Isoline Retrieval uses supervised statistical classification to retrieve isolines from cross-track scanning or similar satellites. It contains software to generate training data using collocation or radiative transfer simulations, as well as routines to interpolate the final fields using a variation of multi-linear interpolation or kernel estimation. The currently-supported satellites are the Advance Microwave Sounding Unit (AMSU) series and, to a lesser extent, the Global Ozone Measurement Experiment (GOME). An ambitious researcher, however, could easily adapt the codes to a similar satellite.
|Tags||inverse methods Statistics supervised learning kernel estimation atmospheric physics remote sensing|
|Operating Systems||OS Independent|
|Implementation||C++ IDL shell scripting Perl|
For those interested in learning more about the method, it is fully described in the following article: Peter Mills (2009). "Isoline retrieval: An optimal sounding method for validation of advected contours." Computers & Geosciences, 35(10):2020-2031. DOI:10.1016/j.cageo.2008.12.015. It can be downloaded from the arXiv.org preprint database: http://arxiv.org/abs/1202.5659 Or from my own homepage: http://peteysoft.org/Mills2009.pdf