Data
The Huracán data repository is on JASMIN (Centre for Environmental Data Analysis):
/gws/nopw/j04/huracan/data/ (apply for access at accounts.jasmin.ac.uk)
Tropical cyclone tracks
Introduction: Below are the tropical cyclone track datasets available for analysis. Tracks are based on the TRACK algorithm (see Hodges et al., 2017), with additional tracking with TempestExtremes. Data located on JASMIN are in the Huracán Group Workspace (above), or at the path given (monospace type). Additional datasets are available on other platforms (RACC, ECMWF, DKRZ), and these will be migrated to JASMIN as soon as possible. New datasets will be added as Huracán progresses. If you encounter any errors or broken paths, please inform Kevin Hodges, Pier Luigi Vidale or Alex Baker.
Data type: tropical cyclone tracks
Domain: Northern Hemisphere / global
Description: date, time, longitude and latitude over full cyclone lifecycles, from precursor to post-tropical cyclone (and, for some datasets, separate files containing pre-TC tracks, which are generated by TRACK prior to tropical cyclone identification, are also provided)
Variables: cyclone intensity metrics (vorticity, minimum sea-level pressure, and maximum wind speed at 925hPa and 10m), cyclone phase-space parameters, and cyclone-associated precipitation (or a subset of these)
Format: NetCDF or ASCII
Software: The use of huracanpy is recommended for reading track data
1. Modern Era Reanalyses
I. ERA5, 1950-2022 (1979-2019 currently on JASMIN: /gws/nopw/j04/hrcm/cache/bakera/PRIMAVERA/TRACK/ERA5/NH/)
II. ERA-Interim, 1979-2018 (Univ. Reading RACC)
III. JRA25, 1979-2013 (Univ. Reading RACC)
IV. JRA55, 1958-2017 (Univ. Reading RACC)
V. MERRA2, 1980-2016 (Univ. Reading RACC)
(MERRA2 and JRA55 will be brought up to date as soon as possible.)
VI. JRA3Q, 1948-2023, 40km
VII. ECMWF operational analyses (2008-2023), variable resolution, currently 9km
2. Longer reanalyses (surface-only data assimilation)
I. ERA20C, 1900-2010 (Univ. Reading RACC)
II. CERA20C, 1901-2010, 10 member ensemble (Univ. Reading RACC)
III. NOAA-20CR V3, 1806-2015, 80 member ensemble (JASMIN: /gws/nopw/j04/cmip6_track/NOAA-20CRv3/TC/)
3. Climate model simulations
I. CMIP6, AMIP (1979-2014), historical (1850-2014), piControl , ssp126, ssp245, ssp370, ssp434, ssp460, ssp585 (2015-2100), multiple models, ensembles (JASMIN: /gws/nopw/j04/cmip6_track/CMIP6-TRACK/)
II. PRIMAVERA/HighResMIP, multiple models, ensembles, SST forced (1950-2014, 2015-2050), coupled (1950-2050) (JASMIN: /badc/highresmip-derived/data/storm_tracks/TRACK/)
III. SPHINX, EC-Earth with and without stochastic physics, atmosphere only, multiple resolutions (T159-T1279), ensembles (1-10), historical (1979-2008), future (2038-2068) (JASMIN: /gws/nopw/j04/hrcm/kihodges/SPHINX/)
IV. d4PDF, MRI, 100 member ensemble, historical (1951-2010), future 4xCO2 (2071-2100) (JASMIN: /gws/nopw/j04/cmip6_track/d4pdf/historical/TC/ and /gws/nopw/j04/cmip6_track/d4pdf/future/TC/)
V. CMIP6-HighResMIP tracks (historical 1950-2014, AMIP and coupled, 6 models; future 2015-2050, AMIP and coupled, 5 models) (available from CEDA)
VI. nextGEMS coupled storm-resolving simulations (ICON 10 and 5km; IFS 9, 4 and 2.5 km); cycle 2 and cycle 3 runs
VII. N1280 (10 km) UM simulations with the (AMIP; 1980–2010; configuration similar to HighResMIP-PRIMAVERA):
— u-ck777 /home/users/plvidale/TRACK-1.5.2/u-ck777/
— u-ck778 /home/users/plvidale/TRACK-1.5.2/u-ck778/
