6

Does anyone have or know a citation for the Landsat 8 striping that affects some scenes (Brazil shown here). It's apparent in all bands (B3 here at 100% stretch) and is present in both RAW and Surface Reflectance imagery. It's impacting an image classification's accuracy and I can't seem to find much info about it in the Known Issues from USGS. I've found this document from a Yale website, but again, can't seem to find a primary source to cite.

Google Earth Engine link to script showing the striping effect at Band 3. The script also allows the export of a subset of the product LC08_L1TP_230066_20170621_20170630_01_T1

enter image description here

  • Did you download it from the USGS Earth Explorer or from the Google Earth Engine ?? – PROBERT Apr 17 '18 at 14:51
  • Google Earth Engine. I provided the link to reproduce it at the bottom. – ThrushJacket Apr 17 '18 at 14:53
  • Can you go back to the USGS Earth Explorer and there should be information there if you see the same scene ? – PROBERT Apr 17 '18 at 14:55
  • 3
    Are you absolutely sure that this is OLI data? This is the type of artifact, after correction, that is observed in ETM+7 due to the scan-alignment error. I have never seen this in OLI. I would recommend downloading the same image from earth explorer to see if the artifacts are actually associated with the image or if it is a result of post-processing in GEE. I would imagine that it is possible that this type of banding could result from nearest neighbor sampling during reprojection. – Jeffrey Evans Apr 17 '18 at 15:29
  • 1
    The striping is apparent in Band 3 of LC08_L1TP_230066_20170621_20170630_01_T1. The striping is visible in EarthEngine as well as in an exported subset - see the extended script – Kersten Apr 18 '18 at 12:24
2

This is so called "Memory effect" and it has been around since the sensor is in orbit. It is quite common for optical space-borne sensors, you can see similar effect in Sentinel-2 MSI and Sentinel-3 OLCI sensors.

  • 1
    I do not think this is memory effect, which would manifest as dark and white stripes perpendicular to the direction of satellite travel. – lpd2 Jun 8 '18 at 15:58
1

Looks more like different OLI sensor segments. This usually can be seen over water, but like in this case sometimes it persists on images taken over land.

  • I don't understand why this was downvoted. It seems to be the most correct answer here. Here is a reference: landsat.usgs.gov/… – iled Jul 30 '18 at 4:06
1

Morfitt et al. 2015 (Remote Sensing) provides a description of the effect and its cause. See Figure 1 for Focal Plane Module (FPM) layout and Section 3.2.2 for causes of discontinuities across FPM boundaries (and Figures 8-10 for examples). The effect is "generally uncorrectable."

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.