Why do the arrays have different sizes?
This is due to the fact that a number of size relevant input parameters used in individual profile retrievals may vary from one elevation scan to the next, e.g., because of latitude dependencies or due to presence of corrupted Level 1B input data. Examples are: – choice of retrieval microwindows and occupation matrices – choice of retrieval grid – total no. of processed spectral data points … Due to the large variabilities of some parameters (e.g., no. of processed spectral points) it is not practicable to fit processing outputs in standard size arrays. Note that all size relevant parameters are reported in the structure ADS, which allows to compute the byte offset of any MDS / ADS data record within a product.
