CAMA toolkit

Welcome to the home page of the CAMA toolkit, otherwise known as “the algorithm torture toolkit”. The CAMA toolkit helps to quickly analyse fairly large volumes of OMI level-2 data.


The purpose of CAMA is to correlate various parameters stored in OMI level-2 granules (orbit files). In addition it will gather statistics on the individual parameters, and create a series of graphs to identify potential issues in your level-2 files. Note that overpass files from the AVDC contain level-2 data, but are not useful to CAMA.

How it works

The CAMA tool is written in IDL and intended to be flexible and versatile. The user configures the analysis in a text file, and CAMA does the processing. The configuration allows for reading from any OMI level-2 product. CAMA can also perform simple arithmetic operations to combine several fields, for example subtracting OMI DOAS O3 and OMI TOMS O3 columns. It can also apply user-written filter functions, for example to select only measurements over sea. Several filters are supplied in the standard distribution.

CAMA calculates comparisons and correlations between parameters for each indicated quantity. This is done for each of the OMI ground pixels and all orbits within the user-specified limits. Also time-averages, statistics, and plots of the distributions are calculated. The results are written to a series of encapsulated postscript graphs (eps) and summarised in a LaTeX report. A feature was recently added to allow viewing of the geolocated averages in Google Earth.


The toolkit is written to be memory efficient, and uses two passes: A first pass to calculate the averages, and a second pass to calculate the (co)variances to avoid the need for keeping all data in memory at once. With a limited number of parameters (6) a year of data (5316 orbits) takes about 16 hours to process on the TLCF cluster. In total this set sampled about 500 milion OMI measurements. Our experience with the O2-O2 cloud product has shown that by processing a single day of OMI data, 15 orbits, the probability distribution functions no longer change significantly, but for other OMI products this may be very different.

You can download a recent EGU poster of CAMA here: EGU 2011 Poster on CAMA.

How to obtain the CAMA software

The links in the sidebar contain links to the cama documentation and the software. CAMA has a license, which you may want to read before using the software. There is a page with installation notes (tested on the TLCF). A short overview of the history of CAMA is given on the history page, with future plans separated to the enhancements page.

Contact information

Please note that we do not provide support for CAMA. However, we are interested in your experiences, and like to receive feedback on the software and its documentation. If you encounter problems, we will try to resolve those, or offer a way to work around them. We are happy to receive code to fix issues, additional filter functions, et cetera.

The main author of CAMA is .

© OMI -- Last update: Tuesday, 09-Aug-2011 01:35:18 UTC. --