RAWACF to FITACF

SuperDARN data is distributed to the community in the rawacf format. For most science applications you will want to process these files into the fitacf format using the make_fit routine.

Choosing a fitting algorithm

Several methods have been developed by the SuperDARN community to determine the physical properties of SuperDARN backscatter (power, velocity, spectral width, elevation, and their errors).

The fitting algorithms available for processing SuperDARN data are:

Name	Notes
fitacf3	Suitable for all scientific applications. Fitted data may appear 'noisy' for some radars due to operational problems (remove with fit_speck_removal). Released: 2017 References: (1) FitACF 3.0 White Paper, (2) SuperDARN noise estimation
fitacf2	Suitable for all scientific applications. Generally results in fewer fitted ACFs compared to fitacf3. Released: 2006 Reference: Annales Geophysicae, 24, 115–128, 2006
lmfit2	Suitable for scientific applications of fit-level data. Levenburg-Marquardt fitting of SuperDARN auto-correlation functions (ACFs). ACF fitting is performed for all range gates. No elevation angles. Takes ~5-10 minutes to process 1 hour of data. Call using separate binary, make_lmfit2 Released: 2018 Reference: Radio Science, 53, 93-111, 2018
lmfit1	Suitable for scientific applications of fit-level data. A model complex ACF (single component, exponential decay) is fitted to the observed one using the Levenberg-Marquardt algorithm. No elevation angles. Released: c.a. 2012 Reference: Radio Science, 48, 274–282, 2013
fitex2	Intended for use with the tauscan multipulse sequence. Phase fitting performed with 120 phase variation models. Released: c.a. 2012 Reference: Radio Science, 48, 274–282, 2013
fitex1	Original fitex algorithm. Not thoroughly tested. No elevation angles.

Basic syntax

make_fit -fitacf3 [inputfile].rawacf > [outputfile].fitacf3

make_fit -fitacf2 [inputfile].rawacf > [outputfile].fitacf2

Use the -vb flag to print some information in the terminal during the data processing.

Supplying TDIFF values

By default, RST reads TDIFF from each radar's hardware file.

To supply TDIFF values from a separate calibration file, use make_fit -tdiff_method [method]. A sample calibration file is available in $RSTPATH/tables/superdarn/tdiff/tdiff.dat.tst.

The user can also supply a fixed TDIFF value in microseconds using make_fit -tdiff [tdiff]

Type make_fit --help in the command line for more information.

Old format files

The Data Analysis Working Group recommends converting old-format dat files to rawacf format before calling make_fit. For example,

dattorawacf 2002020202f.dat > 2002020202f.rawacf

make_fit -fitacf2 2002020202f.rawacf > 2002020202f.fitacf2

It is also possible to process dat files directly using make_fit -old, which will produce an old-format fit file as output. However, lossy compression associated with the old-format fit files will lead to small differences in the fitted parameters compared to the recommended procedure shown above.

make_fit -old -fitacf2 [inputfile].dat [outputfile].fit

Convert multiple files

To perform the rawacf to fitacf conversion for all files in the current directory:

for file in *.rawacf
do 
  make_fit -fitacf3 -vb $file > "${file%.rawacf}.fitacf3"
done

You can then concatenate the 2-hour fitacf files into a daily file, for example:

cat 20181001.*.lyr.fitacf3 > 20181001.lyr.fitacf3