| Binning Algorithm Specification |
|
Algorithm Specification of the science processors of the MERIS/(A)ATSR Toolbox
The maximum likelihood binning algorithm is the one originally used by the SeaWiFS L3 processing.
The temporal binning algorithm sums up all measurements Xi,j for a bin j to the following intermediate variables:
where w is the weight coefficient specified for the processing and n is the number of measurements for a specific bin. Additionally, the number of measurements n and the weight is saved to the bin database.
The temporal binning just sums up these "per scene" intermediate variables to the final bin database:
The final process converts the variables stored in the temporal bin database to the final L3 variables according to:
First calculate two intermediate values
Which the compose the final variables MEAN, SIGMA, MEDIAN and MODE:
The Arithmetic Mean algorithm calculates the arithmetic mean of the input measurements. Additionally, the user has the option to apply a weighting to the measurements.
The temporal binning algorithm sums up all measurements Xi,j for a bin j to the following intermediate variables:
where w is the weight coefficient specified for the processing and n is the number of measurements for a specific bin. Additionally, the number of measurements n and the weight is saved to the bin database.
The temporal binning just sums up these "per scene" intermediate variables to the final bin database:
The final process converts the variables stored in the temporal bin database to the final L3 variables according to:
The Minimum/Maximum algorithm traces the minimum and maximum value of the measurements for the bin. This algorithm has no distinction between spatial and temporal binning. The final L3 variables are composed as follows:
