Using Landsat Data - 3 of 14
Using Landsat Data FAQs - 14 Found
The percentage of cloud cover is calculated when raw mission data from Landsat Thematic Mapper (TM) and Enhanced Thematic Mapper Plus (ETM+) are processed to Level-0R during ingest into the archive using the Automatic Cloud Cover Assessment (ACCA) algorithm. This algorithm uses a number of spectral tests and is based on the premise that clouds are colder than Earth surface features. The ACCA, however, does not perform as well over snow and ice due to the cold surfaces.
ACCA works well in most cases, but occasionally, temperature inversions occur and invalid cloud cover assessments may result. While there are also intermittent problems with the ACCA detection of popcorn clouds and haze, the ACCA algorithm maintains a higher degree of overall accuracy than the previous methods employed for past Landsat processing systems.
Changes were made to the ACCA algorithm in January 2009; these are listed below. Most changes apply only to TM data, but the new logic to address misidentified "snow pixels" was applied to both. The TM changes were of sufficient magnitude that the transition is detectable, whereas the overall impact on the ETM+ scores is minor.
The changes affect raw mission data ingested into the USGS EROS archive after January 21, 2009. This applies to data received by the Landsat ground stations as well as current or historical TM or ETM+ data received from Internal Cooperator (IC) ground stations through the Landsat Global Archive Consolidation (LGAC) initiative.
Data ingested into the USGS archive prior to January 21, 2009 do not have the updated ACCA algorithm applied. Data acquired prior to January 21, 2009 that were not ingested until after this date will have the new ACCA calculations applied. Efforts to recalculate and make all ACCA calculations for data products consistent have yet to be confirmed.
Changes made to Automatic Cloud Cover Assessment (ACCA) algorithm, January 21, 2009:
- Correct conversion of sun elevation from radians to degrees.
- Correct calculation of the band 6 reflectance look-up table.
- Undo multiplication of Bias value by 10 for radiometric correction.
- Correct two thresholds that were being reset to zero on program startup.
- Correct one threshold that was not being used properly.
- Change five thresholds to match L7 values.
- Correct indexing for reading band 6.
- Correct radiometric correction for band 6.
- Use correct band 6 gain/bias values.
- Calculate ACCA score based on the total pixels used instead of full scene size.
- Correct offsets to account for the left hand 'ragged' edge for band 6.
- Correct indexing into band 6 reflectance array for TMA data.
- Add new logic to address misidentified 'snow pixels'.
- Adjust Band45 Threshold to 0.94.