the CRE heating and cooling in the lower layers over the TP (Yan et al. 2016). A strong cooling layer of net CRE appears in the upper troposphere, and a shallow but strong heating layer appears in the lower layer. However, more general and precise information related to their full diurnal cycles and averages is needed to combine with the geostationary satellites and ground-based observations.
The TIPEX-III observations and analyses revealed the contributions of regional photochemistry and long-range ozone transport to the lower-tropospheric ozone and the relative importance of heterogeneous chemical processes near the tropopause and convective transports to the UT-LS ozone. Because of a lack of direct observational evidence for contributions of regional photochemistry reactions and horizontal and vertical transports to tropospheric and lower-stratospheric ozone variations, the intensive observations of TIPEX-III helped to improve understanding of the contributions of these processes. It was found that, over the southeastern TP, the ozone concentration is lower in the middle and lower troposphere compared to the South Asian monsoon region, such as New Delhi, India (Fig. 4b). This lower-tropospheric low ozone over the TP could be mainly attributed to the lower anthropogenic pollution emissions through photochemical ozone production in the area. Over the northeastern TP, the lower-tropospheric ozone maximum could be attributed to both the regional photochemistry processes and the long-range ozone transport from East Asia, Europe, and Africa during the summer (Zhu et al. 2016). Especially in June, the contribution of the regional photochemistry process is almost half of that of the horizontal transport.
Previous studies did not explicitly indicate the relative importance of upward transport from the lower-tropospheric low ozone concentration or in situ photochemical processes caused by anthropogenic pollution emissions to the UT-LS ozone valley during the summer. The intensive profile observations of the TIPEX-III campaign over the southeastern TP revealed a mean temperature of >−78.15°C (the threshold of the highest temperature required for the formation of polar stratospheric clouds) and a low water vapor concentration near the tropopause (Fig. 4a), which indicates a dehydration process induced by a cold temperature trap. In such an environment, phenomena such as polar stratospheric clouds are not likely to occur over the TP, and the heterogeneous chemical reaction of depleting ozone is weak over the TP. Thus, the in situ heterogeneous chemical processes near the tropopause might not be a dominant mechanism responsible for the formation of the UT-LS ozone valley during the summer.
Another mechanism, strong convective transports of the lower-tropospheric low ozone concentration toward the UT-LS, may have had a greater influence. Over the southeastern TP, the upward transports are closely associated with the strongest ascending motion. Under the influence of the plateau surface thermodynamic forcing, the upward transports are stronger in the daytime and summer. The tropopause fold is also a favorable structure for cross-tropopause exchanges. It is known as a vertical intrusion of the dynamical tropopause into the troposphere and is accompanied by the normal tropopause break and two tropopause heights (i.e., the polar tropopause height and the tropical tropopause height). The intensive observations of TIPEX-III revealed that the late rainy season is an important period for tropospherestratosphere exchanges over the TP (Hong et al. 2016). In this season, the tropopause fold is frequently observed over the western TP. The polar tropopause occurs during the entire rainy season, but its height decreases from the early to late rainy season (Fig. 4c); while the tropical tropopause is mainly observed in the late rainy season (Fig. 4d), which is possibly associated with weaker upper-tropospheric jet streams over the western TP. The frequent occurrence of the  Table 4). tropopause fold favors the cross-tropopause exchange of air masses with different ozone concentrations between the troposphere and stratosphere. Meanwhile, the lower-tropospheric pollutants over the TP and its surrounding areas may have also been transported to the UT-LS, which could further affect the chemical processes of the UT-LS, as well as ozone concentrations. Thus, the effects of the increasing aerosol concentration near the tropopause over South Asia on chemical processes and ozone should be closely examined.
The TIPEX-III analyses provided a new insight into the effects of the TP thermodynamic functions on local atmospheric water maintenance and vortex movement, downstream rainfall and haze events, and Northern Hemispheric continent temperature and rainfall. The physical diagnoses and numerical simulations of TIPEX-III gave a maintenance mechanism responsible for the TP atmospheric water tower. When the TP surface heating draws the warm moist air masses from the tropical Indian Ocean toward the plateau, they travel along the southern slope of the TP and form two ladders of the conditional instability of the second kind (CISK) processes with two couplings of apparent heat source-moisture sink over the TP southern slope and main platform, respectively. This feature enforces the convergence of low-level low pressure, and convection systems over the platform and favors the formation and development of local cloud and precipitation (Fig. 5a). In this way, the TP atmospheric water tower is maintained (Xu et al. 2014).
The analyses of TIPEX-III also revealed new influence mechanisms for downstream extreme weather and climate events. It was found that when the extratropical westerlies climb over the TP, they descend on the lee side of the TP, which favors a near-surface "harbor" with weak wind over central eastern China, followed by the accumulation of local air pollutants and the occurrence of extreme haze events (Xu et al. 2016). Moreover, the atmospheric heating intensity over the TP affects the movement of low-level vortices from the central to eastern TP ). In the developing stages of these vortices, the vertical structure of the heat source may determine their intensity and movement direction. The vortices moving away from the plateau usually trigger heavy rainfall to the east of the TP and even give rise to disastrous rainfall events over downstream areas.
