Skip to content

ccamlr/CCAMLRGIS

master
Switch branches/tags
Code

CCAMLRGIS R package

This package was developed to simplify the production of maps in the CAMLR Convention Area. It provides two main categories of functions: load functions and create functions. Load functions are used to import spatial layers from the online CCAMLR GIS such as the ASD boundaries. Create functions are used to create layers from user data such as polygons and grids.

Note on V4 update:

Due to the retirement of some packages that the CCAMLRGIS package use to rely on, since CCAMLRGIS V4.0.0 the package relies on the sf package, users may need to familiarize themselves with it. For those that were using older versions, the main difference is in plotting commands.

Plotting a spatial object MyObject used to be:

plot(MyObject)

Since V4, it will be:

plot(st_geometry(MyObject))

Also, to access the data inside spatial objects, instead of MyObject@data, type MyObject directly. You can revert to sp objects with as_Spatial(MyObject) if preferred.

Using sf objects has advantages such as the ability to use Tidyverse methods. Further, additional plotting methods are available, some of which are described in section 5.5. Using sf.

Installation

You can install the CCAMLRGIS R package from CRAN with:

install.packages("CCAMLRGIS")

Documentation

A package to load and create spatial data, including layers and tools that are relevant to CCAMLR activities.


Table of contents


  1. Basemaps
  2. Create functions
  1. Load functions
  • 3.1. Online use
  • 3.2. Offline use
  1. Other functions
  1. Adding colors, legends and labels

Introduction

First, install the package by typing:

install.packages("CCAMLRGIS")

Then, load the package by typing:

library(CCAMLRGIS)

In order to plot bathymetry data, you will also need to load terra:

library(terra)

All spatial manipulations are made using the South Pole Lambert Azimuthal Equal Area projection (type ?CCAMLRp for more details).

#Map with axes, to illustrate projection

#Set the figure margins as c(bottom, left, top, right)
par(mai=c(1.2,0.7,0.5,0.45),xpd=TRUE)
#plot entire Coastline
plot(st_geometry(Coast[Coast$ID=='All',]),col='grey',lwd=0.1)
#Add reference grid
add_RefGrid(bb=st_bbox(Coast[Coast$ID=='All',]),ResLat=10,ResLon=20,LabLon=-40,fontsize=0.8,lwd=0.5)
#add axes and labels
axis(1,pos=0,at=seq(-4000000,4000000,by=1000000),tcl=-0.15,labels=F,lwd=0.8,lwd.ticks=0.8,col='blue')
axis(2,pos=0,at=seq(-4000000,4000000,by=1000000),tcl=-0.15,labels=F,lwd=0.8,lwd.ticks=0.8,col='blue')
text(seq(1000000,4000000,by=1000000),0,seq(1,4,by=1),cex=0.75,col='blue',adj=c(0.5,1.75))
text(seq(-4000000,-1000000,by=1000000),0,seq(-4,-1,by=1),cex=0.75,col='blue',adj=c(0.5,1.75))
text(0,seq(1000000,4000000,by=1000000),seq(1,4,by=1),cex=0.75,col='blue',adj=c(1.75,0.5))
text(0,seq(-4000000,-1000000,by=1000000),seq(-4,-1,by=1),cex=0.75,col='blue',adj=c(1.75,0.5))
text(0,0,0,cex=0.75,col='blue',adj=c(-0.5,-0.5))
text(5200000,0,expression('x ('*10^6~'m)'),cex=0.75,col='blue')
text(0,4700000,expression('y ('*10^6~'m)'),cex=0.75,col='blue')

The South Pole Lambert Azimuthal Equal Area projection converts Latitudes and Longitudes into locations on a disk with x/y axes and units of meters. The South Pole is at x=0m ; y=0m. The tip of the Peninsula, for example, is around x=-2,500,000m ; y=2,000,000m.

1. Basemaps

Bathymetry:

Prior to detailing the package’s capabilities, a set of basic commands are shown here to display a few core mapping elements. All scripts use the low-resolution bathymetry raster included in the package (‘SmallBathy’). In order to obtain higher resolution bathymetry data, use the Load_Bathy() function:

#Load_Bathy() example:
Bathy=load_Bathy(LocalFile = FALSE,Res=5000)
plot(Bathy, breaks=Depth_cuts,col=Depth_cols,axes=FALSE,legend=FALSE,mar=c(0,0,0,0))

#Please refer to ?load_Bathy for more details, including how to save the bathymetry data so that you
#do not have to re-download it every time you need it.

Statistical Areas, Subareas and Divisions (ASDs):

#Load ASDs and EEZs
ASDs=load_ASDs()
EEZs=load_EEZs()
#Set the figure margins as c(bottom, left, top, right)
par(mai=c(0,0.4,0,0))
#Plot the bathymetry
plot(SmallBathy,breaks=Depth_cuts,col=Depth_cols,legend=F,axes=F,box=F,mar=c(0,0.4,0,0))
#Add reference grid
add_RefGrid(bb=st_bbox(SmallBathy),ResLat=10,ResLon=20,LabLon=0,fontsize=0.75,lwd=0.75,offset = 4)
#Add color scale
add_Cscale(height=90,fontsize=0.75,offset=-500,width=15,maxVal=-1,lwd=0.5)
#Add ASD and EEZ boundaries
plot(st_geometry(ASDs),add=T,lwd=0.75,border='red')
plot(st_geometry(EEZs),add=T,lwd=0.75,border='red')
#Add coastline (for all ASDs)
plot(st_geometry(Coast[Coast$ID=='All',]),col='grey',lwd=0.01,add=T)
#Add ASD labels
add_labels(mode='auto',layer='ASDs',fontsize=0.6,col='red')

Local map (e.g., Subarea 48.6):

#Load ASDs
ASDs=load_ASDs()
#Subsample ASDs to only keep Subarea 48.6
S486=ASDs[ASDs$GAR_Short_Label=='486',]
#Crop bathymetry to match the extent of S486
B486=crop(rast(SmallBathy),ext(S486))
#Plot the bathymetry
plot(B486,breaks=Depth_cuts,col=Depth_cols,legend=F,axes=F,mar=c(1.4,2,1.4,7))
#Add color scale
add_Cscale(height=80,fontsize=0.7,offset=300,width=15,lwd=0.5,maxVal=-1)
#Add coastline (for Subarea 48.6 only)
plot(Coast[Coast$ID=='48.6',],col='grey',lwd=0.01,add=T)
#Add reference grid
add_RefGrid(bb=st_bbox(B486),ResLat=5,ResLon=10,fontsize=0.75,lwd=0.75,offset = 100000)
#Add Subarea 48.6 boundaries
plot(st_geometry(S486),add=T,lwd=1,border='red')
#Add a -2000m contour
contour(B486,levels=-2000,add=T,lwd=0.5,labcex=0.3)
#Add single label at the centre of the polygon (see ?Labels)
text(Labels$x[Labels$t=='48.6'],Labels$y[Labels$t=='48.6'],labels='48.6',col='red',cex=1.5)

2. Create functions

2.1. Points, lines, polygons and grids

These functions are used to transform user data into spatial layers with the appropriate projection. User data should be provided as a dataframe containing Latitudes and Longitudes in decimal degrees. Depending on the function used, some other variables may be required (see help).

