Nothing
#' Calculate Wind Fetch
#'
#' Wind fetch is the unobstructed length of water over which wind can blow, and
#' it is commonly used as a measure of exposure to wind and waves at coastal
#' sites. The \code{fetch} function automatically calculates the wind fetch for
#' marine locations within the boundaries of the specified coastline layer.
#' This allows wind fetch to be calculated anywhere around the globe.
#'
#' The function takes a \code{\link[sp]{SpatialPolygons-class}} object
#' (\code{polygon_layer}) that represents the coastline, surrounding islands,
#' and any other obstructions, and calculates the wind fetch for every specified
#' direction. This is calculated for all the user-defined sites, that are
#' represented as the point geometries in a
#' \code{\link[sp]{SpatialPoints-class}} object.
#'
#' The directions for which the wind fetch are calculated for each site are
#' determined by the number of directions per quadrant (\code{n_directions}).
#' The default value of 9 calculates 9 fetch vectors per quadrant (90 degrees),
#' or equivalently, one fetch vector every 10 degrees. The first fetch vector is
#' always calculated for the northerly direction (0/360 degrees).
#'
#' @param polygon_layer \code{\link[sp]{SpatialPolygons}}* object where the
#' polygon geometries represent any obstructions to fetch
#' calculations including the coastline, islands and/or
#' exposed reefs.
#' @param site_layer \code{\link[sp]{SpatialPoints}}* object where the point
#' geometries represent the site locations.
#' @param max_dist numeric. Maximum distance in kilometres (default 300). This
#' will need to be scaled manually if the units for the CRS are
#' not 'm'.
#' @param n_directions numeric. The number of fetch vectors to calculate per
#' quadrant (default 9).
#' @param site_names character vector of the site names. If missing, the site
#' names are taken from a column of the data associated with
#' \code{site_layer} matching the regular expression
#' \code{^[Nn]ames{0,1}}. If there is no such column, then
#' default names are created ('Site 1', 'Site 2', ...).
#' @param quiet logical. Suppress diagnostic messages? (Default \code{FALSE}).
#'
#' @return Returns a \code{\link{Fetch}} object.
#'
#' @note At least one of the inputs to the \code{polygon_layer} or
#' \code{site_layer} arguments must be projected. If one of the inputs are
#' not projected, then it will be transformed to have the same projection
#' as the other. If both are projected, but do not have identical
#' coordinate reference systems (CRS) then \code{site_layer} will be
#' transformed to the same CRS as \code{polygon_layer}.
#'
#' @seealso \code{\link[rgdal]{spTransform}} for methods on transforming map
#' projections and datum.
#' @seealso \code{\link[sp]{is.projected}} for checking whether a spatial object
#' is projected.
#' @seealso \code{\link{fetchR}} for an overview of this package with an
#' extensive, reproducible example.
#' @seealso \code{\link{summary,Fetch-method}} for summarising the fetch lengths.
#'
#' @importFrom rgdal CRSargs
#' @importFrom sp SpatialPoints CRS over spTransform coordinates SpatialLinesLengths SpatialLinesDataFrame identicalCRS
#' @importFrom rgeos gBuffer gIntersects gIntersection
#' @importFrom utils head
#' @importFrom methods new is
#' @import sp
#' @examples
#'
#' # Create the polygon layer ----------------------------------------
#' #
#' # This is the layer that represents any obstacles that obstruct wind flow.
#'
#' # Import map data for the Philippines.
#' philippines.df = ggplot2::map_data("world", region = "Philippines")
#'
#' # Create a list for each separate polygon
#' philippines.list = split(philippines.df[, c("long", "lat")],
#' philippines.df$group)
#'
#' library(sp)
#'
#' philippines.Poly = lapply(philippines.list, Polygon)
#' philippines.Polys = list(Polygons(philippines.Poly, ID = "Philippines"))
#'
#' # Include CRS information to make it a SpatialPolygons object
#' philippines.sp = SpatialPolygons(philippines.Polys,
#' proj4string = CRS("+init=epsg:4326"))
#'
#' # Create the points layer ----------------------------------------
#' #
#' # The points layer represents the locations for which the wind fetch needs to
#' # be calculated.