— u-ch330 /home/users/plvidale/TRACK-1.5.2/u-ch330/
VIII. ERA20CM, 10 member, 125km, prescribed SSTs and Sea Ice, 1900-2010
Forthcoming:
— DestinE and EERIE high-resolution climate runs
— nextGEMS prototype km-scale coupled model runs
4. Seasonal prediction ensembles
I. GLOSEA5/6, 7month hindcasts, 7 members, 4xmonth, 1993-2015 (JASMIN: /gws/nopw/j04/klingaman/fascinate/TRACK/GLOSEA5/
and /gws/nopw/j04/klingaman/fascinate/TRACK/GLOSEA6/) and GLOSEA6 forecasts, 7 month, 2 ensemble members (on jasmin at: /gws/nopw/j04/westpac/kihodges/GLOSEA6/)
II. ECMWF fully coupled hindcasts (CSF-20C, “guh4”) 1901-2010, 50 ensemble members initialised from CERA-20C, with May and August start dates, 4 months and 12 hourly output. Atmosphere-only simulations using prescribed observed SSTs (SEAS5-20C, “guxf”), 24-month hindcasts (multi-annual forecasts) 1901-2010, 10 ensemble members, November and May start dates. (ECMWF)
III. C3S multi-model hindcast, ensembles (will be re-tracked to include high latitudes)
5. NWP (currently only obs-matched tracks)
I. TIGGE:
— ECMWF:- 2008-2024, 00, 12hr, EPS (51 members, 0-15 days), deterministic (0-10 days)
— UKMO:- 2008-2024, 00, 06, 12, 18hr, EPS (17 members, 0-7.5 days)
II. ECMWF 48-day hindcasts, twice per week, 1995-2022 (ECMWF)
III. ECMWF 20 year high resolution hindcasts 2023/06-2024 == 2003-2023 (2x week), 11 member, 0-15 days, operational resolution (9km)
IV. ECMWF extended forecasts 2023/07-2024, daily, 100 member, 0-46 days, ~ERA5 resolution
V. ECMWF AIFS, deterministic, 28km, 2024/03-now
6. Decadal prediction systems
I. Met Office GLOSEA5 and GLOSEA6 hindcasts, 1993 – 2016, 6 hourly, coupled, 50km (1st, 9th, 17th, 25th of each month initializations, 7 members, 6 month integrations)
II. Met Office GLOSEA6 forecasts (initialized each day, 2 members (6 month), 2 members (2month), 2021-2024)
III. DePreSys4 (JASMIN: /gws/nopw/j04/cmip6_track/DePreSys4/TC/)
IV. Extended hindcasts, based on ECMWF SEAS5, coupled, 12 hourly:
— guxf, 1901-2010, 10 member, 24month simulations initialised 1st May and November from CERA-20C, 50km
— guh4, 1901-2010, 51 member, 4 month simulations, initialised 1st May and August from CERA-20C, 80 km
— gh6j, 1901-2010, 51 member, 4 month simulations, initialised 1st May
— gir6, 1901-2010, 51 member, 4 month simulations, initialised 1st August (currently running)
— b1gl, 1901-2010, 51 member, 4 month simulations, initialised 1st May, August, November
superBT (a super Tropical Cyclone Best Track dataset)
https://github.com/tenkiman/superBT
Developed by: Michael Fiorino, Ph.D.
Commander, US Navy (retired)
Department of Atmospheric, Oceanic and Earth Sciences, George Mason University
Version V03 has R34 data from the JTWC/NHC best tracks only. V10 will have more R34 data from CIRA and analysis of full-res ERA5 surface winds. V10 intended to be ready before the end of 2023, but the work is unfunded. Note that both JTWC & NHC are now ‘best tracking’ the wind radii at least since 2018. The current version V03 is distinctive because of the pTC (9X ‘invests’) data and rainfall from three satellite-based analyses. For more information, see this presentation, which is the talk Michael Fiorino gave at the Huracán kick-off day in March. It shows some of the features of the pTC data, such as how, in developing pTC, the wind shear drops 48 h before genesis.
Python2.7 code developed to list (post-)TCs and do analysis is available here and documented here.