X. Liu , E. D. Rivière , V. Marécal , J. Arteta , G. Durry ... · X. Liu 1, E. D. Rivière 1, V....
1
Abstract During the periods of August 04-05, large MCSs (Mesoscale Convective Systems) were observed in the region of Aïr (central Niger) which later moved toward Niamey. Another large MCS developed and overshooted the tropopause over Chad, upwind of the micro-SDLA and FLASH-B water vapor measurements which show an hydrated layer above the tropopause. In order to study the impact of such a strong convective activity on the amount of water entering the stratosphere, we present results of two high resolution mesoscale simulations (3 nested grids) of these cases with the BRAMS model. Thanks to ECMWF reanalyses including AMMA observations, the model simulates reasonably well the Aïr MCS, its extend, its propagation speed and its lifetime The BRAMS model : description & simulation setup BRAMS = Brazilian version of the RAMS mesoscale model. Dedicated to the tropics (http:// www.cptec.inpe.br/brams) • Nested grids possible • Subgrid scale parameterisation of convection : Grell scheme for Grid 1 & 2 • 4Dvar Nudging (e.g. ECMWF, radiosounding, etc…) [email protected] [email protected] Concluding remarks Setup : • Initial conditions from ECMWF reanalyses including AMMA observations + soil moisture deduced from TRMM + surface data. • Simulation from August 03st 12:00 UT August 05th 12:00 UT • Simulation from August 05st 00:00 UT August 07 th 00:00 UT • 3 grid : 20 km / 4km / 1 km horizontal resolution. • 68 vertical levels up to 30 km. 300 m resolution in the UTLS • Microphysical Scheme : double moment. Setup from G. Pénide (LaMP) Meteorological situation METEOSAT IR brightness temperature and MIT Radar at Niamey Main simulation results ( 4 th August ) 1. A fine resolution (3 grids 1 km) is used to simulate 2 cases of overshoot and reproduce the propagation direction and speed of the MCSs. 2. Each overshoot induces water fluxes of several tons/s. 3. The overshoot over Chad (August 4) highlighted by Khaykin et al., (2009) to be responsible of the hydrated layer sampled by FLASH and micro-SDLA balloon flights on August 5 from Niamey is reproduced by our simulation. ATR flight (11:00Z15:00Z) VOC/OVOC 1c Case of August 4 th over Chad Case of August 5 th over Aïr (Central Niger) August 5th 19:00 Overshoot August 6th 07:00 MCS over Niamey Convection triggers in Aïr from a cloud area. It turns into a MSC which propagates South-westward. Main simulation results ( 5 th August ) 18:55 5th August 16780m ice+liquid+iso-theta Convection starts in the model 4 hours later than the observation in the Chad region. This convective system is fairly well reproduced by BRAMS. The 3 nested grids simulation allows the convection to reach the tropopause level and overshoot into the stratosphere (above the 400K θ-level) in the convective system mentioned in Khaykin et al. (2009) . The RHi profile at the position of overshoot shows that the ice evaporates rapidly once transported into the LS. Sedimentation of ice particles from the LS UT may play a negligible role in the cross - tropopause water budget. Altitude (m) Liquid + ice Aug05 18:55 TU lon = 8.3 Liquid + ice mixing ratio (g/kg) + θ-level August 4th 14:30 Overshoot signature from MSG brightness temperature (Khaykin et al. 2009) August 4th 18:00 August 4th 12:00 Water Budget • Computed with a 5 minute time step on isentropic surfaces. The vertical wind speed used in the calculation is the relative speed of air with respect to the θ-level displacement. • Figures : times series of upward total water flux through isentropic surface at 370K, 380K (~ tropopause), 390K and 400K over the third grid domain of the model during the overshoots period. Order of magnitute compatible with previous studies (Grosvenor et al. 2007, Chaboureau et al., 2007). • Tables : total mass of total water crossing the θ-levels during the whole overshoot