Create points:

For details, type:

?create_Points
#Prepare layout for 4 sub-plots
par(mfrow=c(2,2),mai=c(0,0.01,0.2,0.01))

#Example 1: Simple points with labels
MyPoints=create_Points(PointData)
plot(st_geometry(MyPoints),main='Example 1',cex.main=0.75,cex=0.5,lwd=0.5)
text(MyPoints$x,MyPoints$y,MyPoints$name,adj=c(0.5,-0.5),xpd=T,cex=0.75)
box()

#Example 2: Simple points with labels, highlighting one group of points with the same name
MyPoints=create_Points(PointData)
plot(st_geometry(MyPoints),main='Example 2',cex.main=0.75,cex=0.5,lwd=0.5)
text(MyPoints$x,MyPoints$y,MyPoints$name,adj=c(0.5,-0.5),xpd=T,cex=0.75)
plot(st_geometry(MyPoints[MyPoints$name=='four',]),bg='red',pch=21,cex=1,add=T)
box()

#Example 3: Buffered points with radius proportional to catch
MyPoints=create_Points(PointData,Buffer=1*PointData$Catch)
plot(st_geometry(MyPoints),col='green',main='Example 3',cex.main=0.75,cex=0.5,lwd=0.5)
text(MyPoints$x,MyPoints$y,MyPoints$name,adj=c(0.5,0.5),xpd=T,cex=0.75)
box()

#Example 4: Buffered points with radius proportional to catch and clipped to the Coast
MyPoints=create_Points(PointData,Buffer=2*PointData$Catch,Clip=T)
plot(st_geometry(MyPoints),col='cyan',main='Example 4',cex.main=0.75,cex=0.75,lwd=0.5)
plot(st_geometry(Coast[Coast$ID=='All',]),add=T,col='grey',lwd=0.5)
box()

Create lines:

For details, type:

?create_Lines
#If your data contains line end locations in separate columns, you may reformat it as follows:

#Original data:
MyData=data.frame(
  Line=c(1,2),
  Lat_Start=c(-60,-65),
  Lon_Start=c(-10,5),
  Lat_End=c(-61,-66),
  Lon_End=c(-2,2)
)

#Reformat data to use as input in create_Lines as:
Input=data.frame(
  Line=c(MyData$Line,MyData$Line),
  Lat=c(MyData$Lat_Start,MyData$Lat_End),
  Lon=c(MyData$Lon_Start,MyData$Lon_End)
)
#Prepare layout for 3 sub-plots
par(mai=c(0,0.01,0.2,0.01),mfrow=c(1,3))

#Example 1: Simple and non-densified lines
MyLines=create_Lines(LineData)
plot(st_geometry(MyLines),col=rainbow(nrow(MyLines)),main='Example 1',cex.main=0.75,lwd=2)
box()

#Example 2: Simple and densified lines (note the curvature of the purple line)
MyLines=create_Lines(LineData,Densify=T)
plot(st_geometry(MyLines),col=rainbow(nrow(MyLines)),main='Example 2',cex.main=0.75,lwd=2)
box()

#Example 3: Densified, buffered and clipped lines
MyLines=create_Lines(LineData,Densify=T,Buffer=c(10,40,50,80,100),Clip=T)
plot(st_geometry(MyLines[5:1,]),col=rainbow(nrow(MyLines)),main='Example 3',cex.main=0.75,lwd=1)
plot(Coast[Coast$ID=='All',],col='grey',add=T,lwd=0.5)
box()

Adding a buffer with the argument SeparateBuf set to FALSE results in a single polygon which may be viewed as a footprint:

#Set the figure margins as c(bottom, left, top, right)
par(mai=c(0.01,0.01,0.01,0.01))

#Buffer merged lines
MyLines=create_Lines(LineData,Buffer=10,SeparateBuf=F)
#The resulting polygon has an area of:
MyLines$Buffered_AreaKm2
#> [1] 222654.8

plot(st_geometry(MyLines),col='green',lwd=1)
box()

Create polygons:

For details, type:

?create_Polys
#Prepare layout for 3 sub-plots
par(mfrow=c(1,3),mai=c(0,0.01,0.2,0.01))

#Example 1: Simple and non-densified polygons
MyPolys=create_Polys(PolyData,Densify=F)
plot(st_geometry(MyPolys),col='blue',main='Example 1',cex.main=0.75,lwd=0.5)
text(MyPolys$Labx,MyPolys$Laby,MyPolys$ID,col='white',cex=0.75)
box()

#Example 2: Simple and densified polygons (note the curvature of iso-latitude lines)
MyPolys=create_Polys(PolyData)
plot(st_geometry(MyPolys),col='red',main='Example 2',cex.main=0.75,lwd=0.5)
text(MyPolys$Labx,MyPolys$Laby,MyPolys$ID,col='white',cex=0.75)
box()

#Example 3: Buffered and clipped polygons
MyPolysBefore=create_Polys(PolyData,Buffer=c(10,-15,120))
MyPolysAfter=create_Polys(PolyData,Buffer=c(10,-15,120),Clip=T)
plot(st_geometry(MyPolysBefore),col='green',main='Example 3',cex.main=0.75,lwd=0.5)
plot(st_geometry(Coast[Coast$ID=='All',]),add=T,lwd=0.5)
plot(st_geometry(MyPolysAfter),col='orange',add=T,lwd=0.5)
text(MyPolysAfter$Labx,MyPolysAfter$Laby,MyPolysAfter$ID,col='white',cex=0.75)
box()

