Balloon-borne cryogenic frost-point hygrometer (CFH) observations conducted over three stations Trivandrum, Hyderabad and Kolkata in the Indian region during the period 2014–2017 are used to study the influence of deep convection and monsoon dynamics on the distribution of water vapour in the upper troposphere and lower stratosphere (UTLS) and to quantify the amount of water vapour transported to the lower stratosphere (LS) during summer-monsoon (June to September). In summer monsoon season, CFH observations show a water vapour enhancement of ~40–250% in the upper troposphere (UT) and 0.5–1 ppmv (10–40%) in the LS compared to pre-monsoon. In this season, the spatial pattern of Microwave Limb Sounder (MLS) derived water vapour mixing ratio (WVMR) at 100 hPa peaks north of the deep convection core over the head Bay of Bengal (BoB). In the LS, the water vapour maximum shows a south westward shift with altitude in accordance with the anticyclonic circulation flow. The maximum altitude of deep convective cloud top increases north eastward in accordance with the peak in tropical easterly jet (TEJ) causing the altitudinal shift in water vapour enhancement in the UT. The probability of dehydration (WVMR<3 ppmv) maximizes west of the deep convective core over Head BoB. During summer-monsoon, the day-to-day variability of water vapour in the tropical tropopause layer (TTL) is mainly controlled by the dehydration of air due to freeze drying (temperature dependent) and rehydration due to recurrent overshooting deep convections. Direct injection of water vapour into the LS region is probable when the tropopause is relatively warm (>192 K).

Altitude structure of turbulence in the troposphere and lower stratosphere (TLs) over the Indian peninsula is delineated using radiosonde observations carried out from six stations (August 2013 to December 2017) as part of the Tropical Tropopause Dynamics (TTD) Campaign under the GPS Aided Radiosonde Network Experiment for Troposphere-stratosphere Studies (GARNETS) program. Thorpe analysis applied to the potential temperature profiles, taking into account the impact of atmospheric moisture and instrumental noise, is used to estimate the turbulence parameters. This study shows that while the occurrence of turbulence is high in the lower (0-2 km) and upper (10-15 km) troposphere, the region 3-8 km is relatively devoid of turbulence at all the six stations. In general, similar to 60% of the Thorpe scale (L-T) values in the troposphere are less than 250 m at all the stations, except at Cochin where it extends up to 400 m. Though the altitudinal structure of turbulence shows large variability from station-to-station, it does not show any systematic latitudinal pattern. This study also shows that the Thorpe method appears to be good in identifying convectively induced turbulence than the dynamically (shear) induced turbulence. In the lower troposphere, the turbulence is due to both convective and dynamic instability at all the stations. While turbulence in the altitude region between 10 and 15 km is mainly due to convective instability at Coimbatore, Gadanki and Goa and due to dynamic instability at Cochin, both these instabilities contribute significantly to the generation of turbulence at Trivandrum and Hyderabad. Occurrence of convective instability (turbulence) under clear-air conditions in the upper troposphere, suggests wave breaking as a mechanism.

Diurnal variation of atmospheric water vapour in the troposphere over two tropical stations Trivandrum (8.5 degrees N,76.9 degrees E) and Gadanki (13.5 degrees N, 79.2 degrees E) are studied using radiosonde observations for three years (December 2010-March 2014) carried out as part of Tropical Tropopause dynamics (TTD) campaigns under CAWSES (Climate And Weather of the Sun-Earth System) India Phase II program. Trivandrum is a coastal station and Gadanki an inland station in the Indian Peninsula. Even though the absolute value of columnar integrated water vapour content (IWV) is higher over Trivandrum in all the seasons, the amplitude of its diurnal variation is almost the same (similar to +/- 5-8 kg m(-2)) at both the stations with lower values during the day and higher values during night. Though the absolute humidity decreases exponentially with altitude, the amplitude of normalized diurnal anomaly is more-or-less the same (15-20%) from the surface up to 10 km in all the seasons. There exists a time difference in the peaking hours of water vapour density between the two stations with Gadanki always leading Trivandrum. While the water vapour density maximizes around midnight over Trivandrum, it peaks in the late evening over Gadanki. Harmonic analysis shows that the diurnal and semi-diurnal components together account for almost all the sub-daily variations in IWV and absolute humidity. The contribution of semi-diurnal component is only 25% of that of the diurnal component. Diurnal variation of surface wind speed and direction together with the vertical motion associated with convection regulated by the diurnal variation in temperature seems to play the lead role in causing the diurnal variation in water vapour. While the diurnal variation over Gadanki is mainly controlled by the convectively induced vertical motion, the diurnal variation in the rate of evaporation due to surface wind and local circulation also plays a significant role over Trivandrum.

