IBTrACS v04 status updates

[Ken Knapp]

Status updates on IBTrACS v04 will be provided here.

[Ken Knapp]

IBTrACS v04 BUG

DESCRIPTION

A user noticed that the SEARAD variable is incorrectly provided in the CSV  file

FILES AFFECTED

CSV and Shapefile

netCDF file is OK

RESOLUTION PLANNED

This will be resolved in a build soon and incorporated into IBTrACS. This thread will be updated when the fix is included.

[Ken Knapp]

The SEARAD bug has been fixed as of June 23.

[Ken Knapp]

DOCUMENTATION UPDATE

As of June 20th, we have corrected the CSV column description file.

The previous version documented the wrong order of two agencies: Nadi & Wellington. We have corrected the document to record the proper order of the data.

On Wednesday, June 5, 2019 at 12:10:18 PM UTC-4, Ken Knapp wrote:

[Ken Knapp]

IBTrACS v04 BUG REPORT

DESCRIPTION

There is excessive interpolation ongoing for some variables. This can be seen in Hurricane Michael data:

http://www.atms.unca.edu/ibtracs/ibtracs_v04r00/index.php?name=v04r00-2018280N18273#all

R64 data from 10/11 03:00Z through 10/12 21:00 Z should be listed as missing, but instead, the algorithm interpolated during these two days.

This data will affect variables that are are interspersed with missing values (e.g., this does not happen for Max. Sustained Winds, or MSLP but it occurs more often for wind radii).

FILES AFFECTED

All files: netCDF, CSV and shapefiles since initial release of v04 until now.

RESOLUTION PLANNED

In short, codes we thought that would prevent this from happening didn't go as planned. We have now tested the corrected interpolation and have verified that variables are reported in a  manner consistent with the data providers.

We have developed a solution and will implement it soon (by the end of July 2019).

On Wednesday, June 5, 2019 at 12:10:18 PM UTC-4, Ken Knapp wrote:

[Ken Knapp]

BUG FIX

The data filling error has been fixed as of 7/24/2019.

This can be verified by viewing data from Michael at:

http://ibtracs.unca.edu/index.php?name=v04r00-2018280N18273#all

at 2018-10-11 00Z, this is the last observation above 64 knots, and the radius of 64 kt winds end here until winds exceed it once again on the 10-13. 

If data you have has values filled in for R64 between the 11th and 13th, then your data is in error and you should download the latest version.

Please remember that the web browse data at http://ibtracs.unca.edu ius unofficial and is for browsing purposes only. IBTrACS data access is through: https://www.ncdc.noaa.gov/ibtracs/index.php?name=ib-v4-access

Thanks for your patience on this one.

-Ken

[Ken Knapp]

BUG REPORT

DESCRIPTION

Files Affected

All data files

Variables Affected

Only the WMO wind and pressure reports for Nadi (Fiji) are affected. The impact is that the WMO reports are intermittently missing for portions of some storms in the Nadi area of responsiblity.

Some Nadi (Fiji) data was not reported as the WMO reported intensity. The Nadi data still appears in the data but it is the appearance of the intensity in the WMO intensity that is missing.

The effect is clearly seen in Ului and Freda

http://www.atms.unca.edu/ibtracs/ibtracs_v04r00/index.php?name=v04r00-2010070S15168

http://www.atms.unca.edu/ibtracs/ibtracs_v04r00/index.php?name=v04r00-2012362S08165

(the bottom table reports WMO and the Nadi reports are missing from that column, though they appear in the Nadi data columns. ... These tables will only show the bug until the data are corrected.)

CAUSE

The cause appears to be an errant replacement of a variable where the WMO agency is assigned. 

The root is that if there was BoM data and no Wellington data, then Nadi data was removed from WMO. So it affected very few storms. Also, this part of the code has been stable since the initial release. So this bug goes far back to IBTrACS history. Good work to Kay for finding it and reporting it.

RESOLUTION PLANNED

We will test the fix internally for a week or two and roll it out into IBTrACS before the end of April (2020).

[Ken Knapp]

BUG FIX

Nadi is now identified as the WMO agency for several storms identified as erroneous by a user. 

Corrections include storms such as: Ului, Freda, Victor, Pam, etc.

The can be verified by selecting one of the links (below) for Ului or Freda and you can see that Nadi is listed as the WMO agency.

Please remember that the web browse data at http://ibtracs.unca.edu is unofficial and is for browsing purposes only. IBTrACS data access is through: https://www.ncdc.noaa.gov/ibtracs/index.php?name=ib-v4-access

New users should register their use and feedback using the IBTrACS form. Doing so helps us justify continued updates and support.

Thanks for your patience-

-Ken

[Ken Knapp]

BUG FIX on Aug 7, 2020

We have made a correction to the wind radii. Provisional data prior to August 7 had wind radii that was too large. The problem was that the wind radii of two of our real-time datasets had incorrect units. This has been corrected and the units are now provided in nm (nautical miles) consistent with NHC. This change only impacted data labelled "Provisional" (which means it is based on the operational storm warnings and is subject to change after the season ends and storm data are reanalyzed). Please see IBTrACS Technical document for more details on the difference between Provisional and Best Track.

New users should register their use and feedback using the IBTrACS form. Doing so helps us justify continued updates and support.

Thanks for your patience-

-Ken