#'
#' # We need to calculate wind fetch for the following 3 sites:
#' sites.df = data.frame(lon = c(124.4824, 125.8473, 124.8416),
#' lat = c(9.167999, 9.751394, 11.478243),
#' site = c("Camiguin Island", "Bucas Grande Island",
#' "Talalora"))
#'
#' # Create the SpatialPoints object
#' sites.sp = SpatialPoints(sites.df[, 1:2], CRS("+init=epsg:4326"))
#'
#' # Map projection -------------------------------------------------
#' #
#' # At least one of the polygon or points layers need to be projected to
#' # calculate wind fetch.
#'
#' # All these locations lie within the Philippines zone 5 / PRS92, that has
#' # WGS84 Bounds: 123.8000, 5.3000, 126.7000, 12.7500
#' # (http://spatialreference.org/ref/epsg/3125/)
#' # This suggests that this is a suitable map projection.
#' philippines.proj = spTransform(philippines.sp, "+init=epsg:3125")
#'
#' # Calculate wind fetch -------------------------------------------
#' #
#' # Calculate wind fetch at all the 3 locations for every 10 degrees on the
#' # compass rose, with a maximum distance for any fetch vector of 300 km.
#' my_fetch = fetch(philippines.proj, sites.sp, site_names = sites.df$site)
#' my_fetch
#'
#' # Return only the summary data frame
#' summary(my_fetch)
#'
#' # Transform the fetch vectors back to the original CRS
#' my_fetch_latlon = spTransform(my_fetch, proj4string(philippines.sp))
#'
#' # Return the raw data in the original, lat/lon coordinates
#' my_fetch_latlon.df = as(my_fetch_latlon, "data.frame")
#' my_fetch_latlon.df
#'
#' # Plot the wind fetch vectors ------------------------------------
#'
#' # Plot the fetch vectors in the projected space...
#' plot(my_fetch, philippines.proj, axes = TRUE)
#'
#' # ... or in the original coordinate reference system
#' plot(my_fetch, philippines.sp, axes = TRUE)
#'
#' # Output to KML --------------------------------------------------
#' \dontrun{
#'
#' # Save a KML file in the current working directory.
#' kml(my_fetch, colour = "white")
#' }
#' @export
fetch = function(polygon_layer, site_layer, max_dist = 300, n_directions = 9,
site_names, quiet = FALSE){
if (!is(polygon_layer, "SpatialPolygons"))
stop(paste("polygon_layer must be a SpatialPolygons object.\nSee",
"'?SpatialPolygons' for details on how to create a",
"SpatialPolygons object."), call. = FALSE)
if (!is(site_layer, "SpatialPoints"))
stop(paste("site_layer must be a SpatialPoints object.\nSee",
"'?SpatialPoints' for details on how to create a SpatialPoints",
"object."), call. = FALSE)
if (!is.numeric(max_dist) || length(max_dist) != 1)
stop("max_dist must be a single number.", call. = FALSE)
if (!is.numeric(n_directions) || length(n_directions) != 1)
stop("n_directions must be a single integer.", call. = FALSE)
n_directions = round(n_directions)
if (n_directions < 1 || n_directions > 90)
stop("n_directions must be between 1 and 90.", call. = FALSE)
if (!missing(site_names)){
site_names = as.character(site_names)
if (length(site_names) != length(site_layer)){
warning(paste("lengths differ for the number of sites and site names;",
"using default names instead."), call. = FALSE)
site_names = paste("Site", seq_along(site_layer))
}
} else {
if (any(grepl("^[Nn]ames{0,1}$", names(site_layer)))){
name_col = grep("^[Nn]ames{0,1}$", names(site_layer))
site_names = as.character(data.frame(site_layer)[, name_col[[1]]])
} else {
site_names = paste("Site", seq_along(site_layer))
}
}
quiet = as.logical(quiet[1])