Influence of convection on the thermal structure of Troposphere and Lower Stratosphere (TLS) is investigated using radiosonde data, obtained from Trivandrum (8.5 degrees N, 76.9 degrees E), Gadanki (13.5 degrees N, 79.2 E), Bhubaneswar (20.25 degrees N, 85.83 degrees E), Kolkata (22.65 degrees N, 88.45 degrees E) and Singapore (1.37 degrees N, 103.98 degrees E), collected during different convective categories classified based on the altitude of deep convective cloud tops (CT) in the period 2008-2014. During deep convective events, the temperature showed lower tropospheric cooling, an upper tropospheric warming and an anomalous cooling (warming) below (above) the cold point tropopause (CPT) with respect to the clear-sky value. While warming in the upper troposphere is strongest (similar to 2-4 K) around 10-12 km, anomalous cooling (warming) below (above) the CPT is maximum around 15.5 km (17.51cm) with values in the range of -2 to -4 K (3-6 K). These temperature perturbations are observed 5-6 days prior to the convective events. In response to deep convection, surface cooling up to similar to -4 K is also observed. This study showed that the magnitude of cold and warm anomalies increases with strength of convection. During deep convection, the potential temperature (theta) shows a decrease (<5K) in the tropical tropopause layer (TTL) from the TTL-base up to CPT compared to that on clear-sky days, confirming the vertical mixing of convective air from the lower atmosphere to the TTL-levels. Correlation analysis between different TTL parameters suggests that, as the cloud top altitude increases, along with the adiabatic process, diabatic process also plays a major role in the TTL. An interesting feature observed during deep convection is the ascent of TTL-base by similar to 1.5 km and descent of CPT and TTL-top by 0.5 km, which effectively thins the TTL by similar to 2 km.

Seasonal and diurnal variations in the thermal structure of the troposphere and lower stratosphere with special reference to the tropical tropopause layer (TTL) are studied using data from 3h interval radiosonde launches carried out simultaneously from Trivandrum (8.5 degrees N, 76.9 degrees E) and Gadanki (13.5 degrees N,79 degrees E) during December 2010 to March 2014. TTL is defined as the region extending from the level of minimum stability (LMinS) to the level of maximum stability (LMaxS). Above the cold point tropopause (CPT), temperature showed warm anomaly (4-5K) during summer monsoon (June-September) and cold anomaly (up to -4K) during winter (December-February). The temperature in the troposphere showed a clear diurnal variation (0.5K) with a cold anomaly during early morning and warm anomaly during the day in all the seasons. At Gadanki, the diurnal temperature anomaly in the lower stratosphere showed its phase propagating downward with time, whereas at Trivandrum such variations are not clearly evident. At diurnal timescales, the TTL showed significant variations at LMinS (0.5-1.5km/5-7K) and smaller variations at CPT (0.2-0.5km/2-3K). In general, the amplitude of the diurnal component is greater than the semidiurnal component for all the TTL parameters, and these amplitudes are relatively larger at Trivandrum than at Gadanki. The observed diurnal variations could be the manifestation of tidal oscillations and/or due to the influence of local convection. Correlation analysis between different TTL parameters indicated a slow transition from a governing adiabatic process in the TTL base to a diabatic process at the TTL top.

The structure of tropical tropopause layer (TTL) is delineated based on static stability criteria. It is defined as the region extending from the level of minimum static stability to the level of maximum static stability. The static stability and dynamic instability steadily increase from the base to the top of the TTL. Radiosonde data from three tropical stations and Constellation Observing System for Meteorology, Ionosphere, and Climate GPS radio occultation (COSMIC) data over the tropics are used for this investigation. This study suggests that in terms of stability, the TTL can further be considered as a composite of three sublayers: a bottom layer (BL), a middle layer (ML), and an upper layer (UL) embedded between the potential temperatures similar to 350-360 K, similar to 360-380 K, and 380-420 K, respectively. While the BL is significantly influenced by frequent convective outflows, the influence of convection declines across the ML. Ozone and water vapor mixing ratios show a discernible change in their gradients across these sublayers. Occurrences of small-scale turbulence and cirrus are maximum in the BL and decrease in the ML and UL. Longitudinally, the BL is broad and the ML and UL are narrow over the deep convective cores. Strength of static stability in the UL (BL) is relatively strong (weak) centered about the equator, with pronounced features over the deep convective regions. These features of static stability centered about the equator in the BL could be attributed to the convective outflows in the troposphere and that in the UL to the dynamic and radiative processes in the upper troposphere and lower stratosphere.

