Frequently Asked Questions
- Data Access
- Data Values
- General
- IBTrACS Construction
- Why does the best track for the northern hemisphere use data from the JTWC? Ideally, should it not use data from the JMA and the IMD?
- Why did you choose a 10-min wind?
- Why does IBTrACS differ from HURDAT (or data from other agencies)?
- What are the basin boundaries?
- How are Southern Hemisphere seasons handled?
- I have 50 years of historical best track data in paper form. How can I get this included in IBTrACS?
General
Is IBTrACS a reanalysis?
No. IBTrACS is a collection of best track data from numerous agencies. The data have been objectively combined to facilitate analysis. It does not represent any reanalysis of satellite or surface data.
Why should I use IBTrACS?
The IBTrACS project:
- Contains the most complete global set of historical tropical cyclones available
- Combines information from numerous tropical cyclone datasets
- Simplifies inter-agency comparisons by providing storm data from multiple sources in one place
- Provides data in popular formats to facilitate analysis
- Checks the quality of storm inventories, positions, pressures, and wind speeds, passing the information on to the user
How does IBTrACS relate to HURDAT
In short, HURDAT is a reanalysis of storms in one basin while IBTrACS is a collection of global best track dataset.
As such, IBTrACS incorporates
the existing HURDAT files for the North Atlantic basin. Any and all updates
that are performed on the HURDAT data are thus automatically included and
updated in IBTrACS.
How does IBTrACS relate to HURSAT?
The Hurricane Satellite dataset ( HURSAT) is made up of satellite brightness temperatures over global tropical cyclones. HURSAT uses IBTrACS to determine storm position prior to remapping and subsetting the satellite data.
How should I cite IBTrACS?
See the bibliography.
I think I found an error in IBTrACS. What should I do?
If you have found an error in one or more of our storm track files, thank you! The accuracy of the dataset can only be improved through vigilant data users.
First, check to see if your concern is already noted in this FAQ. If not, please send an email detailing the error to:
IBTrACS.Team [at] noaa.gov
IBTrACS Construction
Why does the best track for the northern hemisphere use data
from the Joint Typhoon Warning Center? Ideally, should it not use data from the
Japan Meteorological Agency and the India Meteorological Department?
The goal of IBTrACS is to inventory each TC in all basins. To that extent, we use data from all available resource. While JMA is the RSMC for the Western Pacific, many other agnecies monitor the region (Hong Kong, China, etc.). Including everyone:
- minimizes the dependence upon the operating procedures at any one agency,
- allows more storms to be included,
- helps identify the variability between agencies, and
- identifies possible errors when differences do occur.
Variability of storm reports between agencies are maintained in IBTrACS by reporting minimum and maximum values for wind and pressure at each time. Furthermore, IBTrACS could include data from other agencies as the data become available.
Why did you choose a 10-min wind?
v02r01 and earlier
Maximum sustained winds are reported using various averaging periods which depends on the operating procedures of a forecast center. For example, while the HURDAT dataset uses a 1-minute wind average, a 2-minute average is employed by the Shanghai Typhoon Institute, and a 10-minute average is used by the Australian Bureau of Meteorology and many others. In order to create a dataset that was globally homogeneous, it was necessary to standardize the maximum winds to a single averaging value. Hence, the 10-minute averaging period was selected because it is the global standard as set by the World Meteorological Organization.
v03r01 and later
The latest IBTrACS provides wind speed data unchanged from the agencies. No wind speed conversions are made, nor are the wind speed combined. Any differences between an agencies intensity and IBTrACS should be brought to the attention of the IBTrACS team
Why does IBTrACS differ from HURDAT (or data from other agencies)?
v02r01 and earlier
First, all winds are reported as 10-minute winds. Data from NOAA (HURDAT, CPHC), JTWC and the RSMC New Delhi (IMD) all provide data as 1-minute. Winds were converted to 10-min using:
V10 = 0.88 * V1
Second, IBTrACS is a compilation of data from all available agencies. So the number storms in IBTrACS will likely be larger than the number of storms from an individual agency.