event. Comparable flux par unit of time and surface. topography Model grids used for the simulation. (5th August) topography Model grids used for the simulation. (4th August) X. Liu 1 , E. D. Rivière 1 , V. Marécal 2 , J. Arteta 2 , G. Durry 1 and H. Hamdouni 1 1) Groupe de Spectrométrie Moléculaire et Atmosphérique (GSMA), UMR CNRS 6089, Faculté des Sciences, BP 1039, F-51687 Reims Cedex 2 - France 2) Laboratoire de Physique et Chimie de l’Environnement (LPCE) CNRS / Université d’Orléans, Franc Hourly accumulated surface rainfall (mm) August 5th 18:00 Grid 2 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 18:35 18:45 18:55 19:05 19:15 19:25 (Ton / s) 4.5 370k 380k 390k 400k Acknowledgements • AMMA database, DT-INSU SDLA team during the campaign. S. Khaykin for the FLASH-B data. References • Khaykin et al., Atmos. Chem. Phys , 9, 2275–2287, 2009 • Chaboureau et al., Atmos. Chem. Phys., 7, 1731–1740, 2007 • Grosvenor et al., Atmos. Chem. Phys., 7, 4977–5002, 2007 • The total amount of water crossing the tropopause is higher for August 4 than for August 5 because of a longer event. 4 th of August 5 th of August 14 16 20 altitude (km) 0 20 20 5 water vapor mixing ratio (ppmv) µ-SDLA Multi cell path corrected FLASH-B Hydrated layer above the tropopause. Khaykin et al. 2009 : link with overshoot over Chad Balloon-borne water vapor measurements Niamey August 5, 2006, 18:00 UT Grid 1 Grid 2 Grid 3 Chad Niger Niger Grid 3 Grid 3 4. Due to the very dry conditions in the LS, the ice particles injected by the overshoots sublimate rapidly. The enhancement of water vapor mixing ratio after the modelled overshoot is comparable with the one measured by FLASH and µ-SDLA. Grid 2 reasonably well the Aïr MCS, its extend, its propagation speed and its lifetime although this latter is shorter than the observed system. The model simulates an overshooting cell reaching 18.3 km. The water budget associated with this overshoot is comparable with the result of previous overshoots modeling studies (several tons/s). Another water budget associated with the overshooting MCS over Chad on August 4 is also given. This convective system is fairly well reproduced (position and propagation direction) by BRAMS, even though the modelled system triggers 4 hours later than the observations. The 3 nested grids simulation allows the convection to reach and cross the tropopause level. 9 Liquid + ice mixing ratio (g/kg) + θ-level Grid 3 Grid 3 Liquid + ice mixing ratio (g/kg) + θ-level MSG overshooting pixels time evolution BT (6.2 µm) –BT(10.8 µm) > 3 K MSG overshooting pixels time evolution Convection starts 1 hours later than the observation in the Aïr region. Its propagation direction and speed is correctly reproduced. According to our BTD calculation, the modeled overshoot appears later than the observations, although the overshooting convection is reproduced by the model. The overshoot reaches 18.3km at the maximum RHi profiles shows under-saturated conditions in the LS rapid sublimation of ice. Altitude (m) Grid 3 RHi (%) Aug04 18:40UT with iso-theta ice , rv , ice+rv Aug04 18:20UT RHi (%) + θ-level 1.3 ppmm ~ 2.08 ppmv Altitude (m) Altitude (m) RHi (%) Altitude (m) Mass mixing ratio (ppmm) Latitude ( ˚N ) Latitude ( ˚N ) Under saturated Latitude ( ˚N ) Theta (K) Water mass flux during 1 hours ( tons ) ( tons/h/km² ) 360 144283 23.29 365 27399 4.42 370 6911 1.12 375 3503 0.57 380 2377 0.39 385 1946 0.32 390 1666 0.27 395 1400 0.23 400 1222 0.20 0 1 2 3 4 5 6 7 8 9 10 11 12 August 4th 14:30 370K 380K 390K 400K 8 7 6 5 2 3 4 0 1 (Ton / s) 19:55 17:00 19:00 18:00 16:00 Theta (K) Water mass flux during 4 hours ( tons ) ( tons/h/km² ) 360K 116090 5.37 365K 45545 2.11 370K 24098 1.12 375K 14093 0.65 380K 9553 0.44 385K 7454 0.35 390K 6044 0.28 395K 5121 0.24 400K 4215 0.20
Transcript of X. Liu , E. D. Rivière , V. Marécal , J. Arteta , G. Durry ... · X. Liu 1, E. D. Rivière 1, V....