The effect of deep convection on the thermal structure and dynamics of the tropical tropopause at sub daily scales is investigated using data from radiosondes launched over two sites in the Indian Monsoon region (Gadanki (13.5°N, 79.2°E) and Trivandrum (8.5°N, 76.9°E)) conducted between December 2010 and March 2014. The data from these soundings are classified into 5 convective categories based on the past, present and future cloudiness over the launching region after the radiosonde has reached tropopause altitude. They are denoted as category 1 (no convection), category 2 (convection may occur in any of the next 3h), category 3 (convection occurred prior 3h), category 4 (convection terminated within 3h of launching) and category 5 (convection persistent throughout the considered period). The anomalies from the background in temperature, relative humidity and wind speed are grouped into the aforementioned five different convective categories for both the stations. Cooling and moisture anomalies are found during the active convection (category 5). The horizontal wind speed showed a strong anomaly indicating the presence of synoptic scale features. Vertical wind obtained simultaneously from the MST radar over Gadanki clearly showed strong updraft during the active convection. The ozone profiles from ozonesondes launched during the same period are also segregated according to the above convective categories. During the active convection, high and low ozone values are found in the upper troposphere and the lower troposphere, respectively. The cross tropopause ozone mass flux and vertical wind at the tropopause and convective outflow level estimated from the ozonesonde, and MST radar/ERA-Interim data showed positive values indicating the transport of ozone between troposphere and stratosphere during deep convection. Similarly, the total mass flux crossing the cold point tropopause over Gadanki showed upward flux during the active convection. The variability of the cross tropopause mass flux is found to be higher over Gadanki compared to Trivandrum.

Diurnal variability of atmospheric stability as well as the occurrence and strength of turbulence in the troposphere and lower stratosphere at two tropical stations, Trivandrum (8.5°N, 76.9°E) and Gadanki (13.5°N, 79.2°E), situated in the Indian Peninsula is studied. For the analysis three years of GPS-radiosonde data, collected as a part of the Tropical Tropopause Dynamics (TTD) Experiment under the CAWSES-India program, has been used. Thorpe method is adopted to estimate the turbulent parameters from radiosonde observations. This study shows that in the atmospheric boundary layer, both stability and turbulence parameters depict a pronounced diurnal variation. Over Trivandrum, the occurrence of turbulence as well as its strength peaks during night and falls off during the day, while at Gadanki it shows an opposite pattern. At both the stations, in the 3–10km altitude layer, the occurrence and strength of turbulence are relatively high during night compared to day. Although the turbulence strength in the 10–15km altitude layer is rather weak at both the stations, the occurrence of turbulence shows a clear diurnal pattern (high during the day), especially over Trivandrum. In the 3–15km altitude layer, while the occurrence of convective instability is fairly the same at both the stations, the wind shear is significantly large at Trivandrum compared to Gadanki, with high values during night compared to the day. This shows that in this altitude region, while convective instability is mainly responsible for the generation of turbulence at Gadanki, wind shear induced dynamic instability is also responsible for the generation of turbulence at Trivandrum especially during night. Above 15km, where wind shear driven instability leads the convective instability, turbulence at both the stations does not show any significant diurnal variability.