Third, since IBTrACS is a compilation of data from many agencies, the IBTrACS maximum sustained winds (MSW) is an average from all available agencies and will likely be different from the reports of each agency. Calculating statistics from IBTrACS (like ACE, number of super typhoons, etc.) will result in values which are more than likely different from those of any particular agency.
v03r01 and later
IBTrACS does not make any modification to HURDAT data. Any differences should be brought to the attention of the IBTrACS team.
What are the basin boundaries?
Basins within IBTrACS are provided for convenience. The latitude and longitudinal extents are provided in the following table. NOTE: The boundary between EP and NA was variable and determined using a polynomial which roughly follows Central America.
| Latitude (deg North) |
Longitude (deg East) |
|||
| Basin | Minimum | Maximum | Minimum | Maximum |
| North Indian | 0 | 90 | 30 | 100 |
| West Pacific | 0 | 90 | 100 | 180 |
| East Pacific | 0 | 90 | 180 | Variable |
| North Atlantic | 0 | 90 | Variable | 30 |
| South Indian | -90 | 0 | 10 | 135 |
| South Pacific | -90 | 0 | 135 | -70 |
| South Atlantic | -90 | 0 | -70 | 10 |
How are Southern Hemisphere seasons handled?
Each storm in the IBTrACS dataset was assigned a season, representative of its location. Storms in the Northern Hemisphere follow the calendar year and storms in the Southern Hemisphere follow a 6-month offset year, from July to June. For example, the Southern Hemisphere 2008 season is spanned by 1 July 2007 to 30 June 2008.
I have 50 years of historical best track data in paper form.
How can I get this included in IBTrACS?
If you have historical best track data that you would like to be considered for inclusion into the IBTrACS dataset, please contact:
IBTrACS.Team [at] noaa.gov
We may be able have the data digitized through a program at NCDC.
If you already have data in digital form, please email it as an attachment to the above address, or include a web site where the data may be downloaded.
Data access
What is the easiest way to ingest the entire dataset?
The best choice is to use the netCDF version which contains the entire data set in one netCDF file. Other formats also have an "AllStorms" file which contains all storms in one file as well. But not all other formats can store all the data generated in merging the IBTrACS dataset.
What is the best format for input into a spreadsheet or database?
The CSV and WMO-CSV format has been generated to facilitate the use of IBTrACS in spreadsheets and databases.
Data Values
What do the values mean in the netDF files?
Values in netCDF files are converted to bytes to save space. The key is provided in the netCDF variable attributes. The v03r04 PDF Key is provided.
How was storm nature assigned (e.g., tropical vs. subtropical vs. extratropical)
The purpose was to delineate vastly different storm types. Some agencies only report wind spped with no information about the nature of the storm. Other agencies only report natures that are directly derived from wind speeds (e.g., gale force, storm force). And some agencies report storm nature (tropical, extratropical, etc.). IBTrACS converts all assignments to a similar set of definitions. This primary set ignores categories that are wind speed dependent. Instead we focus on the storm characteristics. A PDF table shows how the IBTrACS storm nature was reported based on agency reports.
Note: This report of nature only affects the netCDF variable nature_for_mapping and nature assignments for the WMO dataset.
Can I access best track data from the source agency before it is averaged by IBTrACS?
v02r01 and earlier
Yes. The data as reported by each agency are provided in three of the IBTrACS formats: ibtracs, ibtracs_net and ibtracs_csv.
v03r01 and later
Question is not applicable. IBTrACS provides intensities as provided by the agency and does not support a mean intensity.
Why all the missing standard deviation values for the
North Atlantic?
The standard deviation for inter-agency intensity variation is determined from all agencies reporting intensity (winds or pressure) at that time. In the case of the North Atlantic, the only source of wind or pressure data is the HURDAT dataset. The standard deviation does not apply since only one report is available, thus it is reported as missing
This also ocurrs in other basins when the number of agencies reporting the intensity is only one.



indicates a link which will lead you out of the Federal Government webspace