The GOES-16 Day Cloud Phase Distinction RGB can be used during the day to differentiate various cloud types/layers, including over snow cover, a task which is often difficult with visible imagery alone given similarly moderate-high reflectance values. Further, clouds that are most likely producing snowfall can be differentiated from those that are not. The Day Cloud Phase Distinction RGB is a “simple” RGB available in AWIPS that combines the VIS with IR window and snow/ice bands to differentiate these cloud types and snow. A few scenes from 10 November 2017 will be analyzed.
First we discuss the components of the Day Cloud Phase Distinction RGB. The red component to the RGB is the 10.35 um IR window channel. Warmer surfaces (land surface, snow, low clouds) will have little to no red, while progressively colder surfaces (high clouds) will have more red. The green component is the 0.64 um visible channel. Less reflective surfaces (bare ground) will have less green, while highly reflective surfaces (clouds, snow) will have more green. Finally, the blue component is the 1.61 um snow/ice band. Low reflectance surfaces (ice clouds, snow) will have very little blue, while more highly reflective surfaces (water clouds) will have more blue. Bare ground tends to be somewhere in-between.
The Day Cloud Phase Distinction RGB is analyzed for a scene centered over Colorado (Fig 1). The bare ground appears a dark blue given a lack of red (warm in IR), lack of green (not very reflective in VIS), and moderate amount of blue (land surface=moderately reflective in snow/ice band). Bodies of water will appear closer to black given very low reflectance in the VIS and NIR channels. Looking over southern Colorado, the bright red color indicates high cirrus clouds: there is a lot of red (cold in IR), moderate green (moderately reflective in VIS) and low amount of blue (ice particles=not reflective in snow/ice band). The shades of light blue to aqua colors represent low water clouds: there is very little red (warm in IR), moderate green (moderately reflective in VIS), and higher values of blue (water droplets=highly reflective in snow/ice band). Snow cover appears green: there is very little red (warm in IR), moderate green (reflective in VIS), and low amount of blue (ice particles=not reflective in snow/ice band). Looking at this scene using this RGB, one can easily see that there are high, thick ice clouds over much of southeast Colorado, low water clouds in the Arkansas River Valley from the edge of the thick cirrus to Canon City and in the Colorado Springs area, and under the thinner cirrus in far eastern Colorado and western Kansas. The snow on the mountains is also easy to distinguish.

Figure 1: 10 November 2017 GOES-16 Day Cloud Phase Distinction RGB over Colorado. Full res
The various features stand out much better in the RGB when compared to the VIS (Fig. 2).

Figure 2: 10 November 2017 GOES-16 Day Cloud Phase Distinction RGB and VIS over southeast Colorado. Full res
Looking at the scene over Montana, there is widespread snow cover (green), high clouds over the Canadian border (red), and low clouds over the snow (gray). The low clouds appear grayer up north because these low water clouds are in a much colder airmass (more red) compared to south in Colorado, and therefore contain nearly equal amounts of red, green, and blue (Fig. 3).

Figure 3: 10 November 2017 GOES-16 Day Cloud Phase Distinction RGB and VIS over Montana. Full res
Looking east over Lake Michigan, lake effect snow bands stand out as bright green/yellow (low clouds are relatively warm so moderate-low red, contains ice so low blue, cloud so at least moderate green) versus the surrounding pure water clouds that are gray/aqua. Including radar imagery, you can see that the highest reflectivity lies under the bright green/yellow regions where we know ice processes are occurring (Fig. 4).

Figure 4: 10 November 2017 GOES-16 Day Cloud Phase Distinction RGB and VIS and radar reflectivity mosaic over Lake Michigan Full res
Deep convection (thick clouds with ice) will appear yellow or very bright red or green due to the very high levels of red (very cold) and green (very bright) but low amounts of blue (ice not reflective at 1.6 um).
Shallow fog was present in the river valleys along the Canadian/US border in northeast Montana and northwest North Dakota on 13 November. The river valley fog resided on the edge of snow and bare land, making it difficult to track the evolution in any single band. The RGB allowed for easy detection and tracking of this fog during the day (Fig. 5).

Figure 5: 13 November 2017 GOES-16 imagery over Montana/North Dakota/Canada boundaries. Full res
-Bill Line, NWS
“The GOES-16 data posted on this page are preliminary, non-operational data and are undergoing testing. Users bear all responsibility for inspecting the data prior to use and for the manner in which the data are utilized.”
How can I get access to the AWIPS RGBs?
Many of these RGBs can be accessed from the CIRA Slider page: https://col.st/ojVVC