SVRIMG - SeVere Reflectivity IMaGe Dataset

Raw Reflectivity Data in 136 x 136 Images

Home Classify Data Machine Learning Notebooks (GitHub) Mean Storms

Data Methodology

Radar images are centered on SPC severe weather reports and extracted from the closest hourly data in GridRad which can be downloaded from the Research Data Archive. The original ~2x2km 3D data are converted to 2D by calculating the column maximum reflectivity. These values are then converted to 8-bit integers and interpolated to a 3.75 km Lambert conformal conic grid using nearest neighbor. The 136 x 136 dimensions result in a region approximately 512 x 512 km.

Access the Data

136 x 136 pixel (~512 x 512 km) images with units of dBZ. The center of the radar image is the location of the report. You can download the raw data, view individual maps centered on reports, or view all reports on an interactive map for each year.

Year Images and CSVs Browse Single Report Maps Browse Annual Report Maps
1996 1996_svrimg_raw.tar.gz 1996 Single Report Maps 1996 Annual Report Maps
1997 1997_svrimg_raw.tar.gz 1997 Single Report Maps 1997 Annual Report Maps
1998 1998_svrimg_raw.tar.gz 1998 Single Report Maps 1998 Annual Report Maps
1999 1999_svrimg_raw.tar.gz 1999 Single Report Maps 1999 Annual Report Maps
2000 2000_svrimg_raw.tar.gz 2000 Single Report Maps 2000 Annual Report Maps
2001 2001_svrimg_raw.tar.gz 2001 Single Report Maps 2001 Annual Report Maps
2002 2002_svrimg_raw.tar.gz 2002 Single Report Maps 2002 Annual Report Maps
2003 2003_svrimg_raw.tar.gz 2003 Single Report Maps 2003 Annual Report Maps
2004 2004_svrimg_raw.tar.gz 2004 Single Report Maps 2004 Annual Report Maps
2005 2005_svrimg_raw.tar.gz 2005 Single Report Maps 2005 Annual Report Maps
2006 2006_svrimg_raw.tar.gz 2006 Single Report Maps 2006 Annual Report Maps
2007 2007_svrimg_raw.tar.gz 2007 Single Report Maps 2007 Annual Report Maps
2008 2008_svrimg_raw.tar.gz 2008 Single Report Maps 2008 Annual Report Maps
2009 2009_svrimg_raw.tar.gz 2009 Single Report Maps 2009 Annual Report Maps
2010 2010_svrimg_raw.tar.gz 2010 Single Report Maps 2010 Annual Report Maps
2011 2011_svrimg_raw.tar.gz 2011 Single Report Maps 2011 Annual Report Maps
2012 2012_svrimg_raw.tar.gz 2012 Single Report Maps 2012 Annual Report Maps
2013 2013_svrimg_raw.tar.gz 2013 Single Report Maps 2013 Annual Report Maps
2014 2014_svrimg_raw.tar.gz 2014 Single Report Maps 2014 Annual Report Maps
2015 2015_svrimg_raw.tar.gz 2015 Single Report Maps 2015 Annual Report Maps
2016 2016_svrimg_raw.tar.gz 2016 Single Report Maps 2016 Annual Report Maps
2017 2017_svrimg_raw.tar.gz 2017 Single Report Maps 2017 Annual Report Maps

Metadata

Geogrid parameters:

Description of the columns in the CSV are as follows:

The unid column is based on the report time in UTC and "om" identifier from the "1950-2018_actual_tornadoes.csv" file on the SPC severe weather reports page. You should be able to easily join these data with those data to attach more attributes.

Disclaimers and Caveats

The data are provided at no cost, as-is, with no warranty of any kind. No modification of either the SPC reports or the GridRad data (beyond interpolation) is done before these data are hosted on the website. The process is completely repeatable from start to finish, assuming you have patience or access to a supercomputer cluster. Please examine the GridRad and the SPC severe weather reports pages to read about the caveats and issues with those data before using these data.

Known issues include:

Severe reports closer to 30 minutes after the hour will have images with displaced reflectivity and the level of displacement is dependent on storm motion.
No attempt is made to remove reports with incorrect positions or those that have no radar data. A csv is provided that includes basic radar reflectivity statistics which make it possible to filter these events.
The 3.75 km LCC Grid has larger grid spacing than the original data. This will result in some detail being lost. This was a decision made to reduce the size of the dataset on disk and on GPU.
The raw data are originally "upside down" because of how the grid was defined. The image browser on this website provides flipped images. However, the tarball data will have the original orientation that matches up with the LCC grid. Please take the time to examine what you are throwing into your ConvNet!

Data Citations

We are generating these data solely because we think they would be of interest to the meteorology and climatology community. That being said, we would like to get some credit if you find them useful!

If using these data in a paper or project, please cite the methods paper:

Haberlie, A. M., W. S. Ashley, and M. Karpinski, 2020: Mean storms: Composites of radar reflectivity images during two decades of severe thunderstorm events. International Journal of Climatology, In Press.

Please cite the GridRad dataset as well:

Bowman, K. P., and C. R. Homeyer. 2017. GridRad - Three-Dimensional Gridded NEXRAD WSR-88D Radar Data. Research Data Archive at the National Center for Atmospheric Research, Computational and Information Systems Laboratory.