Characteristics of turbulence in the troposphere and lower stratosphere at Trivandrum (8.5°N, 76.9°E) and Gadanki (13.5°N, 79.2°E), two tropical stations located in the Indian Peninsula, are studied using GPS-radiosonde observations during the period of December 2010 to March 2014 as part of the Tropical Tropopause Dynamics (TTD) Experiment under the CAWSES-India program. This study relies on the detection of turbulence applying Thorpe analysis to the temperature profile, taking into account the impact of atmospheric moisture and instrumental noise on static stability. In general, the tropospheric turbulence is largely intermittent in space and time. The altitude region very close to the convective tropopause (COT), 10–15km, is relatively more turbulent than the lower troposphere from 3 to 8km. Though the occurrence of turbulence decreases significantly above the COT, occasionally a rather thin layer of turbulence (thickness <1km) is observed in the tropical tropopause layer (TTL) very close to the cold point tropopause (CPT). Even though broad turbulent layers, with thickness >2km, are the persisting features that can be observed in the 5–15km altitude region in multiple observations at both the sites at least during Asian Summer Monsoon (ASM) season, prominent multiple thin layers of stratified turbulence in the lower troposphere lasting for a day or less are observed only at Trivandrum in all seasons. In general, the turbulence strength in the 5–15km altitude region at Gadanki is generally larger than that at Trivandrum. Below 15km, while the turbulence is mainly governed by the convective instability at Gadanki, wind-shear driven (dynamic) instability also contributes considerably for the generation of turbulence at Trivandrum. While the generation of turbulence above 15km is dominated by dynamic instability, in the lower stratosphere (LS) it is mainly due to strong wind shears.

The role of deep convection on the tropical tropopause parameters at sub-daily scales using radiosonde observations at two locations in South-India affected by monsoon has been investigated. Special experiments were conducted under the Tropical Tropopause Dynamics (TTD) campaigns from two stations, (Gadanki (13.5°N, 79.2°E) and Trivandrum (8.5°N, 76.9°E) as a part of CAWSES India Phase-II programme during December 2010 to September 2013. In addition, data from regular radiosonde launches available from April 2006 to September 2013 are also utilized in the present study. Using satellite based infrared brightness temperature data, convection is classified into six categories based on the life cycle of the convection prevailing 3h before and after the balloon reaching the tropical tropopause. Cold-point and lapse rate tropopause altitudes (CPH, LRH) and temperatures (CPT, LRT), convective outflow level (COH) and tropical tropopause layer (TTL) thickness extracted from individual soundings are grouped into six convection categories. Large amount of water vapour with diabatic cooling prevailed near the CPH during active convection leading to STE processes. At the same time, decrease in TTL thickness is observed not only because of pushing up of the COH but also due to decrease of CPH. On an annual basis a decrease (increase) in CPH and LRH (CPT and LRT) is noticed during active convection. This feature is more significant at Gadanki compared to Trivandrum. During the monsoon and pre-monsoon periods when the convection is rather widespread, CPH (CPT) shows a decrease (increase) at Gadanki while it increases (decreases) over Trivandrum. Large seasonal variation is noticed in the tropopause parameters even when they are segregated into different convective categories mainly due to intensity of the convection being different. During active convection, diabatic and adiabatic processes seem to be dominant at Gadanki and Trivandrum, respectively.

The structure and variability of tropical tropopause over Gadanki (13.5°N, 79.2°E) are delineated using data obtained from Indian MST radar operated in the vertical mode as a part of intense Tropical Tropopause Dynamics (TTD) campaigns conducted under the CAWSES India Phase II (Theme 3) program. Radar measurements for 72h in each month from December 2010 to September 2013 have been considered. The identified tropopause altitude with radar (RTH) is compared with the cold point (CPH) and lapse rate tropopause altitudes (LRH) obtained from simultaneous radiosonde data at three hourly intervals during these campaigns. Most of the time, a very good agreement between the RTH and CPH and/or LRH from radiosonde measurements is observed. The mean difference between RTH and CPH and RTH and LRH is found to be 0.1±1km and 0.5±1km, respectively. The smaller differences between RTH and CPH noticed in the present work when compared to other mid- and polar latitudes might be due to the well defined tropopause structure in the tropical latitudes. As the radar provides reliable data on the tropopause, its long-term variability is investigated using the data from 2007 to 2012 available from the MST radar.

