Long-exposure digital holography applied to study mixing at the - - PDF document

▶
long exposure digital holography applied to study mixing
SMART_READER_LITE
LIVE PREVIEW

Long-exposure digital holography applied to study mixing at the - - PDF document

Apr 02, 2024 205 likes •294 views

23.10.2017 Long-exposure digital holography applied to study mixing at the laboratory analogue of cloud top J.L. Nowak 1 , J. Fugal 2 1 Institute of Geophysics, University of Warsaw, Poland 2 Institute for Atmospheric Physics, Johannes

slide-1
SLIDE 1

23.10.2017 1

Long-exposure digital holography applied to study mixing at the laboratory analogue of cloud top

J.L. Nowak1, J. Fugal2

1 Institute of Geophysics, University of Warsaw, Poland 2 Institute for Atmospheric Physics, Johannes Gutenberg-University Mainz, Germany

How to simultaneously measure cloud droplet size, position and velocity ?

2

What is the structure of the artificial cloud top ?

slide-2
SLIDE 2

23.10.2017 2

Hologram is an interference pattern of 2 beams: reference and scattered light

3

Object Hologram plane CCD 4.2 MP camera Laser 532 nm Single-mode fiber Lens cover Microscope Sample volume ~ 1 cm3

Single long-expos. hologram contains information about particle size, position and 2C velocity

4

  • Vertical position:

depth focusing based on image gradient

  • Horizontal position:

center of the particle trace

  • Horizontal velocity:

length and orientation of the trace

  • Droplet diameter:

width of the trace Exposure time: 5 ms Trace length ~ 110 μm Trace width ~ 18 μm Diameter ~ 18 μm Velocity ~ 1.8 cm/s

slide-3
SLIDE 3

23.10.2017 3

How to find particles in the reconstructed volume ?

5

Raw hologram Quality control Background correction Numerical reconstruction (angular spectrum method) Object detection based

  • n thresholding

3D pixel patches

How to distinguish between droplets and artifacts ?

6

3D pixel patches Depth focusing, Lateral positioning Parameter estimation (~90) Refinement based on patch parameters Random subset (~1%) manual classification Fitting optimal classification model Automated classification with the model Machine learning approach

slide-4
SLIDE 4

23.10.2017 4

Simpler model makes more mistakes in classification

  • f objects from experimental holograms

7

Detection performance depends on droplet size

according to tests on simulated holograms Model Detection error E Decision tree 31.3 % Logistic regression 26.9 % Support vector machine 23.7 %

𝑥 = 1 3 𝑥 = 2 3

Diameter [μm] 5 10 15 20 25 False Overall detection rate 11 % 44 % 75 % 86 % 87 % 7 %

𝐹 = 𝑥𝑆 + 𝑥𝑆 False alarm rate Miss rate

How to produce a cloud in the lab ?

8

slide-5
SLIDE 5

23.10.2017 5

How the laboratory cloud compares to nature ?

9

 cloud - clear air interface  mixing  temperature inversion  droplet size  droplet concentration ? spatial and temporal scales ? pressure changes ? boundary conditions ? droplet production

The cloud mixes with clear dry air from above

10

slide-6
SLIDE 6

23.10.2017 6

Droplets are spatially distributed in filaments

due to mixing and evaporation

11 12

slide-7
SLIDE 7

23.10.2017 7

Statistical cloud top structure seems to be layered

13

101 holograms Mean Std Min Max Error Concentration [cm-3] 305 222 38 1004 ? Horizontal velocity [mm/s] 6.8 3.9 25.6 1.8 Droplet diameter [μm] 12.7 3.2 8.5* 33 4.5

* Resolution limit

Not enough to draw strong conclusion LWC CONC MVD

Acknowledgements S.V. Gilles, O. Schlenczek, W. Schledewitz

  • A. Górska, S. Malinowski

Holosuite software package Erasmus+ student exchange programme

Contact jnowak@igf.fuw.edu.pl

Summary

Size, position and 2C velocity estimated with one measurement Supervised machine learning helps in particle detection Mixing at the cloud top simulated in the laboratory chamber Droplets at the interface spatially distributed in filaments