## Check if the polygon and points layers are projected, and ensure they have
## the same CRS.
which_proj = c(is.projected(polygon_layer), is.projected(site_layer))
if (all(!which_proj))
stop("polygon_layer and/or site_layer must be projected to calculate fetch",
call. = FALSE)
if (all(which_proj) && !identicalCRS(polygon_layer, site_layer)){
warning("the CRS for polygon_layer and site_layer differ; transforming
site_layer CRS to match")
site_layer = spTransform(site_layer, CRS(proj4string(polygon_layer)))
}
if (!which_proj[1]){
if (!quiet)
message("projecting polygon_layer onto the site_layer CRS")
polygon_layer = spTransform(polygon_layer, CRS(proj4string(site_layer)))
}
if (!which_proj[2]){
if (!quiet)
message("projecting site_layer onto the polygon_layer CRS")
site_layer = spTransform(site_layer, CRS(proj4string(polygon_layer)))
}
if (!quiet)
message("checking site locations are not on land")
# Should remove these sites with warning instead of returning error.
if (any(!is.na(over(polygon_layer, site_layer))))
stop("at least one site location is on land")
# Convert max_dist to appropriate units.
# First of all convert max_dist to metres (default)
max_dist = max_dist * 1000
# Double check if metres are the correct units
proj_unit = strsplit(gsub("*.+units=", "",
CRSargs(CRS(proj4string(polygon_layer)))), " ")[[1]][1]
# If not, warn the user that the supplied max_dist should be scaled
# appropriately
if (proj_unit != "m")
warning("the PROJ.4 unit is not metres; ensure max_dist has been scaled
appropriately")
directions = head(seq(0, 360, by = 360 / (n_directions * 4)), -1)
# Return the quadrant the directions belong to
dirs = as.numeric(directions)
dirs_bin = findInterval(dirs, seq(45, 315, by = 90))
quadrant = rep("North", length(dirs))
quadrant[dirs_bin == 1] = "East"
quadrant[dirs_bin == 2] = "South"
quadrant[dirs_bin == 3] = "West"
# Rearrange sequence order to start at 90 degrees to match up with the output
# from gBuffer
directions = unlist(split(directions, directions < 90), use.names = FALSE)
# Create an empty list to store the Fetch objects
fetch_list = vector("list", length(site_layer))
for (i in seq_along(site_layer)){
message("calculating fetch for ", site_names[i], " (", i, " out of ",
length(site_layer), ")")
## Create polygon (approximating a circle) with a given radius. These
## vertices are used for creating the end points for the fetch vectors.
d_bff = gBuffer(site_layer[i, ], width = max_dist, quadsegs = n_directions)
## Calculate end points at the maximum distances.
fetch_ends = head(coordinates(d_bff@polygons[[1]]@Polygons[[1]]), -1)
fetch_ends = fetch_ends[order(directions), ]
## Create a list of Line objects radiating from the site location to the
## maximum distance for each direction.
line_list = create_line_list(site_layer[i, ], fetch_ends)
## Create a SpatialLines object (i.e. add in CRS information)
fetch_sp_lines = create_sp_lines(line_list, sort(directions), polygon_layer)
## Subset polygon coastline layer to only incorporate polygons that
## interfere with any of the fetch vectors. This will speed up computation
## times.
poly_layer_subset =
polygon_layer[which(!is.na(over(polygon_layer, fetch_sp_lines))), ]
if (length(poly_layer_subset) > 0){
## Calculate the fetch bearings that hit land
hit_land = !sapply(gIntersects(fetch_sp_lines, poly_layer_subset,
byid = c(TRUE, FALSE), returnDense = FALSE),
is.null)
## Calculate intersections and identify closest shoreline for those
## vectors that hit land
ints = gIntersection(fetch_sp_lines[hit_land], poly_layer_subset,
byid = c(TRUE, FALSE))
fetch_ends[hit_land, ] = t(sapply(ints@lines, function(x){
coordinates(x)[[1]][1, ]
}))
## Update the line list and spatialLines objects
line_list = create_line_list(site_layer[i, ], fetch_ends)
fetch_sp_lines = create_sp_lines(line_list, sort(directions), polygon_layer)
}
fetch.df = data.frame(site = site_names[i],
fetch = SpatialLinesLengths(fetch_sp_lines) / 1000,
direction = sort(directions),
quadrant = factor(quadrant,
levels = c("North", "East",
"South", "West")))
## Create a SpatialLinesDataFrame object to include the fetch lengths, and
## add it to the Fetch list
fetch_list[[i]] = SpatialLinesDataFrame(fetch_sp_lines, fetch.df)
}
new("Fetch", fetch_list, names = site_names, max_dist = max_dist / 1000)
}
Any scripts or data that you put into this service are public.
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.