Archives

All posts for the day March 23rd, 2023

GOES-18 Barcode Artifact – March 2023 Update

Posted by Bill Line on 03/23/2023
Posted in: Uncategorized. Leave a comment

GOES-18 has been operating as the operational GOES-West Satellite since 4 Jan 2023. A previous (July) blog post characterized the Barcode Artifact (BA), which is present in GOES-18 Ch07 (3.9 um), and most visibly impacts the appearance of imagery that leverages Ch07 in a band difference, most notably the Fog Difference and Nighttime Microphysics RGB Imagery. To summarize, the BA noise appears as N-S oriented warm and cold stripes and is most apparent in cold scenes. While users may notice the BA in imagery, there is no impact on feature detection and therefore minimal impact to NWS. Since the July blog post, the character of the Barcode Artifact has evolved. This blog post will highlight the history of the GOES-18 Barcode Artifact, including its very recent appearance (as of 3/23/2023).

A Satellite Book Club (SBC) webinar was presented on April 6 on this subject. See Recording here.

The July blog post exemplified the BA appearance prior to 9/22/2022. On 9/22, the BA became “less visible”, as is shown in Figure 1, which compares the Nighttime Microphysics RGB on 9/17 and 9/24. Notice the vertical striping on the left become much more subtle on the right.

Figure 1: 17 Sep 2022 (left) and 24 Sep 2022 (right) GOES-18 Nighttime Microphysics RGB over western US, capturing the change in BA appearance that occurred on Sep 22.

On 12/25/2022, the character of the BA appearance changed again, and is shown in Figure 2. Notice that while the vertical striping is present in both examples, occasional “more narrow/compact” stripes appear in the example after 12/25. The BA has appeared more or less the same since 12/25. Additional examples highlighting the current (as of 3/23/2023) appearance of the BA in a variety of imagery products are shown below.

Figure 2: 18 Dec 2022 (left) and 25 Jan 2023 (right) GOES-18 Nighttime Microphysics RGB over Alaska, capturing change in BA appearance that occurred on Dec 25.

Figure 3 captures the appearance of BA overnight in March 2023 for three of the most used B07 imagery products, including B07 alone, 10.3 – 3.9 Fog Difference, and Nighttime Microphysics RGB. The BA is barely noticeable in the single channel imagery, but is occasionally apparent across the broad scene in both the Fog Difference (10.3 – 3.9 um) and the Nighttime Microphysics RGB (which leverages the Fog Diff as the “green” component of the RGB). Notice that the vertical striping is most apparent within the high “cold” clouds.

Figure 3: 22 March 2023 GOES-18 ch07 (left), Fog Diff (center), Nighttime Microphysics RGB (right) over western US.

When we focus in on a smaller geographic area, such as a NWS County Warning Area (SEA), the BA becomes even less apparent during the overnight period, as the vertical stripes are few and far between (Fig 4). Although occasionally noticeable, analysis of low clouds near Seattle is not impacted by the BA.

Figure 4: 22 March 2023 GOES-18 Nighttime Microphysics RGB over Washington state.

Moving further north into an even colder and cloud-ridden airmass, BA may be slightly more apparent overall, but analysis of any given localized cloud can still be completed (Fig 5 and 6).

Figure 5: 20 March 2023 GOES-18 ch07 (left), Fog Diff (center), Nighttime Microphysics RGB (right) over Alaska.
Figure 6: 20 March 2023 GOES-18 Nighttime Microphysics RGB over Alaska.

The Day Snow Fog RGB is a popular RGB leveraged for daytime cloud analysis, and includes the Fog Difference as the “blue” component of the RGB recipe. Despite including the fog difference, no BA is visible when using this RGB since it is leveraged during the daytime (Fig 7).

Figure 7: 20 March 2023 GOES-18 Day Snow Fog RGB over western US.

Similarly, the Day Convection RGB, which also uses the Fog Difference, does not contain any BA influence (Fig 8).

Figure 8: 20 March 2023 GOES-18 Day Convection RGB over western US.

One may have a (valid) concern about wildfire hot spot detection being impacted by BA in Ch07. This is not the case, as we remember that BA is primarily apparent in cold scenes (in clouds at night) in Band 07 differences. Figure 9 compares Ch07 and the Fire T RGB (which uses ch07 as the “red” component), and confirms no BA influence. Fires can continue to be tracked into the evening in either of these two imagery products, with striping very unlikely to be visible in the scene, as fires are surface (warm) features and use only the single band imagery.

Figure 9: 18 March 2023 GOES-18 Ch07 (left) and Fire Temperature RGB (right) over western US.

Zooming in on southeast Washington in the Fire Temperature RGB, no artifacts are visible, and the hot spots are diagnosed per usual (Fig 10).

Figure 10: 18 March 2023 GOES-18 Fire Temperature RGB over southeast Washington state.

A full PACUS sector overnight Fog Difference example shows the BA most visible in cold high clouds, less visible in cold low clouds and over clear-sky cold airmass like in the western US, and least visible over the warmer airmass around Hawaii (Fig 11).

Figure 11: 7 Feb 2023 GOES-18 Nighttime Fog Difference over PACUS.

As has been noted and is shown in these examples, a users’ ability to detect and track clouds across their forecast area using GOES-18 imagery is not significantly impacted by BA. It is important to note that the appearance of BA may continue to change in the future.

Bill Line, NESDIS/STAR