This is an approximately true color composite image of the contiguous United States as captured by GOES-16 on May 17, 2017 (17:15:00 UTC) and remapped to Mercator projection for SSEC RealEarth.
Since there is no green filter on GOES-16 it is necessary synthesize the green channel from blue, red, and near infrared bands. The method planned is far more sophisticated than what was done here. I assume the process will be made available but I expect it will require specialized software to implement. I also expect most of the GOES-16 imagery presented to the general public will have atmospheric correction done (hopefully with the reddish coloration fixed).
It has its usefulness but it doesn't give an accurate impression of how the planet actually appears. With the introduction of Himawari 8 and now GOES-16 it is possible to produce high quality, high resolution imagery of the full disk of Earth and its atmosphere in very close to true color. Inexplicably there doesn't seem to be much importance placed on this so it may be left up to enthusiasts to do rather than the scientific community.
In the meantime some pretty interesting results can be achieved with nothing more than a paint program like Photoshop. Using the blue band as the base for the green channel the red band and near infrared band are added as layers with their opacities set to 45% and 5% respectively. Then the blue, simulated green, and red channels are combined to make a RGB image. Once proper hue, saturation, color balance, and contrast adjustments layers are applied the result can be hard to distinguish from imagery that does include the green band.
The image in the left panel was made by combining blue, green, red, and near infrared bands from Himawari 8. The right panel image uses a synthesized green band. The same technique was incorporated to make the GOES-16 image at the top of this post. One shortcoming is shallow waters like those around the Bahamas will not have very accurate coloration because the green signal is entirely absent. This issue is being addressed in the proposed method mentioned previously.
Finally, here is the US image with pseudo atmospheric correction applied.
Like all the visualization work shown in this blog it is intuitive rather than scientific. It mimics the appearance of Rayleigh-corrected MODIS and VIIRS data as seen on NASA WorldView and other places but it is not intended to be directly comparable. It will be interesting to see if synthesized true color GOES-16 imagery will be able to reproduce all the yellow, orange, and pink color variations in the deserts shown in this VIIRS composite.
CREDIT
NASA
National Oceanic and Atmospheric Administration (NOAA)
Space Science and Engineering Center (SSEC) University of Wisconsin-Madison
Meteorological Satellite Center (MSC) of Japan Meteorological Agency (JMA)
National Institute of Information and Communication Technology (NICT)
RESOURCES
Virtual Green Band for GOES-R
GOES-R Advanced Baseline Imager (ABI) True-Color Capability
http://www.goes-r.gov/resources/Scipubs/docs/2011/053520_1.pdf (broken)
 |
| Credit: NASA / NOAA / SSEC - Modified by LoneSky. |
Since there is no green filter on GOES-16 it is necessary synthesize the green channel from blue, red, and near infrared bands. The method planned is far more sophisticated than what was done here. I assume the process will be made available but I expect it will require specialized software to implement. I also expect most of the GOES-16 imagery presented to the general public will have atmospheric correction done (hopefully with the reddish coloration fixed).
It has its usefulness but it doesn't give an accurate impression of how the planet actually appears. With the introduction of Himawari 8 and now GOES-16 it is possible to produce high quality, high resolution imagery of the full disk of Earth and its atmosphere in very close to true color. Inexplicably there doesn't seem to be much importance placed on this so it may be left up to enthusiasts to do rather than the scientific community.
In the meantime some pretty interesting results can be achieved with nothing more than a paint program like Photoshop. Using the blue band as the base for the green channel the red band and near infrared band are added as layers with their opacities set to 45% and 5% respectively. Then the blue, simulated green, and red channels are combined to make a RGB image. Once proper hue, saturation, color balance, and contrast adjustments layers are applied the result can be hard to distinguish from imagery that does include the green band.
 |
| Credit: MSC of JMA / NICT / SSEC - Modified by LoneSky. |
The image in the left panel was made by combining blue, green, red, and near infrared bands from Himawari 8. The right panel image uses a synthesized green band. The same technique was incorporated to make the GOES-16 image at the top of this post. One shortcoming is shallow waters like those around the Bahamas will not have very accurate coloration because the green signal is entirely absent. This issue is being addressed in the proposed method mentioned previously.
Finally, here is the US image with pseudo atmospheric correction applied.
 |
| Credit: NASA / NOAA / SSEC - Modified by LoneSky. |
Like all the visualization work shown in this blog it is intuitive rather than scientific. It mimics the appearance of Rayleigh-corrected MODIS and VIIRS data as seen on NASA WorldView and other places but it is not intended to be directly comparable. It will be interesting to see if synthesized true color GOES-16 imagery will be able to reproduce all the yellow, orange, and pink color variations in the deserts shown in this VIIRS composite.
 |
| Credit: NASA |
CREDIT
NASA
National Oceanic and Atmospheric Administration (NOAA)
Space Science and Engineering Center (SSEC) University of Wisconsin-Madison
Meteorological Satellite Center (MSC) of Japan Meteorological Agency (JMA)
National Institute of Information and Communication Technology (NICT)
RESOURCES
Virtual Green Band for GOES-R
GOES-R Advanced Baseline Imager (ABI) True-Color Capability
http://www.goes-r.gov/resources/Scipubs/docs/2011/053520_1.pdf (broken)
Comments
Post a Comment