A database for the outer sizes of tropical cyclones over the Middle Americas
Description
Tropical cyclones (TCs) are catastrophic phenomena and represent a constant threat to populations settled in the tropics. We analyzed 222 and 400 TCs over the North Atlantic (NA) and the Eastern North Pacific (EP) basins, respectively, during the 2000–2024 period. This database provides TC location, TC outer sizes defined by two independent algorithms, and TC spatial metrics every 6 hours. This dataset is provided in a tab-delimited text format with six-hourly information on TC location, maximum winds, central pressure, and size metrics over the NA and EP basins. The dataset format follows the structure of HURDAT2 and previously developed TC outer size datasets. Our compilation includes: (1) non-synoptic best-track times, mainly to indicate landfall and peak intensity; (2) non-developing tropical depressions; and (3) precipitation-based radii when available. TC tracks are obtained from HURDAT2. Sizes are presented by quadrant oriented along the cardinal directions (NE, NW, SW, and SE) in kilometers (km). Shape metrics include asymmetry (A), solidity (S), and dispersion (D). Outer size estimates are derived from two methodologies: ROCLOUD and RPB. Due to limitations in the availability of required input datasets, RPB-derived size metrics are not available for the period 2020–2024. However, all remaining variables and ROCLOUD-based metrics are fully computed for the entire 2000–2024 record. The information is organized as follows: EP132006, LANE, 17, EP132006: TC code LANE: TC name 17: Number of entries of this TC in our dataset Example record: 20120704 0600 12.1 -105.3 55 1006 354.91 552.39 645.30 817.04 592.41 0.57 0.24 0.13 144.14 44.53 141.00 899.99 307.42 20120704 : date (YYMMDD) 0600 : hour (hhmm) 12.1 : TC center latitude (Positive: Northern Hemisphere, Negative: Southern Hemisphere) -105.3 : TC center longitude (Negative: Western Hemisphere, Positive: Eastern Hemisphere) 55 : Maximum 1-minute sustained surface wind (knots) 1006 : Minimum sea level pressure (hPa) 354.91 : Maximum extent (km) of the northeastern quadrant defined by ROCLOUD 552.39 : Maximum extent (km) of the northwestern quadrant defined by ROCLOUD 645.30 : Maximum extent (km) of the southwestern quadrant defined by ROCLOUD 817.04 : Maximum extent (km) of the southeastern quadrant defined by ROCLOUD 592.41 : Mean extent (km) of the four quadrants defined by ROCLOUD 0.57 : Asymmetry value (range 0–1) 0.24 : Dispersion value (range 0–1) 0.13 : Solidity value (range 0–1) 144.14 : Maximum extent (km) of the northeastern quadrant defined by RPB 44.53 : Maximum extent (km) of the northwestern quadrant defined by RPB 141.00 : Maximum extent (km) of the southwestern quadrant defined by RPB 899.99 : Maximum extent (km) of the southeastern quadrant defined by RPB 307.42 : Mean extent (km) of the four quadrants defined by RPB Missing data are denoted by −9999, reflecting periods when satellite imagery or required input datasets were not available.
Files
Steps to reproduce
We developed the Radius of the Outermost CLOUD (ROCLOUD) algorithm. This technique is based on Phyton programming language and determines the outer TC size by combining wind field profile data and convective cluster locations. The ROCLOUD technique measures the distance between the TC center reported by HURDAT2 and the farthest convective cluster located inside the outer TC size obtained by PA21. The algorithm automatically carries out the following steps for each TC position: 1. IR images are segmented based on BTs < 233 K. The technique segments pixels or groups of pixels with similar characteristics inside the outer TC size obtained by PA21. 2. Cumulus polygons, which represent areas with intense TC precipitation, are determined when the IR images are greater than 50 pixels (~ 800 km2). 3. The cumulus polygons that intersect the outer TC size are selected to define the maximum extension from the TC center per quadrant. The quadrants are defined according to the TC displacement. 4. TCs that interact with the Intertropical Convergence Zone (ITCZ) are analyzed following these two steps: If the TC position is located in the latitudinal band 0°N-10°N, the cumulus clouds exceeding the outer TC size will not be considered for obtaining the maximum extent per quadrant. Therefore, we assume these cumulus clouds belong to the ITCZ. 5. If the TC position is located northwards 10°N and the outer TC size interacts with the ITCZ cloud band, the cumulus clouds that exceed the outer TC size limits will be discarded for calculating the maximum extent per quadrant. 6. The distance of the edged cumulus polygons from the TC center is determined for the four TC quadrants, which are expressed in kilometers (km) and degrees (°). 7. The mean radius is calculated as an average of the four quadrant radii. If a quadrant radius has no value, it is not considered for the average. The technique of Radius of Precipitation Bands (RPB) was developed as a modification of ROCLOUD. The RPB algorithm follows steps 1-5 of ROCLOUD, but it considers the IMERG rainfall data to define the maximum extension per quadrant of 2.5 mm/h. For more details, see the full paper on the database.
Institutions
- Universidad Nacional Autonoma de MexicoDistrito Federal, Coyoacan