#Convention area
#The locations of vertices are given clockwise, starting from the northwest corner of 48.3
CA=data.frame(
  Name="CA",
  Lat=c(-50,-50,-45,-45,-55,-55,-60,-60),
  Lon=c(-50,30,30,80,80,150,150,-50)
)

#Prepare layout for 2 sub-plots
par(mfrow=c(1,2),mai=c(0,0,0.2,0))

#Example 4: Convention area contour
MyPoly=create_Polys(CA)
plot(st_geometry(MyPoly),col='blue',border='green',main='Example 4',cex.main=0.75,lwd=2)
box()

#Example 5: Convention area contour, coastline clipped
MyPoly=create_Polys(CA,Clip = TRUE)
plot(st_geometry(MyPoly),col='blue',border='green',main='Example 5',cex.main=0.75,lwd=2)
box()

Create grids:

For details, type:

?create_PolyGrids
#Prepare layout for 3 sub-plots
par(mfrow=c(1,3),mai=c(0,0.01,0.2,0.01))

#Example 1: Simple grid, using automatic colors
MyGrid=create_PolyGrids(GridData,dlon=2,dlat=1)
plot(st_geometry(MyGrid),col=MyGrid$Col_Catch_sum,main='Example 1',cex.main=0.75,lwd=0.1)
box()

#Example 2: Equal area grid, using automatic colors
MyGrid=create_PolyGrids(GridData,Area=10000)
plot(st_geometry(MyGrid),col=MyGrid$Col_Catch_sum,main='Example 2',cex.main=0.75,lwd=0.1)
box()

#Example 3: Equal area grid, using custom cuts and colors
MyGrid=create_PolyGrids(GridData,Area=10000,cuts=c(0,50,100,500,2000,3500),cols=c('blue','red'))
plot(st_geometry(MyGrid),col=MyGrid$Col_Catch_sum,main='Example 3',cex.main=0.75,lwd=0.1)
box()

Customizing a grid and adding a color scale:

#Prepare layout for 2 sub-plots
par(mfrow=c(2,1),mai=c(0.2,0.05,0.1,1.3))

#Step 1: Generate your grid
MyGrid=create_PolyGrids(GridData,Area=10000)

#Step 2: Inspect your gridded data (e.g. sum of Catch) to determine whether irregular cuts are required
hist(MyGrid$Catch_sum,100,cex=0.75,main='Frequency distribution of data',
     cex.main=0.5,col='grey',axes=F)
axis(1,pos=0,tcl=-0.15,lwd=0.8,lwd.ticks=0.8,labels=F)
text(seq(0,2500,by=500),-1.5,seq(0,2500,by=500),cex=0.75,xpd=T)

#In this case (heterogeneously distributed data) irregular cuts would be preferable
#Such as:
MyCuts=c(0,50,100,500,2000,2500)
abline(v=MyCuts,col='green',lwd=0.1,lty=2) #Add classes to histogram as green dashed lines

#Step 3: Generate colors according to the desired classes (MyCuts)
Gridcol=add_col(MyGrid$Catch_sum,cuts=MyCuts,cols=c('yellow','purple'))

#Step 4: Plot result and add color scale
#Use the colors generated by add_col
plot(st_geometry(MyGrid),col=Gridcol$varcol,lwd=0.1) 
#Add color scale using cuts and cols generated by add_col
add_Cscale(title='Sum of Catch (t)',cuts=Gridcol$cuts,cols=Gridcol$cols,width=18,
     fontsize=0.6,lwd=0.5,height = 100) 
box()

2.2. Create Stations

This function was designed to create random point locations inside a polygon and within bathymetry strata constraints. A distance constraint between stations may also be used if desired. The examples below use the ‘SmallBathy’ data for illustrative purposes; users should use a higher resolution bathymetry dataset instead, as obtained via the load_Bathy() function.

For details, type:

?create_Stations

First, create a polygon within which stations will be created:

#Create polygons
MyPoly=create_Polys(
        data.frame(Name="mypol",
             Latitude=c(-75,-75,-70,-70),
             Longitude=c(-170,-180,-180,-170))
        ,Densify=T)

#Set the figure margins as c(bottom, left, top, right)
par(mai=c(0,0,0,0))
plot(st_geometry(Coast[Coast$ID=='88.1',]),col='grey')
plot(st_geometry(MyPoly),col='green',add=T)
text(MyPoly$Labx,MyPoly$Laby,MyPoly$ID)
box()

Example 1. Set numbers of stations, no distance constraint:

#optional: crop your bathymetry raster to match the extent of your polygon
BathyCroped=crop(rast(SmallBathy),ext(MyPoly))

#Create stations
MyStations=create_Stations(MyPoly,BathyCroped,Depths=c(-2000,-1500,-1000,-550),N=c(20,15,10))

#add custom colors to the bathymetry to indicate the strata of interest
MyCols=add_col(var=c(-10000,10000),cuts=c(-2000,-1500,-1000,-550),cols=c('blue','cyan'))
plot(BathyCroped,breaks=MyCols$cuts,col=MyCols$cols,legend=F,axes=F,main="Example 1")
add_Cscale(height=90,fontsize=0.75,width=16,lwd=0.5,offset=-130,cuts=MyCols$cuts,cols=MyCols$cols)
plot(st_geometry(MyPoly),add=T,border='red',lwd=2,xpd=T)
plot(st_geometry(MyStations),add=T,col='orange',cex=0.75,lwd=1.5,pch=3)

Example 2. Set numbers of stations, with distance constraint:

#Create Stations
MyStations=create_Stations(MyPoly,BathyCroped,
                           Depths=c(-2000,-1500,-1000,-550),N=c(20,15,10),dist=10)

#add custom colors to the bathymetry to indicate the strata of interest
MyCols=add_col(var=c(-10000,10000),cuts=c(-2000,-1500,-1000,-550),cols=c('blue','cyan'))
plot(BathyCroped,breaks=MyCols$cuts,col=MyCols$cols,legend=F,axes=F,main="Example 2")
add_Cscale(height=90,fontsize=0.75,width=16,lwd=0.5,offset=-130,cuts=MyCols$cuts,cols=MyCols$cols)
plot(st_geometry(MyPoly),add=T,border='red',lwd=2,xpd=T)
plot(st_geometry(MyStations[MyStations$Stratum=='1000-550',]),pch=21,bg='yellow',add=T,cex=0.75,lwd=0.1)
plot(st_geometry(MyStations[MyStations$Stratum=='1500-1000',]),pch=21,bg='orange',add=T,cex=0.75,lwd=0.1)
plot(st_geometry(MyStations[MyStations$Stratum=='2000-1500',]),pch=21,bg='red',add=T,cex=0.75,lwd=0.1)

