I’m a big fan of RGBs, especially the Nighttime Microphysics RGB, as the imagery is fantastic for delineating between various cloud types and highlighting clouds that can have some of the largest impacts…especially fog (cumulonimbus aside). I noticed the other day, however, that the Day/Night Band RGB can be superior for the detection of low clouds in certain situations, particularly when relatively warm, low clouds are observed against a similarly warm ocean background state, or through thin cirrus. For example, take a look at the Nighttime Microphysics RGB image below, centered on the Gulf of Mexico at 0658 UTC Sep 22. Notice the fairly solid area of low and mid-level clouds in the NW part of the Gulf along and near the TX and LA coasts. Much colder, thick convective clouds are also observed in the central NW Gulf (red/orange/yellow colors). Elsewhere, mostly high cirrostratus (or altostratus) and cirrus dominate…although a few low clouds can be observed interspersed around the cirrus canopy in some spots (pinkish-aqua colors).
Now, take a look at the VIIRS Day/Night Band RGB in the same region and valid at the same time (image 2). I’ve annotated where low clouds are clearly more observable.
Here’s a loop of both images, demonstrating how well the low clouds stand out in the Day/Night Band imagery vs the Nighttime Microphysics image.
The Nighttime Microphysics RGB contains a recipe of Infrared channels only, and cold, thin cirrus clouds become very dark blue (near black) in the imagery due to small contributions from red (thin optical depth), green (composed of ice particles) and blue (cold temperatures) colors. The resulting color of these cloud types then masks the underlying low cumulus. Meanwhile, the similar characteristis of warm, liquid-water clouds and the ocean’s surface make these features barely indistinguishable. Thin, cold cirrus are more transparent in visible wavelengths however, which are contained in the Day/Night Band Radiance RGB recipe. Also, the whitish color of clouds in the visible is contrasted against the dark ocean surface at night, making these types of warm, low clouds more distinguishable.
So, why is this important? Well, in sub-tropical latitudes, this may be less relevant. However, relatively low cumuls in the tropics or in very warm, moist, near-tropical environments can produce precipitation. As a forecaster once in the deep tropics (central Pacific) myself, I observed that small bands of low cumulus within the trade winds could produce precipitation once cloud depth exceeded about 5 kft. The assosicated cloud top temperatures were only slightly cooler than the surrounding ocean surface, leading to great difficulty when trying to observe these clouds in standard IR imagery, or even some types of RGB imagery, as demonstrated above. Additionally, thin cirrus clouds, which are often present in these environments due to nearby convection, can mask the underlying low cumulus. However, as shown here, Day/Night Band RGBs show promise in helping to locate low clouds at night that can be important for precipitation probability forecasts, particularly in the tropics or in very warm/moist environments.