This readme file describes the dataset: Data from the Cornell coherent scatter radar imager.

This data is shared openly with the public, but the author requests that anyone re-using the data please contact him before re-use and provide proper attribution.

The dataset was updated in March 2024, to include 2022 and 2023 data (in 4 .date files, and one compressed zipped file with all 2022 and 2023 data), in addition to the 2021 and 2020 data.

The dataset was updated in December 2022, to include 2021 data (in 5 .dat files, or one compressed, zipped file with all 2021 data), in addition to the 2020 data.

Please cite this dataset as:

Hysell, David L. (2024) Data from the Cornell coherent scatter radar imager [Dataset]. Cornell University Library eCommons Repository. https://doi.org/10.7298/8sxf-b977.3

These files contain data acquired by the Cornell radar imager. They
are binary files written in single-precision little-endian format. The
files are divided into header records and data records which are
interleaved.

Header records can be read as (for example in C):

read(f_in,&tstamp,sizeof(int));
read(f_in,&shead,sizeof(shead));
read(f_in,used,6*sizeof(int));
read(f_in,&navg,sizeof(int));
read(f_in,&nbins,sizeof(int));

where 'tstamp' is an integer and contains the time in linux time format
and 'shead' is a header structure defined by:

typedef struct {
  float txa;
  float ipp;
  float h0;
  float dh;
  int nsamp;
  int ncode;
  unsigned int code[20];
} short_head;

Here, 'txa' is the pulse length in km, 'ipp' is the interpulse period
in km, 'h0' is the first sample range in km, 'dh' is the range
increment in km, 'nsamp' is the number of range samples, 'ncode' is
the number of distinct pulse codes used, and 'code' contains the actual
binary phase codes transmitted in binary format.

Also, 'used' is an integer array which contains 0's or 1's depending
on whether each one of six receivers was used for the experiment. The
number of receivers N is therefore the total number of 1's, and the
number of interferometry baselines is NB = N(N-1)/2. Finally, 'nbins'
is the number of spectral bins used, and 'navg' is the number of
spectra incoherently averaged.

Spectral and cross-spectral data for each receiver and baseline may be
read into arrays defined respectively as follows:

power=(float*)malloc(sizeof(float)*shead.nsamp*nbins*N);
product=(fcomplex*)malloc(sizeof(fcomplex)*shead.nsamp*nbins*NB);

Filenames specify the UT date of the experiment, and extensions run
serially in order starting from 000.

For further information about these data, consult:

Hysell, D., Larsen, M., Fritts, D. et al. Major upwelling and
overturning in the mid-latitude F region ionosphere. Nat Commun 9,
3326 (2018). https://doi.org/10.1038/s41467-018-05809-x

or contact:

David Hysell
3114 Snee Hall
Cornell University,
Ithaca, NY 14853 USA
david.hysell@cornell.edu

This research is sponsored by NSF award AGS-2011304 to Cornell
University.