1. Daniel Grosvenor, Thomas Choularton, Martin Gallagher (University of Manchester, UK); Thomas Lachlan Cope and John King (British Antarctic Survey). Daniel Grosvenor, T. Choularton, M. Gallagher, K. Bower, J. Crosier (University of Manchester, UK); Thomas Lachlan Cope and Russell Ladkin (British Antarctic Survey). In-cloud aircraft observations over the Antarctic Peninsula

2. ContentsContents Antarctic cloudsAntarctic clouds Flight 102 – In situ observations of lenticular cloudsFlight 102 – In situ observations of lenticular clouds Ice number observations – can they give an estimate of Ice Nuclei concentrations?Ice number observations – can they give an estimate of Ice Nuclei concentrations? How do they compare to current ice parameterisations (based on non-Antarctic clouds)?How do they compare to current ice parameterisations (based on non-Antarctic clouds)? Are Antarctic clouds different?Are Antarctic clouds different? Flight 104Flight 104 Ice observations at colder temperaturesIce observations at colder temperatures Ice formed by the Hallet Mossop process.Ice formed by the Hallet Mossop process.

3. Antarctic Clouds – why the interest? Antarctica covers a very large area and cloud cover is extensive. Therefore they are important in determining the radiation balance of the Earth. Yet, little is known about Antarctic clouds due to a lack of in-situ observations – representing them in climate models is therefore difficult. Satellite observations have the potential to give us widespread cloud information, but retrievals require assumptions about cloud properties often based on mid- latitude clouds.

4. The BAS cloud instruments The CAPS instrument Consists of 3 instruments  CAS (Cloud Aerosol Spectrometer)  Size distributions of particles 0.61-50 μm in diameter  CIP (Cloud Imaging Probe)  Takes images of particles 25-1550 μm in diameter (mainly ice)  Can calculate size distributions from these  Hotwire probe  Measures the liquid water content How do Antarctic clouds differ from mid-latitude ones?  Cloud Condensation Nuclei (CCN) concentrations?  Ice Nuclei (IN) concentrations?  Different CCN/IN sources – e.g. Bio IN? Knowing these is important as they determine cloud reflectivity

5. The Antarctic Peninsula region Larsen B Larsen C Wilkins 1540 km 1750 km Topography Scale comparison Consists of a long ridge of high mountains (up to ~2000 m high). = Rothera BAS base

6. Case study – lenticular clouds in Marguerite bay Approx wind direction Altitude (m)  Deep low in the North Weddell Sea.  Led to a strong cross Peninsula flow (east to west).  Large stacks of lenticulars were developing over the mountains.  But flew through bands of lenticulars developing out into Marguerite Bay.  Clouds most likely formed on the crests of lee waves. = Rothera BAS base

7. Flow over mountain sets off vertical motions Stable air on downwind side allows vertical oscillations Lee wave clouds Since such clouds are likely to have been recently formed and are not likely to be deep they are quite simple May therefore be useful to look at ice nucleation

8. The overall picture Approx wind direction Blue - indicates small particles, probably droplets Green – large particles, probably ice Grey - both.

9. Flight segment 20:20-20:40 UTC Approx wind direction Blue - indicates small particles, probably droplets Green – large particles, probably ice Grey - both.

10. Examining the lee waves Gravity waves of temperature amplitude 2-5 o C. Horizontal wavelengths of 9-10 km. Predominately liquid formed at the crests of the gravity waves. But some ice too Ice present on the downward part of the waves – likely sedimentation from above RHi > RH at these temperatures – so would be supersaturated w.r.t. ice if are forming liquid

12. Not many observed ice crystals to base the statistics on.

13. Blue - indicates small particles, probably droplets Green – large particles, probably ice Grey - both.

14. 3500 m 3000 m 2500 m 2000 m 1500 m Blue - indicates small particles, probably droplets Green – large particles, probably ice Grey - both.

