I want to classify aerosol types from CALIPSO- CALOP data. Which software is suitable for analysis of this data and for processing suggest some literature?
On the CALIPSO pages you can find not only the data, but also a lot of documentation and tools for MATLAB and IDL. Unfortunately, although MATLAB is easy to handle, it is not free (far from it!).
Try reading this first:
http://www-calipso.larc.nasa.gov/resources/calipso_users_guide/essential_reading/index.php
Then look here:
http://www-calipso.larc.nasa.gov/resources/calipso_users_guide/tools/index.php
Software is only a tool; depends on what actually you would like to do. How do you propose to classify the aerosol types? The tool should be selected based on the physics of the scheme
Mr. Krishna Moorthy - I assume that the aerosol types meant are those defined in the CALIOP retrieval algorithm, which would be clean marine, clean/polluted dust, clean/polluted continental, and smoke. These types are based on the detected depolarization ratio and color ratio, but in cases where the measurement does not contain enough information, (also) on a climatology.
Level-2 CALIOP data can most conveniently be read using tools developed for MATLAB or IDL (as mentioned above), because the data structure is not very intuitive. Maybe these tools can be easily adapted to Octave (or Python?), but I don't have any experience with that.
A word of warning, though: CALIOP results are not always as reliable as they may appear! The extinction and aerosol type suffer grievous errors under certain circumstances. There are several papers on these problems in JGR (I think) - try looking for papers (co-) authored by D. Winker. Here's one from ACP:
http://www.atmos-chem-phys.net/11/3981/2011/acp-11-3981-2011.html
Hello Marioes, The last paragraph of your comment is the most important fact. All operational algorithms are developed considering certain zero-order features, which could be valid under ideal conditions.. like ideal dust/ ideal sea-salt etc. Operational algorithms are meant for quick look products, which might form ancillary data, rather than primary. So to infer aerosol types from the spectral extinction and depolarization, one should really consider the physics behind the process (scattering and absorption) and how it would manifest in the backscattered lidar signals.
Indeed - and don't forget the terrible signal-to-noise ratio of CALIOP! I (naively) thought that CALIOP would essentially provide the answers to all our aerosol questions until I started working with the data. That doesn't mean that it isn't good --- just not as good as I thought!
- Badges
- Science topic
- Similar topics
- Geoscience
- Remote Sensing