sdafilter {argosfilter}

Filter Argos locations

Description

This function filters location data obtained from Argos, using the Freitas et al. (2008) algorithm.

Usage

sdafilter(lat, lon, dtime, lc, vmax = 2, ang = c(15, 25), distlim = c(2500, 5000))

Arguments

lat

a numeric vector of latitudes, in decimal degrees

lon

a numeric vector of longitudes, in decimal degrees

dtime

a vector of class POSIXct with date and time for each location

lc

a numeric or character vector of Argos location classes. Argos locations Z can be entered as "Z", "z" or -9

vmax

speed threshold, in m/s. Default is 2 m/s

ang

angles of the spikes to be removed. Default is c(15,25). No spikes are removed if ang=-1

distlim

lengths of the above spikes, in meters. Default is c(2500,5000)

Details

Locations are filtered using the algorithm described in Freitas et al. (2008). The algorithm first removes all locations with location class Z (-9), which are the points for which the location process failed. Then all locations requiring unrealistic swimming speeds are removed, using the MacConnell et al. (1992) algorithm, unless the point is located at less than 5 km from the previous location. This procedure enables retaining good quality locations for which high swimming speeds result from location being taken very close to each other in time. The default maximum speed threshold is 2 m/s. The last step is optional, and enables to remove unlikely spikes from the animal's path. The angles of the spikes should be specified in ang, and their respective length in distlim. The default is c(15,25) for ang and c(2500,5000) for distlim, meaning that all spikes with angles smaller than 15 and 25 degrees will be removed if their extension is higher than 2500 m and 5000 m respectively. No spikes are removed if ang=-1. ang and distlim vectors must have the same length.

Value

Returns a vector with the following elements: "removed" (location removed by the filter), "not" (location not removed) and "end_location" (location at the end of the track where the algorithm could not be applied).

Author(s)

Carla Freitas, with contributions from Anne Goarant and Catriona MacLeod

References

Freitas, C., Lydersen, C., Ims, R.A., Fedak, M.A. and Kovacs, K.M. (2008) A simple new algorithm to filter marine mammal Argos locations Marine Mammal Science 24:315-325.

McConnell, B.J., Chambers, C. and Fedak, M.A. (1992) Foraging ecology of southern elephant seals in relation to the bathymetry and productivity of the Southern Ocean. Antarctic Science 4:393-398.

See Also

vmask

Examples

data(seal)
lat<-seal$lat
lon<-seal$lon
dtime<-seal$dtime
lc<-seal$lc

# plot unfiltered data
plot(lon,lat,col="lightgrey",type="l",xlim=c(5,18),
ylim=c(77.1,79.1),xlab="Longitude",ylab="Latitude")

# filter by speed only
mfilter<-vmask(lat,lon,dtime,2)
mfilter[1:10]
lines(lon[which(mfilter=="not")],lat[which(mfilter=="not")],col="red")

# filter data using sdafilter
cfilter<-sdafilter(lat, lon, dtime, lc)
cfilter[1:20]
lines(lon[which(cfilter=="not")],lat[which(cfilter=="not")],col="blue")

# check number of locations (by location class) removed by each filter
table(lc,mfilter)
table(lc,cfilter)