Documentation on IEM generated NEXRAD composites


Example composite from 1 March 2007.

Quick Information:

N0RN0Q
Product: 4 bit Base Reflectivity 5dbz 8 bit Base Reflectivity 0.5dbz
Format: PNG image (GeoTiff available)
Size: 6000x2600 pixels 12000x5200 pixels
12200x5400 after 8 Aug 2014
Time: Every 5 minutes since 1 Jan 1995 Every 5 minutes since 13 Nov 2010 16:25 UTC
Download: Current and Archive for today
Precipitation N0R (4bit) Reflectivity N0Q (8bit) Reflectivity
Ramp Index values [inch] RampIndex values [dBZ] RampIndex values [dBZ]
0
0.25
0.5
0.75
1.0
1.25
1.5
1.75
2.0
2.25
2.5
2.75
3.0
3.25
3.5
3.75
4
4.8
5.6
6.4
7.2
8.0
8.8
9.6
10.4
11.2
12
13
14
15
16
17
18
19
20
21
22
23
24+
NODATA
-30
-25
-20
-15
-10
-5
0
5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
NODATA
-32 dBZ
through
90 dBZ
0.5 dBZ increment

IEM Applications using this product

Since 2003, the IEM has been generating and archiving nationwide composites of NEXRAD base reflectivity every 5 minutes. These composites are generated by combining the 130+ individual RADARs that comprise the NEXRAD network. This page provides background information on this dataset and highlights various ways this dataset can be and is used. Please note: Errors can and do occur during the automated generation of these images, please use them with care.

Important changes to this product.

  • 1 August 2003: 5 minute interval archive begins. The initial pixel resolution is roughly ~2 km.
  • 14 November 2005: The pixel resolution is increased to ~1 km thanks to the support of a Unidata Equipment grant.
  • 10 May 2007: Implemented a clutter suppression algorithm to remove most anomalous propogation during the summertime.
  • 11 Nov 2008: Implemented a crude check against RUC model surface temperature analysis for near freezing areas to stop the clutter suppression algorithm from removing areas of winter precipitation.
  • 14 May 2009: Completed a 7 month process of backfilling the composites back to 1 Jan 1995. Also regenerated composites prior to 2006 to produce them on a ~1 km grid as well. Big thanks to Steve Ansari at NCDC for providing the level3 archive.
  • 25 Aug 2010: Reprocessed 2005 + 2006 data to improve product quality and to backfill the archive with clutter suppression enabled during the summer months (Apr-Sep).
  • 16 July 2014: Modified the clutter removal routine to use the higher resolution EET level III product. Previously NET was used.
  • 8 Aug 2014: The spatial domain of the N0Q composite was expanded 1 degree east and south. The reason is that the clipping was occurring on the eastern and southern borders of the domain.
  • 7 Sep 2016 1940 UTC: The original intent of the N0Q composites was to provide imagery at a resolution of 0.5 dBZ. After some investigation, it was discovered that the N0Q composites only contained data at values of whole integer values of dBZ. After some debugging of GEMPAK, the issue was hopefully fixed within GEMPAK. The good news is that this bug did not impact any processing of the file. The conversion of color index and RGB colors to dBZ was OK. There just was not any data at those 0.5 dBZ, non-int values.
  • 21 Aug 2017: The NWS intends on shutting down the N0R product on 31 December 2017. When this happens, the IEM will stop producing the composite. You should consider using the n0q product as its replacement.

Product Generation:


Composite of composites for 18 Oct 2007. Click image for regular size.

The IEM receives a feed of NEXRAD Level III products from the Unidata Internet Data Distribution. This feed includes the Base Reflectivity (N0R) and Net Echo Top (NET) products which are saved to disk for later processing. Every 5 minutes, a script runs a GEMPAK program called nex2img. nex2img searches the on-disk data for current (within 15 minutes of runtime) products and then composites them into a large image.

The result of the nex2img process is a simple raster image with size of 6000 pixels in width by 2600 pixels in height. This raster image is then compared with a similiar raster of NET to remove any suspected false echos. In the wintertime, this comparison is skipped for most of the country. The final result is a PNG formatted image looking much like the example image shown above.

The final image is shipped to the IEM webfarm for archival and use. You can always find the 10 most recently generated rasters here (n0r_0* is the newest). You can also find current and historical images in the archive. The format for the URL is:
https://mesonet.agron.iastate.edu/archive/data/YYYY/mm/dd/GIS/uscomp/
For example, here are the images generated for today: https://mesonet.agron.iastate.edu/archive/data/2017/09/19/GIS/uscomp/
All dates in the archive are in UTC time.

Product Format:

The images are stored in PNG format. They have a 4-bit colormap and are non-interlaced. Images for the past 50 minutes are also available in TIFF, GeoTIFF, and gzipped GeoTIFF format. Most people want to use these images in a GIS, so world files (.wld or .tfw) are available in the same current and archive folders. The projection of these images is in EPSG:4326 (WGS84 longlat).

For some people, the PNG format is not readable by their GIS, so the IEM provides a server-side conversion of the PNG files for delivery of a GeoTIFF. This application is called like so:
https://mesonet.agron.iastate.edu/request/gis/n0r2gtiff.php?dstr=200711010025
where dstr specifies the {year}{month}{day}{hour}{minute} in UTC of the requested raster. Rasters are only available at minutes modulo 5.

Colormap Information:

The raster image has a colormap associated with it. The raw pixel values increase monotonically from -30 dbz to 75 dbz every 5 dbz. All values below 0 dbz are assigned black (rgb 0 0 0). We have tried to match the commonly used NEXRAD display as what you see on National Weather Service page. Please note: We do not include a concept of missing data or "no-data". So spots with beem blockage in the mountains or out away from the CONUS are just "black".

Web Map Service (WMS)

A special way you can interface with this data archive is via the Web Map Service that also supports time based retrieval (WMS-T). Point your WMS-T aware app at:
https://mesonet.agron.iastate.edu/cgi-bin/wms/nexrad/n0r-t.cgi?
The nexrad-n0r-wmst layer is time aware!

Example WMS-T Application

Here is an example interface utilizing the WMS-T NEXRAD service. The interface code is a part of the OpenLayers toolkit. The default time display shows Hurricane Katrina making landfall.