Ongoing projects
Auxiliary Data for a Better Characterisation of MIPAS Level 2 Products:
measurements from January 2005 on
measurements from January 2005 on
Averaging Kernels
The Averaging Kernels for the 2002-2004 measurements were computed a-priori for different standard atmospheres.
For the measurements from January 2005 on, the AKs are provided for each retrieved profile in the Level 2 products. This is possible because with the new measurements regularization is needed to reduce instability of the retrieval (the vertical sampling grid is finer than the IFOV of the instrument, and retrieval is performed in correspondence of the measurement sampling grid). In this case, the Averaging Kernels computed on the retrieval grid are not equal to the Identity matrix, as is the case when no regularization is applied, and hence the computation of the Averaging Kernel is not too much time computing.
For the measurements from January 2005 on, the AKs are provided for each retrieved profile in the Level 2 products. This is possible because with the new measurements regularization is needed to reduce instability of the retrieval (the vertical sampling grid is finer than the IFOV of the instrument, and retrieval is performed in correspondence of the measurement sampling grid). In this case, the Averaging Kernels computed on the retrieval grid are not equal to the Identity matrix, as is the case when no regularization is applied, and hence the computation of the Averaging Kernel is not too much time computing.
Systematic Errors
The systematic errors have been estimated by Oxford University and are downloadable here.
For a description of the error analysis performed by Oxford University we refer to the website: http://www.atm.ox.ac.uk/group/mipas/err/.
For a description of the error analysis performed by Oxford University we refer to the website: http://www.atm.ox.ac.uk/group/mipas/err/.
pT error propagation matrices
Still not available