You may have noticed that high-altitude cloud features (thunderstorms) in your GOES-16 imagery appear displaced (to the north and east or west) from other data sets such as radar and lightning. This is not something unique to GOES-16 imagery, and has always been a known effect in satellite imagery. The displacement is due to to a phenomena known as parallax, and is diagrammed/explained below.

Diagram depicting parallax error. Satellite is over equator, and looking north at a thunderstorm. Top of thunderstorm will appear further north then it actually is. Feature not to scale.

The geostationary satellite is located over the equator, and therefore objects it is viewing over the CONUS are located well north of the satellite sub-point. Tall objects such as thunderstorms will be placed further north (in the imagery) than they actually are due to the parallax effect, typically up to around 10 km. The parallax also exists in the W-E direction. Currently, the GOES-16 satellite is located at 89.5 degrees West, roughly the same longitude as the Mississippi River. Elevated objects to the west will appear further west than they actually are, and objects to the east will appear further east than they actually are.

The parallax error grows as you move further away from the satellite (N-S and W-E), and for taller objects. Given the current position of GOES-16, a thunderstorm in Oregon will have a greater parallax error (north and west) than the same storm in Iowa (north) which will have a greater error than the same storm in Louisiana (north). The parallax effect exists for all satellite channels/bands.

Below is an example of a thunderstorm in Nebraska early on 19 April 2017 poking through a layer of high cirrus clouds. The image includes GOES-16 visible imagery with radar and lightning data overlaid. You can see that the core/peak of lightning activity is occurring over the radar reflectivity core, as expected. However, the texture in the satellite imagery, indicative of the updraft region, is shifted to the north and slightly west of the radar and lightning features. This shift is due to the parallax effect, and is an expected result.

19 April 2017 GOES-16 VIS, national radar mosaic, Earth Networks cloud to ground (+/-) and in cloud (dots) lightning. Full resolution: https://satelliteliaisonblog.files.wordpress.com/2017/04/20170419_nebraska_convection_anno.gif

There are algorithms (not currently available) that make corrections for parallax. However, these will lead to at least slight delays in imagery availability due to processing. Additionally, they introduce issues of their own. Parallax corrected GOES-14 (105W) 1-min imagery was tested by NWS forecasters in the NOAA Hazardous Weather Testbed during the 2016 Hazardous Weather Testbed in Norman, OK. In the example below in Virginia (north and east of satellite sub-point), the correction correctly places the thunderstorm top over the core of lightning activity (correction is to the south and west).

Comparison of corrected and uncorrected GOES-14 visible imagery in Virginia. Also plotted are Earth Networks Cloud to Ground and Cloud Flash’s.

However, the correction can mask important features that are occurring at the low levels, where the parallax error is minimal. In the example below in North Carolina, the correction to the south and west masked flanking line development that could otherwise be seen on the southwest portion of the storm. Additionally, blocky artifacts are introduced into the imagery with the correction.

Comparison of corrected and uncorrected GOES-14 visible imagery in North Carolina.

The parallax error is nothing new, but it is important to keep in mind that this displacement of satellite imagery does exist.

Example below from May 4, 2017 of severe convection over the Pacific Northwest, far from the satellite sub-point. Actual storm location determined from radar and lightning data is to the south and east of deep updraft region seen in satellite imagery (bubbling, overshooting tops).

-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.”