Computation
Infoterra MERIS HiProGen system to generate L3 composites of MERIS L2 data deployed on unknown computer
Abstract
This computation involved: Infoterra MERIS HiProGen system to generate L3 composites of MERIS L2 data deployed on unknown computer. Infoterra Ltd, backed by the European Space Agency (ESA) and the British National Space Centre (BNSC), has set up a Level 3 product generation and dissemination service.
The production service is based around the High level Product Generation (HiProGen) system which was initially developed under an ESA funded Research and Technology Development contract. Upgraded to use Near Real-Time data and to work in the UK-PAC operational environment, the HiProGen system is now generating Level 3 composites of MERIS Level 2 geophysical data. Level 3 products are generated in GeoTIFF format, on lat-long geocoded grids, accompanied by an XML metadata file and a JPEG browse image.
The Infoterra Level 3 production service combines individual passes (or swathes) of data from satellite instruments into single composite products. These Level 3 composites generated at global and regional scales using two distinct types of compositing algorithm. At the global scale, large (nominally 4.6km) sample sizes are used with each input Level 2 pixel accumulated into a geographic 'bin'. This is the standard form of compositing carried out where the output bin is much larger than the size of the input instrument samples. Unfortunately at higher scales, where the input and output sample sizes are comparable, the input samples overlap several cells of the output grid to a significant degree. If this is not taken into account then both the fine scale values and the overall product statistics are distorted (i.e. the final product is not flux conserving).
For the regional products, with nominal sample spacings of 1km or less, an area weighted, flux conserving compositing algorithm is used. This algorithm was developed by Tom McGlynn of NASA.
The HiProGen Level 3 Production system makes use of the Basic ENVISAT (A)ATSR and MERIS (BEAM) toolkit. Mathematically, different input data values are combined using a Maximum Likelihood, Arithmetic Mean or Min-Max algorithm. The Infoterra service uses the arithmetic mean algorithm for all Level 3 product generation.
| keywords: | |
|---|---|
| inputDescription: | None |
| outputDescription: | None |
| softwareReference: | None |
| Previously used record indentifiers: |
http://badc.nerc.ac.uk/view/neodc.nerc.ac.uk__ATOM__dpt_11750991626419753
http://badc.nerc.ac.uk/view/neodc.nerc.ac.uk__ATOM__obs_11737171090917008
|
More Information (under review)
The Infoterra Level 3 production service combines individual passes (or swathes) of data from satellite instruments into single composite products. These Level 3 composites generated at global and regional scales using two distinct types of compositing algorithm. At the global scale, large (nominally 4.6km) sample sizes are used with each input Level 2 pixel accumulated into a geographic 'bin'. This is the standard form of compositing carried out where the output bin is much larger than the size of the input instrument samples. Unfortunately at higher scales, where the input and output sample sizes are comparable, the input samples overlap several cells of the output grid to a significant degree. If this is not taken into account then both the fine scale values and the overall product statistics are distorted (i.e. the final product is not flux conserving). For the regional products, with nominal sample spacings of 1km or less, an area weighted, flux conserving compositing algorithm is used. This algorithm was developed by Tom McGlynn of NASA. The HiProGen Level 3 Production system makes use of the Basic ENVISAT (A)ATSR and MERIS (BEAM) toolkit. Mathematically, different input data values are combined using a Maximum Likelihood, Arithmetic Mean or Min-Max algorithm. The Infoterra service uses the arithmetic mean algorithm for all Level 3 product generation.