Crisscross cloud patterns over American Midwest
24 September 2018 13:12–16:52 UTC
The GOES-16 satellite observed multiple intertwined wave patterns in the cloud field in September 2018, at the advanced resolution of 500 m.
04 May 2023
24 September 2018
By Ivan Smiljanic (SCISYS)
Very complex wave patterns were observed over the US states of Nebraska and Kansas. Stratus fields that formed during the night hours started to take different wave forms during the day.
Interesting dynamics of the cloud formation/dissipation in the area of interest can be observed through animated imagery combining the high resolution visible (HRV) 0.64 µm channel and 24hr Microphysics RGB (Figure 1 and Figure 2).
There were four emerging wave patterns, focused at designated areas on Figure 3 (slider left):
- Cyan area (1): Gravity wave clouds (stretching W-E direction) formed due to the main low level south-southwesterly atmospheric flow (confirmed with near-by radiosounding measurement ) that moved over the elongated (west-east) Kansas topography .
- Green area (2): Wave pattern with very small distance between the crests (roughly 1km) in a form of the cloud streets over almost perpendicular wave pattern in area 1.
- Purple area (3): Dominating wave pattern of cloud streets that formed in the stratus field (which formed in the area during the previous night) that takes up the majority of the domain. This is a similar wave pattern to that in area 2 (similar dynamics), but with ‘bigger’ wave, that follow the main low level flow.
- Yellow area (4): Gravity waves that formed as result of the advection of a Rossby wave from the north that moved towards the south east. This is the ‘biggest’ wave pattern in the domain with 8–10km distance between wave crests, elongated in the south west-north east direction.
Image comparison
Figure 3: Comparison of combined HRV (0.64µm channel) and 24hr Microphysics RGB imagery at nominal resolution of 500m (left ) and 1km (right).
Comparing the imagery at 1km (highest Meteosat Second Generation SEVIRI instrument resolution) and 500m (future Meteosat Third Generation (MTG) Flexible Combined Imager (FCI) instrument capability) (Figure 3), it becomes obvious that very fine scale waves are only detected at 500m resolution. The wave patterns in areas 2 and 3 are barely visible.
In area 5 (red box) on Figure 3 one can see the example of possible misleading information inferred from satellite data. In the reduced resolution image it appears (most vividly in area 5, but not only) that the alignment and the distance between waves becomes different when comparing to the original image (at 500m resolution). In another words, suddenly there are different clouds waves appearing when reducing resolution of the image (‘wider’ waves leaning to the left). In this case this is most likely due to a re-sampling of the pixels done by visualisation software. But the question is if these artefacts may also be caused by the specific sampling of the instrument itself or, for instance, by processing of the data.
Zooming in to almost pixel level over some parts of domain (Figure 4 and Figure 5) the resolving power of higher resolution imagery is more vividly shown, the lines in the annotated imagery reveal the approximate position of the cloud waves (taken from random sub-areas of the observed domain). Note: nominal resolution of 500 m translates roughly to pixel size of 0.8 x 0.6km (lat/long) around the observed domain.