15. 3500 m 3000 m 2500 m 2000 m 1500 m Blue - indicates small particles, probably droplets Green – large particles, probably ice Grey - both. 3500 m 3000 m 2500 m 2000 m 1500 m

19. Ice Nuclei concentration parameterisations Heterogeneous Ice Nuclei (IN) These include deposition IN (direct nucleation from vapour phase) and condensation IN (liquid droplet nucleated first, which then freezes)  Numbers of up to ~0.15-0.35 per litre predicted for the temperature range of the lenticulars for the WRF scheme  Actual ice concentrations of 0.1-0.45 per litre observed. Thus IN parameterisation seem to be of the right magnitude.

20. Immersion and contact IN Ice Nuclei concentration parameterisations Bigg’s (immersion IN) - IN already contained within droplets Contact IN – when droplets collide with airborne IN and freeze  Gives a rate of freezing – need to estimate a period of nucleation to get a concentration

21. Estimation of ice duration of formation Icy regions at gravity wave crests are ~5 km wide. Wind speeds of ~20 m/s. Gives an ice forming time of ~250 s.  Gives an ice concentration of ~0.015-0.040 per litre for -11 to -14 o C.  For higher LWC of 0.2 g m -3 get 0.015-0.080 per litre.  Actual ice concentrations of 0.1-0.45 per litre observed. Thus immersion ice parameterisations are on the low side of the observed ice concentrations.  Heterogeneous IN likely slightly dominant over Biggs and contact freezing, according to parameterisations.  Overall, observed ice concentrations similar to predicted IN concentrations

22. Blue - indicates small particles, probably droplets Green – large particles, probably ice Grey - both. 3500 m 3000 m 2500 m 2000 m 1500 m

32. Blue - indicates small particles, probably droplets Green – large particles, probably ice Grey - both.

34. Summary for this flight  Gravity waves allowed liquid and ice supersaturation and thus droplet and ice nucleation.  Ice numbers that approximately agree with typical IN parameterisations  Deposition/condensation IN predicted to be dominant over immersion nucleation.  Except in the likely aircraft seeding region  How representative of IN concentrations are the observed ice crystal concentrations?  No ice observed in some of the liquid regions – why?

35. Summary of another flight  Cold temperatures of cloud over the mountain.  Fairly low concentrations and large ice particles:- Ice mass (mg m -3 ) Ice number

36.  Hallet Mossop splinter production zone  Plenty of liquid water available.  Lots of ice splinter columns observed:- Ice number Liquid water Ice mass (mg m -3 )  Cold temperatures of cloud over the mountain.  Fairly low concentrations and large ice particles:-

37. Ice Nuclei parameterisation comparison  Observed ice generally < 1 per litre  Suggests possible overestimation of parameterisations? Heterogeneous IN parameterisations estimate ~2.25 L -1 at T= -20 o C Biggs freezing (immersion IN) of up to ~0.02 L -1 s -1 predicted for T= -20 o C and LWC=0.3 g m -3.  Wind speeds of 18 m/s at T=-20 o C. Ice present over distances of ~25 km.  Gives estimate of ~>20 per litre at coldest temperatures.

38. Conclusions Lee wave (lenticular clouds) likely provide a good “natural laboratory” to look at Ice Nuclei numbers. Mostly liquid formed in the gravity wave crests, but some ice was observed in the crests and in the troughs (likely precipitated from above). Ice numbers were consistent with IN parameterisations for the -11 to -14 o C temperature range. Parameterisations suggested that deposition/condensation IN likely the biggest source of ice. Likely seeding of ice from aircraft exhaust – caution required in data interpretation and flight track planning. For the colder clouds at -20 o C the parameterisation numbers were considerably higher than those observed (factor of 20 or so). In the -3 to -8 o C temperature range the Hallet Mossop process was observed producing more ice particles – concentrations up to 3.5 per litre. This process is likely to be important for glaciating Antarctic clouds given the likely low IN concentrations.