These results provide valuable information about the skills of downstream weather and climate forecasts. For example, improving the land surface and PBL processes over the TP in numerical forecast models can reduce an overestimated SH effect and a cold bias in the land surface temperature over the TP, so the models can better capture the characteristics of precipitation over central eastern China and the coasts . Moreover, TIPEX-III also uncovered that when the new intensive radiosonde data at Shiquanhe, Gaize, and Shenzha stations of the western TP are assimilated in the mesoscale Weather Research and Forecasting (WRF) Model system with three-dimensional variational data assimilation, the forecast skill of rainfall in the TP and adjacent areas is remarkably enhanced, with a decrease in the rootmean-square error of the 24-h forecast rainfall by 11% over the TP (Fig. 5b).
TIPEX-III demonstrated an important modulation of the TP heating on large-scale atmospheric waves, teleconnections, and climates. Surface sensible heat and atmospheric latent heating over the TP could enhance the meridional circulation of the Asian summer monsoon, produce eastward-propagating Rossby waves along the extratropical westerlies, and modify largescale climates over the Northern Hemisphere (Wu et al. 2016). It was also discovered that during summer, a strong surface heating over the TP could trigger a Northern Hemispheric extratropical teleconnection like the Asia-Pacific Oscillation (APO) (Liu et al. , 2017, with an increased tropospheric temperature and strengthened ascent over Asia and a decreased tropospheric temperature and strengthened descent over the central eastern North Pacific (Fig. 5c). It has been known that associated with APO are significant anomalies in lows over the Asian-African monsoon region and subtropical anticyclones over the North Pacific and Atlantic, as well as surface air temperature and rainfall over Africa, South Asia, East Asia, and extratropical North America (Zhao et al. 2012). TIPEX-III further revealed that these APO-related anomalies in atmospheric circulation, rainfall, and temperature could be also forced by an individual surface heating change over the TP (Figs. 5d-f), which suggests the importance of the TP forcing to global-scale climate anomalies, which should be further explored.

SUMMARY AND DISCUSSION.
Meteorological observational data are scarce over the TP. Although a few field experiments over the TP have made significant progress, they have mostly focused on land surface and PBL observations. Routine surface and tropospheric observation stations are still sparse over the western TP. There is a lack of direct observations of cloud microphysical and troposphere-stratosphere exchange characteristics. All of these hinder the better understanding of the TP Earth-atmosphere coupled system and its effects.
To promote Tibetan meteorological research, in 2013 the CMA, NSFC, and CAS jointly initiated TIPEX-III for a duration of 8-10 years. It has been implemented as an integrated observation of the land surface, PBL, troposphere, and lower stratosphere by coordinating ground-, air-, and space-based facilities from the CMA operational networks and previous scientific experiment observation networks in China. The implementation of TIPEX-III has greatly expanded the CMA operational observation networks and provided important support for the NSFC comprehensive research program during 2013-22 entitled "The Earth-atmosphere coupled system over the Tibetan Plateau and its global climate effects." TIPEX-III has achieved noticeable progress in research and data application. Its preliminary achievements provide a special reference to advance the understanding of the plateau land-atmosphere coupled system, as well as new insight into how to improve model physical parameterization schemes and enhance the skills of weather and climate forecasts. In the future, TIPEX-III will include further field experiments and comprehensively study interactions between tropospheric cloud-precipitation physical processes, surface heating, and atmospheric environments over the TP. It will involve further examination of the vegetated surface aerodynamic conductance in land surface models and the importance of the heat source intensity over the TP to monsoon onsets and global and regional extreme weather and climate events; improve parameterization schemes of the land surface, PBL, cloud-precipitation, and troposphere-stratosphere exchange processes over the TP; and validate assimilated and remote sensing products of soil temperature and moisture and atmospheric ozone, aerosol, and water vapor.
Equally importantly, TIPEX-III aims to promote scientific exchanges and collaborations with international research communities and broader organizations. Coordinated observational experiments in the countries neighboring the TP are especially encouraged. Scientists from international communities are invited to participate in the follow-up field campaigns and to use the TIPEX-III data in their research. Validated TIPEX-III datasets will be open to the domestic and international scientific communities after a data-protection period of one year (starting from the dates of completion of the data quality control or the product generation). A dedicated data archival center of the CMA, the National Meteorological Information Center (NMIC), is responsible for collecting the raw and processed datasets and distributing them to users. The website for downloading the TIPEX-III data (http://data.cma.cn/tipex) has been preliminarily built and will be further improved. Users are also encouraged to directly contact the corresponding authors of this paper to obtain data and further details.