It is widely accepted that the tropical tropopause is closely linked to climate change. Several campaigns have already been conducted and also are being planned to address various issues related to the tropical tropopause layer (TTL). Despite many campaigns, several scientific issues still remain unexplained including Indian summer monsoon dynamics, cirrus clouds and the trace gas distribution across the tropopause etc. In order to address some of the issues, particularly over the Indian region, intensive observational campaigns called ‘tropical tropopause dynamics (TTD)’ are being conducted since December 2010 at two stations namely Gadanki (13.5°N, 79.2°E) and Trivandrum (8.5°N, 76.9°E) under CAWSES India Phase-II programme. This overview article aims to bring out the current understanding on the tropical tropopause, issues addressed through the TTD campaigns and the details of the data collected in these campaigns using collocated instruments as well as complementary satellite data. So far 32 campaigns have been completed successfully and in this paper main focus is given for describing the systematic data collected using various techniques (MST radar, Mie lidar, Radiosonde, ozonesonde) simultaneously in each month. In general, over the study region affected by the monsoon, a prominent updraft prevails in the middle and upper troposphere regions covering TTL affecting the transport of minor species across the tropopause. The behavior of the cold point tropopause (CPT) at Gadanki and Trivandrum reveals that there are significant differences in the CPT characteristics even within the monsoon region. Cold point tropopause shows stronger sub-daily scale variation over Trivandrum than Gadanki though no indication of deep convection is present at the former location particularly in winter.

The mean spatiotemporal variations in tropopause parameters over the tropics (±35°, in latitude) in the Indian monsoon region are examined using the upper air data for an extended period obtained from radiosonde and Radio Occultation measurements. In general, the altitude of cold point tropopause (CPT) is a minimum near the equator and increases with latitude on either side. While CPT over the entire southern tropical latitudes and northern equatorial region is cooler (higher) during boreal winter and warmer (lower) during boreal summer, the annual pattern of CPT-temperature reverses in the northern hemispheric off-equatorial region. The temperature of lapse rate tropopause (LRT) is always negatively correlated with its altitude. While the annual variation of LRT-temperature in tropics is always positively correlated with CPT-temperature, the annual variation of LRT-altitude differs mainly in the off-equatorial regions. While the altitude of the convective tropopause is positively correlated with CPT-altitude over the latitude region 20°S–5°N, they are negatively correlated at the north of 10°N. In general, the tropical tropopause layer (TTL) is very thin (~3 km) near the equator and its thickness increases with latitude on either side of the equator to reach a peak value (of ~6 km) around ±30°. A pronounced decrease in TTL-thickness observed over the northern off-equatorial region during the ASM period can be attributed to the manifestation of very deep convection over the land near the Head Bay-of-Bengal region. The TTL-lapse-rate (γTTL) is large in the equatorial region and decreases with increase in latitude. While γTTL in the northern hemispheric off-equatorial region is low during winter, it increases and becomes comparable to that over equatorial region during the ASM period. The annual variations in CPT parameters as well as the TTL- thickness are significantly modulated by quasi-biennial oscillation and the El Niño Southern Oscillation.

The notions that (a) the frequency of occurrence of clouds is little affected by the sea surface temperature (SST) over the warm tropical oceans when it is maintained above ∼27.5°C and (b) over these regions the cloudiness is mainly governed by the large‐scale atmospheric dynamics, are well‐founded. This study investigates the association among SST, atmospheric wind divergence, vertical wind and cloudiness (convection) over the Arabian Sea, the Bay of Bengal and the tropical Indian Ocean during different seasons. Analysis carried out on data from all the above regions during 1996–2008 shows that the cloudiness and cloud top altitude increase with SST for SST > 26.5°C and become saturated for SST in the range of 28.5–29.75°C. Further increases in SST reduce cloudiness. Similar variations are obtained by analyzing the data during the Asian summer monsoon season (June–September) alone. In contrast, the cloudiness and cloud top altitude increase with SST even for SST > 29.5°C during winter (December–February). The base value of SST for a significant enhancement in cloudiness during winter is found to be ∼1°C larger than that during the summer monsoon season. Seasonal variations in the SST‐dependence of cloudiness over warm oceanic regions are primarily due to changes in the large‐scale atmospheric circulation which plays a pivotal role in regulating cloudiness, especially through surface‐level moisture convergence, mid‐tropospheric vertical winds, upper tropospheric divergence and orographic effects in the upwind direction. Regions of the warmest SST, maximum surface wind convergence and the largest cloudiness are generally not collocated.

Impact of the long duration noontime annular solar eclipse on 15 January 2010 on the vertical distribution of aerosols and mixing layer height (HM) in a well-developed convective atmospheric boundary layer (ABL) has been investigated using continuous Lidar observations over a tropical coastal station, Thumba (8.5°N, 76.9°E). This study shows that HM has decreased from its peak value of ∼1800m at 12:00h to ∼1000m following the annular phase of the eclipse (13:17h), while the corresponding decrease in the total aerosol abundance of ABL is ∼29%. The post-eclipse increase of HM is rapid compared to that during forenoon.