Satellite Remote Sensing of Small Ice Crystal Concentrations in Cirrus Clouds David L. Mitchell

1. Satellite Remote Sensing of Small Ice Crystal Concentrations in Cirrus Clouds David L. Mitchell Desert Research Institute, Reno, Nevada Robert P. d’Entremont Satellite Meteorology Group Atmospheric and Environmental Research Lexington, Massachusetts R. Paul Lawson SPEC, Bouider, Colorado Further details of this work may be found at http://www.dri.edu/Projects/Mitchell/

2. Problem 1: Instruments aboard aircraft that measure the ice particle size distribution (PSD) cannot accurately measure the concentrations of small ice crystals. Larger ice particles may shatter on the probe inlet, producing tiny artifact ice crystals. Problem 2: Global climate models are very sensitive to the concentrations of small ice crystals in cirrus. Possible solution: Improved instruments + new remote sensing methods that estimate small ice crystal concentrations.

3. How Photon Tunneling Can be Used to Remotely Detect Small Ice Crystals in Cirrus Photon tunneling is the process by which radiation beyond the physical cross-section of a particle is absorbed. Tunneling is strongest when: Effective size and wavelength are comparable Particle shape is spherical or quasi-spherical The real refractive index is relatively large Therefore tunneling contributions at terrestrial wavelengths are greatest for smallest (D < 60 μm)‏ ice crystals. To detect the tunneling signal is to detect small ice crystals.

4. Size dependence of tunneling for single ice crystals

5. Ice particle size distribution (PSD) scheme for evaluating potential tunneling contributions

6. Decomposition of the two PSD modes

7. MADA calculation using climatological ice crystal shapes

8. Method for Estimating Small Crystal Concentrations - using effective emissivities -

9. 18 UTC Sun 22 Jul 07 200-mb NWP Winds Cloudsat (radar) overpass Aircraft flight tracks Flights between 1400 and 2000 UTC 0 2 4 6 8 10 12 14 6-hr Precipitation (mm)‏ TC4 Case Study - Sunday 22 July 2007

10. 2D-S Probe for measuring ice particle size distributions From Lawson et al. 2006, JTech.

11. Blue = cirrus; white = CBs; yellow = low cloud

12. MODIS Retrieval Results

13. MODIS Retrieval Results

14. 2D-S measurements for 22 July appear consistent with retrieved estimates of small-mode contributions

15. MODIS Retrieval Results homogeneous freezing nucleation begins?

16. Summary 1. The absorption of radiation by photon tunneling affects the small mode of the PSD but not the large mode, making thermal emission due to tunneling an ideal signal for detecting small ice crystals. 2. The tunneling signal is the emissivity difference between 12 and 11 μm wavelength. This was not previously understood. 3. A satellite retrieval methodology has been developed whereby it is possible to retrieve the small-to-large mode ice crystal concentration ratio, the ice crystal concentration for a given IWC, the IWP and De. The PSD is also estimated, even when it is bimodal. 4. Retrievals using a new satellite remote sensing methodology suggest a pronounced PSD small mode is often a measurement artifact. 5. The retrieved dependence of De and N on cloud temperature appear to support the homogeneous freezing nucleation mechanism. The absorption optical depth ratio (12 to 11 µm) of 1.08 could be used as a metric for the new MG microphysics.

17. 22 July 2007 Case Study: MODIS Retrieval Results

18. Remote Sensing of Small Ice Crystals (D < 60 μm) and the Process of Photon Tunneling Incident Ray Surface Wave Tunneling Tunneling depends on real refractive index, which changes abruptly between 11 and 12 μm wavelength. Critical Angle Diffracted Ray Tunneling can account for up to 45% of absorption in ice clouds for terrestrial radiation. Surface Wave Large-angle Diffraction

19. MODIS Retrieval Results

20. Cloud Optical Properties Depend on the Ice Crystal Mass- and Projected Area-Dimension Relationships That Characterize the PSD β • Mass = m = α D σ • Projected area = P = γ D Constants giving P & m = function of temperature for mid-latitude cirrus. Based on 22 cirrus flight missions, 104 horizontal legs and 15,000 km of in-cloud sampling, and Heymsfield et al. (2007).

21. Method for Estimating Small Crystal Amounts • Begin with satellite retrievals of cloud temperature and emissivity (ε) • at 11 and 12 μm wavelength channels. • Use the retrieved cloud temperature to estimate the PSD mean sizeD and dispersion (ν) for large and small mode. Difference between the solid and dashed curves results primarily from differences in the contribution of the PSD small mode to the IWC. This also determines effective diameter, De. Note that the large modeD and ν have little effect on above curves. • Locate the retreived Δε and the ε(11 μm) by (1) incrementing the modeled IWP to increase ε(11 μm) and (2) incrementing the small mode IWC, which elevates the curve. 4. If all IWC is in the small mode and retrieved Δε and ε(11 μm) is still not located, then decrease small modeDto locate them.

22. Method for Estimating Small Crystal Amounts - continued - • If the retrieved point lies below the “large mode only” curve (e.g. a dashed curve), then systematically decreaseDfor large mode until a match is obtained. Negative Δε values correspond to maximum allowed large modeDvalues. • This methodology retrieves IWP, De, the small-to-large mode ice crystal concentration ratio, and the ice particle concentration for a given IWC. It also estimates the complete PSD (even when it is bimodal).

23. Tunneling Efficiencies for Different Crystal Shapes (from Mitchell et al. 2006, JAS)‏ CRYSTAL TUNNELING EFFICIENCIES SHAPE Small Mode of SD Large Mode Mean of SD . D = 14 to 20 μm D = 53 μm D = 170 μm Aggregate 0.35 0.30 0.15 Bullet Rosette 0.70 0.40 0.15 Hex. Column 0.90 0.65 0.40 Hollow Column 0.70 0.50 0.15 Hex. Plate 0.60 0.15 0.00 Droxtal 1.00 0.80 0.40 Red circles indicate typical crystal shapes found in the small or large mode. Efficiencies are one reason tunneling is mostly manifested in the small mode.

24. MODIS Retrieval Results