Abstract During the periods of August 04-05, large MCSs (Mesoscale Convective Systems) were observed in the region of Aïr (central Niger) which later moved toward Niamey. Another large MCS developed and overshooted the tropopause over Chad, upwind of the micro-SDLA and FLASH-B water vapor measurements which show an hydrated layer above the tropopause. In order to study the impact of such a strong convective activity on the amount of water entering the stratosphere, we present results of two high resolution mesoscale simulations (3 nested grids) of these cases with the BRAMS model. Thanks to ECMWF reanalyses including AMMA observations, the model simulates reasonably well the Aïr MCS, its extend, its propagation speed and its lifetime
The BRAMS model : description & simulation setup BRAMS = Brazilian version of the RAMS mesoscale model. Dedicated to the tropics (http://www.cptec.inpe.br/brams) • Nested grids possible • Subgrid scale parameterisation of convection : Grell scheme for Grid 1 & 2 • 4Dvar Nudging (e.g. ECMWF, radiosounding, etc…) �
[email protected] [email protected]
Concluding remarks
Setup : • Initial conditions from ECMWF reanalyses including AMMA observations + soil moisture deduced from TRMM + surface data. • Simulation from August 03st 12:00 UT August 05th 12:00 UT • Simulation from August 05st 00:00 UT August 07th 00:00 UT • 3 grid : 20 km / 4km / 1 km horizontal resolution. • 68 vertical levels up to 30 km. 300 m resolution in the UTLS • Microphysical Scheme : double moment. Setup from G. Pénide (LaMP)
Meteorological situation METEOSAT IR brightness temperature and MIT Radar at Niamey
Main simulation results ( 4th August )
1. A fine resolution (3 grids 1 km) is used to simulate 2 cases of overshoot and reproduce the propagation direction and speed of the MCSs.
2. Each overshoot induces water fluxes of several tons/s. 3. The overshoot over Chad (August 4) highlighted by Khaykin et al., (2009) to be
responsible of the hydrated layer sampled by FLASH and micro-SDLA balloon flights on August 5 from Niamey is reproduced by our simulation.
ATR flight (11:00Z15:00Z) VOC/OVOC
1c
Case of August 4th over Chad
Case of August 5th over Aïr (Central Niger)
August 5th 19:00 Overshoot
August 6th 07:00 MCS over Niamey
Convection triggers in Aïr from a cloud area. It turns into a MSC which propagates South-westward.
Main simulation results ( 5th August )
18:55 5th August 16780m ice+liquid+iso-theta
Convection starts in the model 4 hours later than the observation in the Chad region.
This convective system is fairly well reproduced by BRAMS.
The 3 nested grids simulation allows the convection to reach the tropopause level and overshoot into the stratosphere (above the 400K θ-level) in the convective system mentioned in Khaykin et al. (2009) .
The RHi profile at the position of overshoot shows that the ice evaporates rapidly once transported into the LS. Sedimentation of ice particles from the LS UT may play a negligible role in the cross -tropopause water budget.
Alti
tude
(m)
Liquid + ice Aug05 18:55 TU lon = 8.3
Liquid + ice mixing ratio (g/kg)
+ θ-level
August 4th 14:30
Overshoot signature from MSG brightness temperature
(Khaykin et al. 2009)
August 4th 18:00 August 4th 12:00
Water Budget • Computed with a 5 minute time step on isentropic surfaces. The vertical wind speed used in the calculation is the relative speed of air with respect to the θ-level displacement.
• Figures : times series of upward total water flux through isentropic surface at 370K, 380K (~ tropopause), 390K and 400K over the third grid domain of the model during the overshoots period. Order of magnitute compatible with previous studies (Grosvenor et al. 2007, Chaboureau et al., 2007).
