Composition and Thermal Structure of the Upper Troposphere and Lower Stratosphere in a Penetrating Mesoscale Convective Complex Determined by Satellite Observations and Model Simulations

We describe here the composition and thermal structure of the upper troposphere and lower stratosphere during a penetrating mesoscale convective complex (MCC) event that occurred in southern China on 8 June 2009. Our results are based on satellite observations andWeather Research and Forecasting model simulations. Ice-rich and ozone-poor air reached as high as 17 km.The air was −5C colder inside the mature MCC than outside at the first cold-point tropopause near 17 km, −2C colder inside the mature MCC than outside at the second cold-point tropopause, and 3C warmer inside the mature MCC than outside between the two cold-point tropopauses. Corresponding to the temperature structure, there were two lower water vapor contents inside the MCC than outside near 17 km and 19 km while there was a higher water vapor content inside the MCC than outside near 18 km.


Introduction
Deep convective systems are associated with severe weather and can also affect stratospheric-tropospheric exchange and the radiation balance of the Earth by changing the composition and thermal structure of the upper troposphere and lower stratosphere.This topic has attracted much recent attention with the development of satellite observations [1][2][3][4][5][6].Penetrating mesoscale convective complexes (MCCs) are deep convective systems that have important effects on stratospheric-tropospheric exchange and the structure of the upper troposphere and lower stratosphere.
Most of deep convective activity occurs below the coldpoint tropopause (CPT) at 16-17 km.Convective activity rarely penetrates the tropical tropopause layer, which is found between the lapse rate minimum at 10-12 km and the CPT [7] and is a crucial area for stratospheric-tropospheric exchange [8].The deep convection system can significantly change the atmospheric structure of the tropical tropopause layer, increasing the height of the lapse rate minimum and lowering the height of the CPT [7].Penetrating MCCs are larger and longer lived than thunderstorm cells and therefore have a greater effect on the chemical and thermodynamic structure of the upper troposphere and lower stratosphere [3].Some satellites sample MCCs infrequently because of their viewing geometry, although, ultimately, the rarity of some types of sampling (particularly of tropopausepenetrating storms) is the result of the timing of satellite measurements.Most satellite-based systems are in Sunsynchronous orbits and make observations at local times before and after the time period in which tropopausepenetrating storms are most common (i.e., near the time of maximum surface temperature and highest instability).Therefore, there have been few studies of penetrating MCCs, but more studies of hurricanes in the upper troposphere and lower stratosphere [9,10].
A penetrating MCC was observed on 8 June 2009 by the NASA A-Train satellite constellation, which is widely used to study the structure of the atmosphere [11][12][13][14][15][16][17].The Microwave Limb Sounder Version 4 product, released in 2 Advances in Meteorology 2015 (https://mls.jpl.nasa.gov/),has led to a great improvement in data quality in the upper troposphere and lower stratosphere [18,19] and provides a good opportunity to understand the fine structure of the upper troposphere and lower stratosphere in a penetrating MCC.Satellite remote sensing and model simulation, together, can complement each other nicely to study the unique upper troposphere and lower stratosphere structure in a rare penetrating MCC.

NASA A-Train Satellite Data.
The A-Train (Afternoon Constellation) is a group of satellites (GCOM-W1, Aqua, CALIPSO, CloudSat and Aura) operated by NASA and its international partners (http://atrain.nasa.gov/).The satellites are all in polar orbit and all cross the equator northward at about 1:30 p.m. local time.The near-simultaneous observations of a wide variety of parameters by these satellites can be used for a single purpose [20].
The Microwave Limb Sounder (MLS) instrument on board NASA's Earth Observation System Aura satellite measures the composition of the atmosphere, the temperature, humidity, and amount of cloud ice at a horizontal resolution of c. 150 km.There are 10 pressure levels in the upper troposphere and lower stratosphere at 316, 261, 215, 178, 147, 121, 100, 83, 68, and 56 hPa.
The Atmospheric Infrared Sounder (AIRS) Level 2 standard product is derived from a combination of the AIRS product and the Advanced Microwave Sounding Unit on NASA's Aqua satellite.The horizontal resolution along the track is c.50 km and the spatial resolution of the AIRS-Advanced Microwave Sounding Unit at the nadir view is 13.5 km (http://airs.jpl.nasa.gov/)[18].The outgoing longwave radiation of the AIRS product was used in this study.
The Cloud-Aerosol Lidar with Orthogonal Polarization (CALIOP) instrument on NASA's CALIPSO satellite is a two-wavelength polarization-sensitive light detection and ranging instrument that provides high-resolution vertical profiles of aerosols and clouds.CALIOP uses three receiver channels: one measuring the 1064 nm backscattered intensity and another two measuring the orthogonally polarized components of the 532 nm backscattered signal.Level 2 products were used in this study [21].
The Cloud Profiling Radar (CPR) instrument on NASA's CloudSat satellite is a 94 GHz nadir-looking radar that measures the power backscattered by clouds as a function of distance from the radar.The original requirements for the CPR instrument are sensitivity defined by a minimum detectable reflectivity factor of −30 dBZ, along-track sampling of 2 km, a dynamic range of 70 dBZ, 500 m vertical resolution, and a calibration accuracy of 1.5 dBZ.The CPR Level 2 products were used in this study (http://cloudsat.atmos.colostate.edu/)[22].