Example 3. Automatic numbers of stations, with distance constraint:

#Create Stations
MyStations=create_Stations(MyPoly,BathyCroped,Depths=c(-2000,-1500,-1000,-550),Nauto=30,dist=10)

#add custom colors to the bathymetry to indicate the strata of interest
MyCols=add_col(var=c(-10000,10000),cuts=c(-2000,-1500,-1000,-550),cols=c('blue','cyan'))
plot(BathyCroped,breaks=MyCols$cuts,col=MyCols$cols,legend=F,axes=F,main="Example 3")
add_Cscale(height=90,fontsize=0.75,width=16,lwd=0.5,offset=-130,cuts=MyCols$cuts,cols=MyCols$cols)
plot(st_geometry(MyPoly),add=T,border='red',lwd=2,xpd=T)
plot(st_geometry(MyStations[MyStations$Stratum=='1000-550',]),pch=21,bg='yellow',add=T,cex=0.75,lwd=0.1)
plot(st_geometry(MyStations[MyStations$Stratum=='1500-1000',]),pch=21,bg='orange',add=T,cex=0.75,lwd=0.1)
plot(st_geometry(MyStations[MyStations$Stratum=='2000-1500',]),pch=21,bg='red',add=T,cex=0.75,lwd=0.1)

2.3. Create pies

The function create_Pies() generates pie charts that can be overlaid on maps. The Input data must be a dataframe with, at least, columns for latitude, longitude, class and value. For each location, a pie is created with pie pieces for each class, and the size of each pie piece depends on the proportion of each class (the value of each class divided by the sum of values). Optionally, the area of each pie can be proportional to a chosen variable (if that variable is different than the value mentioned above, the Input data must have a fifth column and that variable must be unique to each location). If the Input data contains locations that are too close together, the data can be gridded by setting GridKm (see Examples 6-8). Once pie charts have been created, the function add_PieLegend() may be used to add a legend to the figure.

For details, type:

?create_Pies
?add_PieLegend
#The examples below use the following example datasets:
View(PieData)
View(PieData2)

Example 1. Pies of constant size, all classes displayed:

#Plot the bathymetry (See section 'Local map' where B486 was created)
plot(B486,breaks=Depth_cuts,col=Depth_cols,legend=FALSE,axes=FALSE,mar=c(6,0,0,0))
#Add coastline
plot(Coast[Coast$ID=='48.6',],col='grey',lwd=0.01,add=T)
#Create pies
MyPies=create_Pies(Input=PieData,
                   NamesIn=c("Lat","Lon","Sp","N"),
                   Size=50
                   )
#Plot Pies
plot(st_geometry(MyPies),col=MyPies$col,add=TRUE)
#Add Pies legend
add_PieLegend(Pies=MyPies,PosX=-0.1,PosY=-1.6,Boxexp=c(0.5,0.45,0.12,0.45),
              PieTitle="Species")


Example 2. Pies of constant size, selected classes displayed:

#Plot the bathymetry (See section 'Local map' where B486 was created)
plot(B486,breaks=Depth_cuts,col=Depth_cols,legend=FALSE,axes=FALSE,mar=c(6,0,0,0))
#Add coastline
plot(Coast[Coast$ID=='48.6',],col='grey',lwd=0.01,add=T)
#Create pies
MyPies=create_Pies(Input=PieData,
                   NamesIn=c("Lat","Lon","Sp","N"),
                   Size=50,
                   Classes=c("TOP","TOA","ANI")
                   )
#Plot Pies
plot(st_geometry(MyPies),col=MyPies$col,add=TRUE)
#Add Pies legend
add_PieLegend(Pies=MyPies,PosX=-0.1,PosY=-1.6,Boxexp=c(0.6,0.6,0.12,0.55),
              PieTitle="Selected species")


Example 3. Pies of constant size, proportions below 25% are grouped in a ‘Other’ class (N.B.: unlike Example 2, the ‘Other’ class may contain classes that are displayed in the legend. Please compare Example 1 and Example 3):

#Plot the bathymetry (See section 'Local map' where B486 was created)
plot(B486,breaks=Depth_cuts,col=Depth_cols,legend=FALSE,axes=FALSE,mar=c(6,0,0,0))
#Add coastline
plot(Coast[Coast$ID=='48.6',],col='grey',lwd=0.01,add=T)
#Create pies
MyPies=create_Pies(Input=PieData,
                   NamesIn=c("Lat","Lon","Sp","N"),
                   Size=50,
                   Other=25
                   )
#Plot Pies
plot(st_geometry(MyPies),col=MyPies$col,add=TRUE)
#Add Pies legend
add_PieLegend(Pies=MyPies,PosX=-0.1,PosY=-1.6,Boxexp=c(0.55,0.55,0.12,0.45),
              PieTitle="Other (%) class")


Example 4. Pies of variable size (here, their area is proportional to ‘Catch’), all classes displayed, horizontal legend:

#Plot the bathymetry (See section 'Local map' where B486 was created)
plot(B486,breaks=Depth_cuts,col=Depth_cols,legend=FALSE,axes=FALSE,mar=c(6,0,0,0))
#Add coastline
plot(Coast[Coast$ID=='48.6',],col='grey',lwd=0.01,add=T)
#Create pies
MyPies=create_Pies(Input=PieData,
                   NamesIn=c("Lat","Lon","Sp","N"),
                   Size=18,
                   SizeVar="Catch"
                   )
#Plot Pies
plot(st_geometry(MyPies),col=MyPies$col,add=TRUE)
#Add Pies legend
add_PieLegend(Pies=MyPies,PosX=-0.1,PosY=-1.6,Boxexp=c(0.16,0.1,0.1,0.4),
              PieTitle="Species",SizeTitle="Catch (t.)")


