Observation of a Saharan dust outbreak on 1-2 August 2007: determination of microphysical particle parameters Paolo Di Girolamo 1, Donato Summa 1, Rohini Bhawar 1, Tatiana Di Iorio 2, Marco Cacciani 2, Igor Veselovskii 3, Alexey Kolgotin 3 1 DIFA, Università degli Studi della Basilicata, Potenza, Italy, 2 Dipartimento di Fisica, Università degli Studi di Roma “La Sapienza”, Roma, Italy 3 Physics Instrumentation Center, Troitsk, Moscow Region, Russia 6 th COPS Workshop, 27 – 29 February 2008 University of Hohenheim, Stuttgart
Particle Backscatter Ratio at 1064 nm, 1-2 August TIME (UTC) ALTITUDE A.G.L. (m) 8 BASIL
18:45 Water Vapour Mixing Ratio 1-2 August :16 TIME (UTC) g/Kg ALTITUDE A.G.L. (m) out-flow boundary
Measured parameters: particle backscattering 355, 532 and 1064 nm 3 particle extinction 355 and 532 nm 2 depolarization 355 & 532 nm, atmospheric temperature water vapour mixing ratio relative humidity from simultaneous measurements of temperature and water vapor mixing ratio BASIL Raman Lidar Raman lidar measurements (25 May – 30 August 2007) More than 500 hours of measurements distributed over 58 days particle size and microphysical parameters COPS Web Page Operational Products
The retrieval scheme employs Tikhonov’s inversion with regularization Algorithm developed at the Physics Instrumentation Center Inversion algorithm Particle size distribution parameters: Mean radius r mean Effective radius r eff Number concentration N Surface concentration S Volume concentration V Complex refractive index m r and m i Parameters of a bimodal size distribution 3 + 23 + 2 In the solution of the inverse problem, particle size distribution f(r) is approximated by the superposition of base functions B j (r) as: where c j (z) are the weight coefficients. Base functions have a triangular shape on a logarithmic-equidistant grid Veselovskii et al., Appl. Opt. 41, 3685–3699, 2002.
Inversion with regularization r min =0.05 m, r max =15 m 1.3 <m r < 1.6 0<m i <0.04 f(r) Mean radius r mean Effective radius r eff Number concentration N Surface concentration S Volume concentration V numerically integrating f(r) over the radius interval [r min, r max ]
Particle Backscatter Ratio at 1064 nm, 1-2 August TIME (UTC) ALTITUDE A.G.L. (m) 8 Focus: two specific times when aerosol loading was higher 21:00-21:30 UTC on 1 August :00-00:30 UTC on 2 August 2007 (red dashed lines in figure)
1 August 2007, 21:00-21:30 UTC 2 August 2007, 00:00-00:30 UTC averaging layers m m m m m m m m 400 m thick
1 August 2007, 21:00-21:30 UTC2 August 2007, 00:00-00:30 UTC Particle size distribution 0.1 m <r mean < 0.2 m, 0.1 m <r eff < 1.0 m 5 m 3 /cm 3 <V< 40 m 3 /cm 3 Particle size distribution fine mode: m coarse mode: 1-4 m
1.44 <m r < <m i < August 2007, 21:00-21:30 UTC Method: Inversion with regularization, performed by generalized cross-validation. In this retrieval, the combination of particle extinction and backscatter coefficients becomes especially important. The number of backscatter coefficients in the retrieval procedure should exceed the number of extinction coefficients by a factor of 2–3.
Backward trajectories ending at 00:00 UTC on 2 August 2007 NOAA-ARL HYSPLIT Lagrangian trajectory model The air masses observed in Achern in the altitude region km a.g.l. originated in the mixed layer over the Saharan desert
21:30 01:15 23:22 Particle Backscattering coefficient at 355nm (log scale) ALTITUDE A.G.L. (m) TIME (UTC) 1-2 August 2007 Dust particle hygroscopicity 22:00-23: km
21:30 01:1523: ALTITUDE A.G.L. (m) TIME (UTC) Relative Humidity 1-2 August : :00-23: km
Particle backsc. coeff. at 355 nm vs RH, 1 August 07, 22:00-23:00 UTC, t=2min, km RH [%] Particle back 355 nm [m-1 sr-1] x x10 -6 Substantial increase in particle backscattering when RH > 75 % Swelling tendency of hygroscopic aerosol particles at large RH values Trend compatible with partially soluble aerosol particles Back-trajectories show that airmasses originated in the Saharan desert transited for several days over the Atlantic Ocean Aged dust particles presumably mixed with maritime aerosol during the advection to the meaurement site and partially coated with hygroscopic material
lidar dark band freezing level radar bright band freezing level University of Manchester Radio Wind Profiler, 1290 MHz UHF Doppler radar BASIL Raman Lidar radar bright band freezing level MIRA 36, Radar Reflectvity at 36 GHz radar bright band freezing level MIRA 36, Linear Depolarizatio Ratio 9 m/s 4.5 m/s 4000 ALTITUDE (m) radar bright band freezing level
10:46 11:37 Shear IOP 9c – 20 July 07 Passage of the frontal zone, with a Mesoscale Convective System inbedded The waves like structures seen in the data just prior to the arrival of the thunderstorm are due to shear between inflow and outflow regions ALTITUDE A.G.L. (m) TIME (UTC) THUNDERSTORM Range corrected signals at 1064 nm
21:00 04:00 00:30 BASIL – Rhine Valley Supersite (Lat: ° N, Long: 8.06 E, Elev.: 140 m) July 2007 – Water vapour mixing ratio g/kg 1 0 Height a.s.l. (m) TIME (UTC) T = 5 min, z = 150 m mixing ratio (g/kg)
10:46 11:37 10:46 11:37 Range corrected signals at 1064 nm Water vapour mixing ratio
1064 532 355 355 532
Particle Backscatter Ratio at 1064 nm
Particle Backscatter Ratio at 532 nm
1-D approach 2-D approach
mRmR