Black-Body
Temperature from the FY2C Satellite.Blackbody temperature (BBT) data recorded by the FY2C geostationary satellite was obtained from the National Satellite Meteorological Centre, China.The hourly data with a horizontal resolution of 0.1 ∘ shows convection on a variety of scales [23].More details can be found at http://www .nsmc.org.cn/.

Global Positioning System Radio Occultation COSMIC
Data.Global positioning system (GPS) radio occultation temperature profiles were obtained from http://cdaac-www .cosmic.ucar.edu/cdaac/products.html.Anthes et al. [24] introduced the Constellation Observing System for Meteorology, Ionosphere, and Climate (COSMIC) data.We used Level 2 wet profile data with a vertical resolution up to 100 m.

Weather Research and Forecasting Model. The Weather
Research and Forecasting model WRF-ARW V3.5.1 was used to simulate the MCC.A simulation with a time step of 60 s started at 00:00 GMT on 8 June 2009.The model grids were set as shown in Figure 1, with horizontal resolutions of 10 km inside and 30 km outside the MCC.Seventy vertical layers were appointed under the model top at 20 hPa.The simulation schemes included the GF cumulus convection scheme, the RRTM longwave radiation scheme, the DUDHIA longwave radiation scheme, the YSU boundary layer scheme, and the NSSL 2-mom cloud microphysics scheme (http://www2.mmm.ucar.edu/wrf/users/).A sponge layer scheme with a thickness of 5 km was constructed in the upper boundary layer to avoid the reflection of gravity waves.The simulation was driven by the initial conditions and boundary conditions from the Global Forecasting System with 0.5 ∘ horizontal resolution.was >17 km inside the MCC (cyan dots, where the integrated attenuated backscatter coefficient of the top cloud was >0.01 sr −1 ⋅km −1 according to Dessler [11] and Pan and Munchak [26]), that is, higher than that outside (blue dots).The red dots indicate a large concentration of particles in the cloud top; the integrated attenuated backscatter coefficients of the top clouds in the MCC were generally c. 0.03 sr −1 ⋅km −1 larger than the background value of 0.01 sr −1 ⋅km −1 .The cloud was therefore a penetrating MCC.

Upper Troposphere and Lower Stratosphere Atmospheric
Structure in the MCC from Satellite Observations.Penetrating MCCs have a large effect on the structure of the upper troposphere and lower stratosphere, including the distribution of ozone (O 3 ), the ice water content (IWC), and the temperature and the amount of water vapor (H 2 O).
The concentration of O 3 and the IWC is strongly affected by upward motion in the area of the MCC.The cloud convection system can decrease the amount of O 3 at lower atmospheric levels via the transport of O 3 -poor air upward into the upper troposphere and lower stratosphere because ozone concentrations are significantly higher in the stratosphere than in the troposphere.The concentration of O 3 in the area of the MCC was <100 ppbv below 100 hPa.However, the concentration of O 3 outside the MCC was significantly >100 ppbv (Figure 3(a)).The missing value somewhere inside the MCC is based on the data reliability of ozone from MLS.
By contrast, air with a high IWC of 50 mg⋅m −3 was transported upward to 14 km and air with an IWC of 10 mg⋅m −3 was transported up to 16.5 km within the MCC (Figure 3(b)).
The temperature profile of the MLS (Figures 3(c) and 3(e)) shows a cold center (lower than −85 ∘ C) in the MCC at 17 km corresponding to the cloud top (Figure 2(d)).However, there was a warm anomaly at 2 km above the cold center, which is a typical characteristic of an MCC cloud top [10].The cloud top temperature in the MCC was 2-6 ∘ C lower than the temperature at the same level outside the MCC.The temperature inside the MCC was 2-3 ∘ C warmer than outside at 2 km above the first CPT.
The GPS radio occultation temperature profiles with a high vertical resolution support the MLS data.GPS radio occultation only detected the MCCs on 7 and 9 June 2009.The two temperature profiles are similar.Hence we used the 7 June 2009 MCC data (Figure 4(a)) as a reference for the MCC on 8 June 2009.When the 7 June 2009 MCC occurred, the first CPT dropped from 17 to nearly 16 km (Figure 4(b)).The temperatures inside the MCC at the first and second CPTs were lower than the temperature outside, but the temperature inside the MCC between the two CPTs was 3-4 ∘ C warmer than the temperature outside the MCC (Figure 4(b)).
The distribution of H 2 O measured by the MLS instrument (Figures 3(d) and 3(f)) shows a low concentration of H 2 O (<2 ppmv) at the cloud top at 100 hPa and a high concentration of H 2 O (>6 ppmv) at 68 hPa in the MCC.The distribution of H 2 O is coincident with the temperature distribution because the temperature has a large impact on the concentration of H 2 O, especially near the frozen layer.In an environment with a high IWC, the concentration of H 2 O in the upper troposphere and lower stratosphere is directly related to the saturation vapor pressure on the surface of ice crystals.When the temperature increases, the saturation  vapor pressure also increases and the concentration of water vapor increases as a result of evaporation.Hence the distribution of H 2 O followed the temperature distribution.