Example 5. Pies of variable size (here, their area is proportional to ‘Catch’), all classes displayed, vertical legend:

#Plot the bathymetry (See section 'Local map' where B486 was created)
plot(B486,breaks=Depth_cuts,col=Depth_cols,legend=FALSE,axes=FALSE,mar=c(0,0,0,10))
#Add coastline
plot(Coast[Coast$ID=='48.6',],col='grey',lwd=0.01,add=T)
#Create pies
MyPies=create_Pies(Input=PieData,
                   NamesIn=c("Lat","Lon","Sp","N"),
                   Size=18,
                   SizeVar="Catch"
                   )
#Plot Pies
plot(st_geometry(MyPies),col=MyPies$col,add=TRUE)
#Add Pies legend
add_PieLegend(Pies=MyPies,PosX=2.32,PosY=0.1,Boxexp=c(0.35,0.32,0.02,0.15),
              PieTitle="Species",SizeTitle="Catch (t.)",Horiz=FALSE,LegSp=0.6)


Example 6. Pies of constant size, all classes displayed. Too many pies (see next example for solution):

#Plot the bathymetry (See section 'Local map' where B486 was created)
plot(B486,breaks=Depth_cuts,col=Depth_cols,legend=FALSE,axes=FALSE,mar=c(6,0,0,0))
#Add coastline
plot(Coast[Coast$ID=='48.6',],col='grey',lwd=0.01,add=T)
#Create pies
MyPies=create_Pies(Input=PieData2,
                   NamesIn=c("Lat","Lon","Sp","N"),
                   Size=5
                   )
#Plot Pies
plot(st_geometry(MyPies),col=MyPies$col,add=TRUE)
#Add Pies legend
add_PieLegend(Pies=MyPies,PosX=0.4,PosY=-1.5,Boxexp=c(0.5,0.45,0.12,0.45),
              PieTitle="Species")


Example 7. Pies of constant size, all classes displayed. Gridded locations (in which case numerical variables in the Input are summed for each grid point):

#Plot the bathymetry (See section 'Local map' where B486 was created)
plot(B486,breaks=Depth_cuts,col=Depth_cols,legend=FALSE,axes=FALSE,mar=c(6,0,0,0))
#Add coastline
plot(Coast[Coast$ID=='48.6',],col='grey',lwd=0.01,add=T)
#Create pies
MyPies=create_Pies(Input=PieData2,
                   NamesIn=c("Lat","Lon","Sp","N"),
                   Size=5,
                   GridKm=250
                   )
#Plot Pies
plot(st_geometry(MyPies),col=MyPies$col,add=TRUE)
#Add Pies legend
add_PieLegend(Pies=MyPies,PosX=0.4,PosY=-1.3,Boxexp=c(0.5,0.45,0.12,0.45),
              PieTitle="Species")


Example 8. Pies of variable size (here, their area is proportional to ‘Catch’), all classes displayed, vertical legend, gridded locations (in which case numerical variables in the Input are summed for each grid point):

#Plot the bathymetry (See section 'Local map' where B486 was created)
plot(B486,breaks=Depth_cuts,col=Depth_cols,legend=FALSE,axes=FALSE,mar=c(0,0,0,10))
#Add coastline
plot(Coast[Coast$ID=='48.6',],col='grey',lwd=0.01,add=T)
#Create pies
MyPies=create_Pies(Input=PieData2,
                   NamesIn=c("Lat","Lon","Sp","N"),
                   Size=3,
                   GridKm=250,
                   SizeVar='Catch'
                   )
#Plot Pies
plot(st_geometry(MyPies),col=MyPies$col,add=TRUE)
#Add Pies legend
add_PieLegend(Pies=MyPies,PosX=2.8,PosY=0.15,Boxexp=c(0.38,0.32,0.08,0.18),
              PieTitle="Species",Horiz=FALSE,SizeTitle="Catch (t.)",
              SizeTitleVadj=0.8,nSizes=2)


3. Load functions

3.1. Online use

Download the up-to-date spatial layers from the online CCAMLR GIS and load them to your environment.

For details, type:

?load_ASDs
?load_Bathy
#Load ASDs, EEZs, and Coastline
ASDs=load_ASDs()
EEZs=load_EEZs()
Coastline=load_Coastline()

#Set the figure margins as c(bottom, left, top, right)
par(mai=c(0,0,0,0))
#Plot
plot(st_geometry(ASDs),col='green',border='blue')
plot(st_geometry(EEZs),col='orange',border='purple',add=T)
plot(st_geometry(Coastline),col='grey',add=T)
add_labels(mode='auto',layer='ASDs',fontsize=0.75,col='red')
box()

3.2. Offline use

Since the ‘load_’ functions require an internet connection, users may desire to save layers on their hard drive for offline use. This may be done, at the risk of not having the most up-to-date layers, as follows:

#Load all layers
ASDs=load_ASDs()
EEZs=load_EEZs()
Coastline=load_Coastline()
SSRUs=load_SSRUs()
RBs=load_RBs()
SSMUs=load_SSMUs()
MAs=load_MAs()
MPAs=load_MPAs()

#Save as .RData file (here in the temp directory, but users might want to chose their own directory)
save(list=c('ASDs','EEZs','Coastline','SSRUs','RBs','SSMUs','MAs','MPAs'),
     file = file.path(tempdir(), "CCAMLRLayers.RData"), compress='xz')

#Later, when offline load layers:
load(file.path(tempdir(), "CCAMLRLayers.RData"))

The load_Bathy() function may also be used to download and store bathymetry data for later use, see ?load_Bathy for details.

4. Other functions

4.1. get_depths

Given a bathymetry raster and an input dataframe of latitudes/longitudes, this function computes the depths at these locations. The examples below use the ‘SmallBathy’ data for illustrative purposes; users should use a higher resolution bathymetry dataset instead, as obtained via the load_Bathy() function.

