Smoky hurricane
16 September 2020 10:00–15:00 UTC
A large smoke cloud was sucked into Hurricane Paulette, changing microphysical properties of associated cloud bands.
07 December 2022
16 September 2020
By Ivan Smiljanic (CGI)
Hurricane season 2020 is provided not only high number of Tropical Cyclones, but also rarely seen large-scale smoke intrusion into the cyclonic system.
Large-scale advection of smoke from the fires in California and Oregon met the path of hurricane Paulette. A low pressure system pulled the smoke in, setting up the conditions for more special cloud formation — a type of DIBS (Dust‐Infused Baroclinic cyclone Storm) cloud, this time seeded with smoke rather than dust particles. The mixture of smoke and water/ice particles gave birth to smaller cloud particles, longer lasting high ice clouds and more granular texture of cloud tops.
It appears that smoke, unlike dust, produces a finer granular cloud structure. The size of individual cloud crystals also seemed to not be very small. However, comparing two contemporary hurricanes, Teddy and Paulette, one can notice smoother cloud tops with a sort of cirrus veil persisting aloft. Also in the Convection RGB Paulette's cloud bands take on more yellow hues (generally, not by looking at isolated smaller cloud patches), a sign of smaller cloud particles.
Image comparison
Figure 2: Comparison of between hurricane Paulette and Teddy, GOES-16 Convection RGB and Natural Colour RGB, 16 September 12:00 UTC.
Looking closer, through 500m resolution of the ABI VIS0.64 channel, the cloud bands appear more 'milky', having a texture similar to whipped cream (Figure 3 and 4).
The hurricane's cirrus outflow also took on a prominent shape – showing very persistent, linear cirrocumulus formations.
Long-distance smoke transport was happening in the higher levels of atmosphere, detected also by NIR1.3 ABI channel imagery. This channel is ‘blind’ to any particle reflection taking place in the lower troposphere (roughly lower than 500hPa), due to water vapour absorption in around the 1,300nm spectral region. Reflectance range of this channel needs to be enhanced down to only 0-1 % for the full extent of smoke to be seen, even 0-3 % range is not completely sufficient for that (comparison in Figure 6).
Image comparison
Figure 6: GOES-16 NIR1.3 channel, range 0-1% (0-3% for background image), 16 September 12:00 UTC.