• Tables : total mass of total water crossing the θ-levels during the whole overshoot event. Comparable flux par unit of time and surface.
topography
Model grids used for the simulation. (5th August)
topography
Model grids used for the simulation. (4th August)
X. Liu 1 , E. D. Rivière 1, V. Marécal 2, J. Arteta 2, G. Durry1 and H. Hamdouni1
1) Groupe de Spectrométrie Moléculaire et Atmosphérique (GSMA), UMR CNRS 6089, Faculté des Sciences, BP 1039, F-51687 Reims Cedex 2 - France 2) Laboratoire de Physique et Chimie de l’Environnement (LPCE) CNRS / Université d’Orléans, Franc
Hourly accumulated surface rainfall (mm)
August 5th 18:00 Grid 2
4.0
3.5
3.0
2.5
2.0
1.5
1.0
0.5
18:35 18:45 18:55 19:05 19:15 19:25
(Ton / s)
4.5
370k
380k
390k
400k
Acknowledgements • AMMA database, DT-INSU SDLA team during the campaign. S. Khaykin for the FLASH-B data.
References • Khaykin et al., Atmos. Chem. Phys , 9, 2275–2287, 2009 • Chaboureau et al., Atmos. Chem. Phys., 7, 1731–1740, 2007 • Grosvenor et al., Atmos. Chem. Phys., 7, 4977–5002, 2007
• The total amount of water crossing the tropopause is higher for August 4 than for August 5 because of a longer event.
4th of August 5th of August
14
16
18
20
altit
ude
(km
)
0 10 15 20 20 5 water vapor mixing ratio (ppmv)
µ-SDLA Multi cell path corrected
FLASH-B
Hydrated layer above the tropopause. Khaykin et al. 2009 : link with overshoot over Chad
Balloon-borne water vapor measurements Niamey August 5, 2006, 18:00 UT
Grid 1
Grid 2
Grid 3
Chad Niger
Niger
Grid 3
Grid 3
4. Due to the very dry conditions in the LS, the ice particles injected by the overshoots sublimate rapidly. The enhancement of water vapor mixing ratio after the modelled overshoot is comparable with the one measured by FLASH and µ-SDLA.
Grid 2
reasonably well the Aïr MCS, its extend, its propagation speed and its lifetime although this latter is shorter than the observed system. The model simulates an overshooting cell reaching 18.3 km. The water budget associated with this overshoot is comparable with the result of previous overshoots modeling studies (several tons/s). Another water budget associated with the overshooting MCS over Chad on August 4 is also given. This convective system is fairly well reproduced (position and propagation direction) by BRAMS, even though the modelled system triggers 4 hours later than the observations. The 3 nested grids simulation allows the convection to reach and cross the tropopause level.
9
Liquid + ice mixing ratio (g/kg)
+ θ-level
Grid 3 Grid 3
Liquid + ice mixing ratio (g/kg)
+ θ-level
MSG overshooting pixels time evolution
BT (6.2 µm) –BT(10.8 µm) > 3 K
MSG overshooting pixels time evolution
Convection starts 1 hours later than the observation in the Aïr region. Its propagation direction and speed is correctly reproduced.
According to our BTD calculation, the modeled overshoot appears later than the observations, although the overshooting convection is reproduced by the model. The overshoot reaches 18.3km at the maximum
RHi profiles shows under-saturated conditions in the LS rapid sublimation of ice.
Alti
tude
(m)
Grid 3 RHi (%) Aug04 18:40UT with iso-theta
ice , rv , ice+rv Aug04 18:20UT
RHi (%) + θ-level
1.3 ppmm ~ 2.08 ppmv
Alti
tude
(m)
Alti
tude
(m)
RHi (%)
Alti
tude
(m)
Mass mixing ratio (ppmm)
Latit
ude
( ˚N
)
Latit
ude
( ˚N
)
Under saturated
Latit
ude
( ˚N
)
Theta (K) Water mass flux during 1 hours ( tons ) ( tons/h/km² ) 360 144283 23.29 365 27399 4.42 370 6911 1.12 375 3503 0.57 380 2377 0.39 385 1946 0.32 390 1666 0.27 395 1400 0.23 400 1222 0.20
0 1 2 3 4 5 6 7 8 9 10 11 12
August 4th 14:30
370K
380K
390K
400K
8
7
6
5
2
3
4
0
1
(Ton / s)
19:55 17:00 19:00 18:00 16:00
Theta (K) Water mass flux during 4 hours ( tons ) ( tons/h/km² ) 360K 116090 5.37 365K 45545 2.11 370K 24098 1.12 375K 14093 0.65 380K 9553 0.44 385K 7454 0.35 390K 6044 0.28 395K 5121 0.24 400K 4215 0.20