For details, type:

?get_depths
#Generate a dataframe
MyData=data.frame(Lat=PointData$Lat,
                  Lon=PointData$Lon,
                  Catch=PointData$Catch)
#The input data looks like this:
head(MyData)
#>         Lat       Lon    Catch
#> 1 -68.63966 -175.0078 53.33002
#> 2 -67.03475 -178.0322 38.66385
#> 3 -65.44164 -170.1656 20.32608
#> 4 -68.36806  151.0247 69.81201
#> 5 -63.89171  154.4327 52.32101
#> 6 -66.35370  153.6906 78.65576

#Get depths of locations
MyDataD=get_depths(Input=MyData,Bathy=SmallBathy)
#The resulting data looks like this (where 'd' is the depth):
head(MyDataD)
#>         Lat       Lon    Catch          d
#> 1 -68.63966 -175.0078 53.33002 -3790.7695
#> 2 -67.03475 -178.0322 38.66385 -3959.3145
#> 3 -65.44164 -170.1656 20.32608 -3014.6553
#> 4 -68.36806  151.0247 69.81201  -336.2152
#> 5 -63.89171  154.4327 52.32101 -3234.9985
#> 6 -66.35370  153.6906 78.65576 -1955.7587

#Plot Catch vs Depth
plot(MyDataD$d,MyDataD$Catch,xlab='Depth',ylab='Catch',pch=21,bg='red')

4.2. seabed_area

Function to calculate planimetric seabed area within polygons and depth strata in square kilometers. Its accuracy depends on the input bathymetry raster. The examples below use the ‘SmallBathy’ data for illustrative purposes; users should use a higher resolution bathymetry dataset instead, as obtained via the load_Bathy() function. Higher accuracy will be obtained with raw, unprojected bathymetry data.

For details, type:

?seabed_area
#create some polygons
MyPolys=create_Polys(PolyData,Densify=T)
#compute the seabed areas
FishDepth=seabed_area(SmallBathy,MyPolys,PolyNames="ID",depth_classes=c(0,-200,-600,-1800,-3000,-5000))
#Result looks like this (note that the 600-1800 stratum is renamed 'Fishable_area')
head(FishDepth)
#>      ID 0|-200 -200|-600 Fishable_area -1800|-3000 -3000|-5000
#> 1   one      0  19100.01      41400.01    40500.01    92700.03
#> 2   two      0    200.00       1800.00     9300.00    93300.03
#> 3 three    700   1600.00       8100.00   229400.07   138000.04

4.3. assign_areas

Given a set of polygons and a set of point locations (given in decimal degrees), finds in which polygon those locations fall. Finds, for example, in which ASD the given fishing locations occurred.

For details, type:

?assign_areas
#Generate a dataframe with random locations
MyData=data.frame(Lat=runif(100,min=-65,max=-50),
                  Lon=runif(100,min=20,max=40))
#The input data looks like this:
head(MyData)
#>         Lat      Lon
#> 1 -53.11870 34.20721
#> 2 -55.81513 25.69306
#> 3 -61.87161 29.29898
#> 4 -64.10882 34.58778
#> 5 -55.62069 22.56286
#> 6 -52.94103 38.68591

#load ASDs and SSRUs
ASDs=load_ASDs()
SSRUs=load_SSRUs()

#Assign ASDs and SSRUs to these locations 
MyData=assign_areas(MyData,Polys=c('ASDs','SSRUs'),NamesOut=c('MyASDs','MySSRUs'))
#The output data looks like this:
head(MyData)
#>         Lat      Lon  MyASDs   MySSRUs
#> 1 -53.11870 34.20721 58.4.4a 58.4.4a D
#> 2 -55.81513 25.69306    48.6    48.6 G
#> 3 -61.87161 29.29898    48.6    48.6 F
#> 4 -64.10882 34.58778  58.4.2  58.4.2 A
#> 5 -55.62069 22.56286    48.6    48.6 G
#> 6 -52.94103 38.68591 58.4.4a 58.4.4a D

#count of locations per ASD
table(MyData$MyASDs) 
#> 
#>    48.6  58.4.2 58.4.4a 
#>      53       7      40

#count of locations per SSRU
table(MyData$MySSRUs) 
#> 
#>    48.6 F    48.6 G  58.4.2 A 58.4.4a D 
#>        17        36         7        40

4.4. project_data

A simple function to project user-supplied locations. Input must be a dataframe, outputs may be appended to the dataframe.

For details, type:

?project_data
#The input data looks like this:
head(PointData)
#>         Lat       Lon  name    Catch Nfishes n
#> 1 -68.63966 -175.0078   one 53.33002     460 1
#> 2 -67.03475 -178.0322   two 38.66385     945 2
#> 3 -65.44164 -170.1656   two 20.32608     374 3
#> 4 -68.36806  151.0247   two 69.81201      87 4
#> 5 -63.89171  154.4327 three 52.32101     552 5
#> 6 -66.35370  153.6906  four 78.65576      22 6
#Generate a dataframe with random locations
MyData=project_data(Input=PointData,NamesIn=c('Lat','Lon'),
                    NamesOut=c('Projected_Y','Projected_X'),append=TRUE)
#The output data looks like this:
head(MyData)
#>         Lat       Lon  name    Catch Nfishes n Projected_Y Projected_X
#> 1 -68.63966 -175.0078   one 53.33002     460 1    -2361962  -206321.41
#> 2 -67.03475 -178.0322   two 38.66385     945 2    -2545119   -87445.72
#> 3 -65.44164 -170.1656   two 20.32608     374 3    -2680488  -464656.29
#> 4 -68.36806  151.0247   two 69.81201      87 4    -2100218  1162986.84
#> 5 -63.89171  154.4327 three 52.32101     552 5    -2606157  1246832.20
#> 6 -66.35370  153.6906  four 78.65576      22 6    -2349505  1161675.96

4.5. get_C_intersection

Get Cartesian coordinates of lines intersection in Euclidean space. This may have several uses, including when creating polygons with shared boundaries. Uses the coordinates of line extremities as input.

For details, type:

?get_C_intersection
#Prepare layout for 4 sub-plots
par(mfrow=c(2,2),mai=c(0.8,0.8,0.2,0.05))

#Example 1 (Intersection beyond the range of segments)
get_C_intersection(Line1=c(-30,-55,-29,-50),Line2=c(-50,-60,-40,-60))
#> Lon Lat 
#> -31 -60
text(-40,-42,"Example 1",xpd=T)
box()
#Example 2 (Intersection on one of the segments)
get_C_intersection(Line1=c(-30,-65,-29,-50),Line2=c(-50,-60,-40,-60))
#>       Lon       Lat 
#> -29.66667 -60.00000
text(-40,-41,"Example 2",xpd=T)
box()
#Example 3 (Crossed segments)
get_C_intersection(Line1=c(-30,-65,-29,-50),Line2=c(-50,-60,-25,-60))
#>       Lon       Lat 
#> -29.66667 -60.00000
text(-38,-41,"Example 3",xpd=T)
box()
#Example 4 (Antimeridian crossed)
get_C_intersection(Line1=c(-179,-60,-150,-50),Line2=c(-120,-60,-130,-62))
#> Warning in get_C_intersection(Line1 = c(-179, -60, -150, -50), Line2 =
#> c(-120, : Antimeridian crossed. Find where your line crosses it first, using
#> Line=c(180,-90,180,0) or Line=c(-180,-90,-180,0).
#>        Lon        Lat 
#> -260.47619  -88.09524
text(-180,-37,"Example 4",xpd=T)
box()

