Operational and Parallel Track Files

These track files are for operational models, and are all either made from operational runs by NCEP Central Operations (NCO) or retrospective reruns by EMC of the NCO operational systems. Unless the link says otherwise, the tracks are the original, 126hr forecast tracks from the model.

In some cases, interpolated tracks are available as well, made using the NHC interpolator. Generally, you want the original tracks since they give the exact output of the model, and hence are better for analyzing the model's forecast characteristics. However, if you are comparing to NHC/JTWC forecasts or on-time models (such as statistical models), then you must do so with the interpolated tracks to take into account the fact that HWRF delivers its tracks almost six hours after those models. The interpolator will also sometimes correct track or intensity biases in the model forecast, improving the forecast skill.

H215: (original) (interpolated): The 2015 HWRF retrospectives, run on the NOAA Zeus, Jet, Tide or Gyre clusters using the 2015 GFS retrospectives as initial and boundary conditions.
H15Z: (original) (interpolated): The 2014 HWRF run retrospectively with the 2015 upgraded GFS on the NOAA WCOSS Tide or Gyre clusters. This configuration was also called H14Z in late 2014.
G215: The 2015 GFS run retrospectively for the 2011-2014 hurricane seasons (generally, late May to early December). This is the parent model of the H215 and H15Z, and is also the operational GFS as of 18 UTC, January 14, 2015. This includes basins and storms during the simulated time period, with up to a 10.5 day forecast. The 2011 simulation was done on the NOAA Zeus cluster while the 2012-2014 simulations used the NOAA WCOSS Tide and Gyre clusters. The tracks use non-standard two-letter basin specifiers for the SH and IO basins to differentiate between the two halves of each basin. Instead of Indian Ocean (IO), they use the Bay of Bengal (BB) and North Arabian Sea (NA). Instead of Southern Hemisphere (SH), they use the South Pacific (SP) and South Indian Ocean (SI).
H214: (original) (interpolated): The 2014 HWRF and retrospectives, run on the NOAA Zeus, Jet, Tide or Gyre clusters using the 2012 GFS. This has been updated on April 9, 2015 to include all real-time runs of the 2014 version of HWRF by NCO. There will be no more updates of this archive - any runs of the 2014 HWRF in April or May 2015 will be added to H15Z, above.
Real-Time HWRF on Jet: An alternative configuration of the HWRF system run in real-time for non-NWS basins (SL, SH, IO, WP) in 2012-2015 and delivered to the Joint Typhoon Warning Center (JTWC). Nearly all cycles were run on the NOAA Jet cluster, with a few very rare exceptions run on WCOSS. This contains all forecasts up to April 8, 2015. Note that these are real-time runs in a development environment, so they are inherently unreliable: some cycles may be missing forecasts or may have been cold starts and hence lower quality. Real-time tracks for this ongoing test can be obtained on the AOML storm or NCEP EMC HTTP servers or the NCEP EMC FTP server.
2012 GFS (pre13h): The 2012 operational GFS system, which is the parent model of H130-H134, H213 and H214 (but NOT H215 or H15Z). It was also the parent model of H212 from mid June onward. This is either the operational NCO GFS or retrospective reruns. Reruns were done on CCS when available or Zeus otherwise.
H213: The 2013 final operational configuration, run in real-time on WCOSS or Jet, or run in retospective mode on various machines. Reruns were done on WCOSS when available, and otherwise on Jet or Zeus.
H212: The 2012 operational HWRF system, or retrospective runs of the system for prior years. Reruns were done on CCS when available, or Jet otherwise.

2014 HWRF Major Tests

This section contains 2014 tests, as well as the collected real-time and retrospective runs of the 2013 operational HWRF.

2014 Baseline T14C: A 2014 experimental configuration that added data assimilation improvements. This is the 2014 Baseline: it is the configuration against which other planned upgrades are compared.

2013 HWRF Major Tests

This section contains links to track (ATCF) files for some major HWRF experimental configurations. These are the non-interpolated track files (the raw model output track).

2013 H130: The 2013 baseline: a three year (2010-2012) test of North Atlantic and North East Pacific storms using the baseline configuration. This was the control run for all atmospheric and ocean model changes. It is essentially the 2012 operational HWRF with a variety of data assimilation, vortex initialization, technical and product upgrades.
2013 H131: The 2013 planetary boundary layer (PBL) upgrade test. This test reduced the per-domain "tuning" applied to the existing PBL implementation, replacing it with a variable critical Richardson number. The same storms and years were run as for the H130 test
2013 H132: The 2013 Mesoscale Simplified Arakawa Schubert (MesoSAS) upgrade test. This test enabled convection in all three atmospheric domains, replacing the original Simplified Arakawa Schubert (SAS) scheme with the MesoSAS scheme and added some changes to the PBL tuning (different changes than H131). This test was originally going to include all storms and years as H130, but was aborted after about eight hundred forecast cycles due to poor performance. It over-intensified sheared storms and overly axisymmetrized all storms.