Upper Troposphere and Lower Stratosphere Atmospheric
Structure in the MCC from WRF Simulations.The positions and heights of the MCC were well simulated by the WRF simulations (Figures 2 and 5).The convective system began to develop at 13:00, the cloud top of MCC reached 17 km after 14:00, and the MCC matured at about 15:00, slightly ahead of the satellite detection.The atmospheric structure of the MCC (e.g., the temperature, IWC, and H 2 O distribution) were well represented in the model output.The model output was a good match for the observed divergence, vertical velocity, and potential pseudo equivalent temperature.Upwelling was −0.8 Pa⋅s −1 near 200 hPa in the mature stage of the MCC (Figure 6(a)).There was a strong divergence center (>3 × 10 −3 s −1 ) corresponding to a cold meridional deviation center (−5 ∘ C) at the top of the upwelling region in the MCC near 100 hPa.The model showed a disturbance structure for temperature with a negative meridional deviation anomaly of −5 ∘ C at the first CPT and −2 ∘ C at the second CPT and a positive meridional deviation anomaly of 3 ∘ C between the two CPTs in the mature stage relative to that in the early stage (Figures 6(a

Conclusions and Discussions
A variety of satellite observations and WRF simulations were used to study the upper tropospheric and lower stratospheric composition and thermal structure of an MCC that occurred over southern China on 8 June 2009.
The WRF simulations supported the upper troposphere and lower stratosphere temperature structure obtained via GPS radio occultation measurements and MLS observations.The air was −5 ∘ C colder inside the MCC than outside at the first CPT near 17 km, −2 ∘ C colder inside the MCC than outside at the second CPT, and 3 ∘ C warmer inside the MCC than outside between the two CPTs.There was a lower water vapor concentration inside the MCC than outside near 17 km and a higher water vapor concentration inside the MCC than outside near 18 km, corresponding to the temperature structure.This is because the water vapor distribution is controlled by the temperature distribution via microphysical processes when a large number of ice crystals are delivered into the upper troposphere and lower stratosphere by the MCC.The ozone-poor air in the upper troposphere and lower stratosphere reflected the convective transport of the MCC.
This study presented the fine structure of the upper troposphere and lower stratosphere in a penetrating MCC during Asia summer monsoon which was rarely reported.Due to the more larger scale of MCC than a common convective activity, the second CPT was located in a higher position than the background climatological CPT, while the location of the second CPT in common convective activity is near the background climatological CPT [10].

Figure 1 :
Figure 1: Location of study area and grids used in the MCC simulation.

Figure 2 :Figure 3 :Figure 4 :
Figure 2: Observations from satellites at 18:00 GMT on 8 June 2009.(a) Outgoing longwave radiation (OLR, W⋅m −2 ) from the AIRS instrument (the middle thick red line is the orbit of MLS, CPR, and CALIOP instruments).(b) Black-body temperature contours from the FY2C satellite (green, blue, red, and black contours denote −20, −32, −50, and −65 ∘ C, resp.).(c) Cross-section of radar reflectivity from CPR (dBZ, color-shaded) and the potential pseudo equivalent temperature (K, black contours) along orbit.(d) Cross-section of data from the CALIOP instrument; left-hand -axis shows the tropopause height (black dots) and the cloud top height (blue and cyan dots; cyan for overshooting cloud top) and the right-hand -axis shows the integrated attenuated backscatter coefficients in the top cloud layer (red dots, sr −1 ⋅km −1 ).

Figure 5 :
Figure 5: Zonal cross-section of the evolution of reflectivity (unit: dBz) along 25.4 ∘ N from the WRF simulations.