5. Adding colors, legends and labels

5.1. Bathymetry colors

Coloring bathymetry requires a vector of depth classes and a vector of colors. Colors are applied between depth classes (so there is one less color than there are depth classes). Two sets of bathymetry colors are included in the package. One simply colors the bathymetry in shades of blue (Depth_cols and Depth_cuts), the other adds shades of green to highlight the Fishable Depth range (600-1800m; Depth_cols2 and Depth_cuts2). The examples below use the ‘SmallBathy’ data for illustrative purposes; users should use a higher resolution bathymetry dataset instead, as obtained via the load_Bathy() function.

Simple set of colors:

#Set the figure margins as c(bottom, left, top, right)
par(mai=c(0,0.4,0,0))
#Plot the bathymetry
plot(SmallBathy,breaks=Depth_cuts,col=Depth_cols,axes=FALSE,box=FALSE,legend=FALSE)
#Add color scale
add_Cscale(cuts=Depth_cuts,cols=Depth_cols,fontsize=0.75,height=80,offset=-500,width=16,maxVal=-1)

Highlighting the Fishable Depth range:

#Set the figure margins as c(bottom, left, top, right)
par(mai=c(0,0.4,0,0))
#Plot the bathymetry
plot(SmallBathy,breaks=Depth_cuts2,col=Depth_cols2,axes=FALSE,box=FALSE,legend=FALSE)
#Add color scale
add_Cscale(cuts=Depth_cuts2,cols=Depth_cols2,fontsize=0.75,height=80,offset=-500,width=16,maxVal=-1)

5.2. Adding colors to data

Adding colors to plots revolves around two functions:

?add_col
#and
?add_Cscale

add_col() generates colors for a variable of interest as well as a set of color classes and colors to be used as inputs to the add_Cscale() function. Colors and color classes may be generated automatically or customized, depending on the intended appearance. Knowing the names of colors in R would be useful here (http://www.stat.columbia.edu/~tzheng/files/Rcolor.pdf).

#Adding color to points

#Prepare layout for 3 sub-plots
par(mfrow=c(3,1),mai=c(0.1,0.1,0.1,1))
#Create some points
MyPoints=create_Points(PointData)

#Example 1: Add default cols and cuts
MyCols=add_col(MyPoints$Nfishes) 
plot(st_geometry(MyPoints),pch=21,bg=MyCols$varcol,main='Example 1:',cex.main=0.75,cex=1.5,lwd=0.5)
box()
add_Cscale(title='Number of fishes',
           height=95,fontsize=0.75,width=16,lwd=1,offset=0,
           cuts=MyCols$cuts,cols=MyCols$cols)

#Example 2: Given the look of example 1, reduce the number of cuts and round their values (in add_Cscale)
MyCols=add_col(MyPoints$Nfishes,cuts=10) 
plot(st_geometry(MyPoints),pch=21,bg=MyCols$varcol,main='Example 2:',cex.main=0.75,cex=1.5,lwd=0.5)
box()
add_Cscale(title='Number of fishes',
           height=95,fontsize=0.75,width=16,lwd=1,offset=0,
           cuts=round(MyCols$cuts,1),cols=MyCols$cols)

#Example 3: same as example 2 but with custom colors
MyCols=add_col(MyPoints$Nfishes,cuts=10,cols=c('black','yellow','purple','cyan')) 
plot(st_geometry(MyPoints),pch=21,bg=MyCols$varcol,main='Example 3:',cex.main=0.75,cex=1.5,lwd=0.5)
add_Cscale(title='Number of fishes',
           height=95,fontsize=0.75,width=16,lwd=1,offset=0,
           cuts=round(MyCols$cuts,1),cols=MyCols$cols)
box()

#Adding colors to a grid with custom cuts (see also the last example in section 2.1.)

#Step 1: Generate your grid
MyGrid=create_PolyGrids(GridData,Area=10000)

#Step 2: Inspect your gridded data (e.g. hist(MyGrid$Catch_sum,100))
#to determine whether irregular cuts are required.
#In this case (heterogeneously distributed data) irregular cuts 
#would be preferable, such as:
MyCuts=c(0,50,100,500,2000,2500)

#Step 3: Generate colors according to the desired classes (MyCuts)
Gridcol=add_col(MyGrid$Catch_sum,cuts=MyCuts,cols=c('blue','white','red'))

#Step 4: Plot result and add color scale
par(mai=c(0,0,0,1.5)) #set plot margins as c(bottom, left, top, right)
#Use the colors generated by add_col
plot(st_geometry(MyGrid),col=Gridcol$varcol,lwd=0.1) 
#Add color scale using cuts and cols generated by add_col
add_Cscale(title='Sum of Catch (t)',cuts=Gridcol$cuts,cols=Gridcol$cols,width=22,
     fontsize=0.75,lwd=1) 

5.3. Adding legends

To add a legend, use the base legend() function:

?legend

To position the legend, the add_Cscale() function can generate legend coordinates which correspond to the top-left corner of the legend box. These may be adjusted using the ‘pos’, ‘height’ and ‘offset’ arguments within add_Cscale(), e.g.:

Legend_Coordinates=add_Cscale(pos='2/3',offset=1000,height=40,mode="Legend")
#Adding a color scale and a legend

#Create some point data
MyPoints=create_Points(PointData)

#Crop the bathymetry to match the extent of MyPoints (extended extent)
BathyCr=crop(rast(SmallBathy),extend(ext(MyPoints),100000))
#Plot the bathymetry
plot(BathyCr,breaks=Depth_cuts,col=Depth_cols,legend=F,axes=F,mar=c(0,0,0,7))
#Add a color scale
add_Cscale(pos='3/8',height=50,maxVal=-1,minVal=-4000,fontsize=0.75,lwd=1,width=16)

