create_full_track: Impute hurricane tracks to finer time scale

Description Usage Arguments Details Value Note Examples

View source: R/impute_track.R

Description

Inputs data on a hurricane's track and imputes to a finer time resolution. For example, if the hurricane tracks are recorded at 6-hour intervals, this could be used to impute locations and windspeeds at 15-minute intervals. This function also does some reformatting necessary for later functions in the stormwindmodel package.

Usage

1
2
create_full_track(hurr_track = stormwindmodel::floyd_tracks,
  tint = 0.25)

Arguments

hurr_track

Dataframe with hurricane track data for a single storm. The dataframe must include columns for date-time (year, month, day, hour, minute; e.g., "198808051800" for August 5, 1988, 18:00 UTC), latitude, longitude, and wind speed (in knots). The column names for each of these must be date, latitude, longitude, and wind. See the example floyd_tracks dataset for an example of the required format.

tint

Interval (in hours) to which to interpolate the tracks. The default is 0.25 (i.e., 15 minutes)

Details

The function uses natural cubic splines for interpolation, both for location and for wind speed. Degrees of freedom are based on the number of available observations for the storm.

Value

A version of the storm's track data with latitude, longitude, and wind speed interpolated between observed values. Also, wind speed is converted in this function to m / s and the absolute value of the latitude is taken (necessary for further wind speed calculations). Finally, the names of some columns are changed (tclat for latitude, tclon for longitude, and vmax for wind speed.)

Note

This function imputes between each original data point, and it starts by determing the difference in time between each pair of data points. Because of this, the function can handle data that includes a point that is not at one of the four daily synoptic times (00:00, 06:00, 12:00, and 18:00). Typically, the only time hurricane observations are given outside of synoptic times for best tracks data is at landfall.

After imputing the tracks, longitude is expressed as a positive number. This is so the output will work correctly in later functions to fit the wind model. However, be aware that you should use the negative value of longitude for mapping tracks from the output from this function.

Examples

1
2
3
4
5
data("floyd_tracks")
full_track <- create_full_track(hurr_track = floyd_tracks)

# Interpolate to every half hour (instead of default 15 minutes)
full_track <- create_full_track(hurr_track = floyd_tracks, tint = 0.5)

geanders/stormwindmodel documentation built on Oct. 15, 2018, 12:06 p.m.