#Plot points with different symbols and colors (see ?points)
Psymbols=c(21,22,23,24)
Pcolors=c('red','green','blue','yellow')
plot(st_geometry(MyPoints[MyPoints$name=='one',]),pch=Psymbols[1],bg=Pcolors[1],add=T)
plot(st_geometry(MyPoints[MyPoints$name=='two',]),pch=Psymbols[2],bg=Pcolors[2],add=T)
plot(st_geometry(MyPoints[MyPoints$name=='three',]),pch=Psymbols[3],bg=Pcolors[3],add=T)
plot(st_geometry(MyPoints[MyPoints$name=='four',]),pch=Psymbols[4],bg=Pcolors[4],add=T)

#Add legend with position determined by add_Cscale
Loc=add_Cscale(pos='7/8',height=40,mode='Legend')
legend(Loc,legend=c('one','two','three','four'),
       title='Vessel',pch=Psymbols,pt.bg=Pcolors,xpd=T,
       box.lwd=1,cex=0.75,pt.cex=1,y.intersp=1.5)

5.4. Adding labels

To add labels, use the add_labels() function:

?add_labels

Three modes are available within the add_labels() function:

  • In ‘auto’ mode, labels are placed at the centres of polygon parts of spatial objects loaded via the load_ functions.
  • In ‘manual’ mode, users may click on their plot to position labels. An editable label table is generated to allow fine-tuning of labels appearance, and may be saved for later use. To edit the label table, double-click inside one of its cells, edit the value, then close the table.
  • In ‘input’ mode, a label table that was generated in ‘manual’ mode is re-used.
#Example 1: 'auto' mode
#label ASDs in bold and red
ASDs=load_ASDs()
#set plot margins as c(bottom, left, top, right)
par(mai=c(0,0,0,0))
plot(st_geometry(ASDs))
add_labels(mode='auto',layer='ASDs',fontsize=0.75,fonttype=2,col='red')
#add MPAs and EEZs and their labels in large, green and vertical text
MPAs=load_MPAs()
EEZs=load_EEZs()
plot(st_geometry(MPAs),add=TRUE,border='green')
plot(st_geometry(EEZs),add=TRUE,border='green')
add_labels(mode='auto',layer=c('EEZs','MPAs'),fontsize=1,col='green',angle=90)

#Example 2: 'auto' and 'input' modes
#This example is not executed here because it needs user interaction.
#Please copy and paste it in the Console to see how it works.

#Prepare a basemap
plot(SmallBathy)
ASDs=load_ASDs()
plot(st_geometry(ASDs),add=T)

#Build your labels
MyLabels=add_labels(mode='manual') 

#Re-use the label table generated (if desired)
plot(SmallBathy)
plot(st_geometry(ASDs),add=T)
add_labels(mode='input',LabelTable=MyLabels)

5.5. Using sf

Depending on the function used, the CCAMLRGIS package computes data summaries and includes them in the resulting spatial object. For example, create_Polys takes any numerical values included in the Input data frame and computes, for each polygon, the minimum, maximum, mean, median, sum, count and standard deviation of values associated with each polygon. The sf package has some useful plotting methods, some of which are shown below.

#First, let's create some example polygons
MyPolys=create_Polys(PolyData)

#MyPolys is an sf object; it is a data frame that includes a column named 'geometry':
kableExtra::kable(MyPolys,row.names = F)
ID Catch_min Nfishes_min n_min Catch_max Nfishes_max n_max Catch_mean Nfishes_mean n_mean Catch_sum Nfishes_sum n_sum Catch_count Nfishes_count n_count Catch_sd Nfishes_sd n_sd Catch_median Nfishes_median n_median geometry AreaKm2 Labx Laby
one 52.61262 11 1 71.65909 329 4 64.17380 172.5000 2.5 256.6952 690 10 4 4 4 9.084736 153.3917 1.290994 66.21175 175.0 2.5 POLYGON ((-290035.9 -164487… 187281.3 -170519.8 -1949051
two 23.12032 116 5 73.49383 954 8 51.94951 505.0000 6.5 207.7980 2020 26 4 4 4 22.264999 428.9188 1.290994 55.59195 475.0 6.5 POLYGON ((-423880.7 -240394… 95294.2 0.0 -2483470
three 10.23393 13 9 95.57774 988 14 52.50313 412.3333 11.5 315.0188 2474 69 6 6 6 32.152675 382.8685 1.870829 54.15367 341.5 11.5 POLYGON ((480755.1 -2726497… 361556.2 786933.1 -2846388

The ‘geometry’ column contains the locations of each point of a given polygon (each row), and can be plotted using plot(st_geometry(MyPolys)), as shown previously in this document. Alternatively, one can use plot(MyPolys) directly:

plot(MyPolys)
#> Warning: plotting the first 9 out of 25 attributes; use max.plot = 25 to plot
#> all

This results in a warning Warning: plotting the first 9 out of 25 attributes… and a 9-panel plot as shown above, with each panel corresponding to each column present in MyPolys and automatic colors generated according to the values in each column. In order to plot only one variable, it must be named in the plotting command:

plot(MyPolys["Catch_mean"])

There are several available options, for example:

Gr=st_graticule(MyPolys,lon=seq(-180,180,by=5),lat=seq(-80,0,by=2.5))
plot(MyPolys["Catch_mean"],
     graticule=Gr,axes=T,key.pos=1,key.width=0.2,key.length=0.8,breaks=seq(50,65,by=2.5))

Where:

  • key.pos controls the color legend position as 1=below, 2=left, 3=above and 4=right,

  • key.width and key.length control the size of the color legend,

  • breaks controls the classes,

  • The function st_graticule generates a Lat/Lon grid.

Additionally, sf objects can be plotted using ggplot2. For example:

library(ggplot2)
ggplot() + 
  geom_sf(data = MyPolys, aes(fill = Catch_mean))

Using ggplot2 and gridExtra, multi-panel plots can be drawn:

library(gridExtra)
map1 <- ggplot() +
  geom_sf(data = MyPolys, aes(fill = Catch_mean)) + 
  labs(title="Mean catch")

map2 <- ggplot() +
  geom_sf(data = MyPolys, aes(fill = Catch_sd)) + 
  labs(title="S.D. of catch")

map3 <- ggplot() +
  geom_sf(data = MyPolys, aes(fill = AreaKm2)) + 
  labs(title="Polygon area")

grid.arrange(map1, map2, map3, ncol=2)

About

A package to load and create spatial data, including layers and tools that are relevant to CCAMLR activities.

Resources

Stars

Watchers

Forks

Packages

No packages published

Contributors 4

  •  
  •  
  •  
  •  

Languages