rworldmap coordinates, how to match NetCDF data to the map? - r

Question

Rworldmap looks like exactly what I need for mapping climate data, but I'm having a problem lining up the base map with the climate data. What I am mapping is ocean temperature data from JAMSTEC for August, 2015 from here:
http://www.jamstec.go.jp/ARGO/argo_web/ancient/MapQ/Mapdataset_e.html
The dataset name is TS_201508_GLB.nc. The R script I'm using is below. The country outlines are fine, but the data is for the oceans only and the data does not show in the oceans it is offset somehow. Can you tell me how to align the data to the map?
I've read lots of articles but I cannot tell how to align the two. The data I have is longitude and latitude. South latitude is negative and west longitude is negative, I don't see how they could be confused. How is the map shown, is there some sort of special convention for the lat/longs?
Thanks for any help you can provide.
The code:
library(RNetCDF)
library(sp)
library(rworldmap)
library(rgeos)
library(RColorBrewer)
library (classInt)
library(grid)
library(spam)
library(maps)
library(maptools)
library(fields)
library(methods)
library(rgdal)
library(rworldxtra)
fname <- "G:/Climate_Change/Ocean_Warming/MOAA_GPV_Jamstec_Temperature/TS_201508_GLB.nc"
moaa <- open.nc(fname)
# moaa
print.nc(moaa)
file.inq.nc(moaa)
#TOI is the temperature array extracted from the NCDF file
TOI = var.get.nc(moaa,"TOI",start=c(1,1,1),count=c(360,132,25))
TOI[1,1,1]
Long = var.get.nc(moaa,"LONGITUDE")
Lat = var.get.nc(moaa, "LATITUDE")
Pres = var.get.nc(moaa,"PRES")
# create grid
offset=c(-179.5,-60.50)
cellsize = c(abs(Long[1]-Long[2]),abs(Lat[1]-Lat[2]))
cells.dim = c(dim(Long), dim(Lat))
# create gt
gt <- GridTopology(cellcentre.offset=offset,cellsize=cellsize,cells.dim=cells.dim)
# create map window
mapDevice()
# Create a color pallette
colourPalette=c('blue','lightblue','white',brewer.pal(9,'YlOrRd'))
# Values at 2000 decibar for August 2015
ncMatrix <- TOI[,,25]
# Gridvalues
gridVals <-data.frame(att=as.vector(ncMatrix))
# create a spatialGridDataFrame
sGDF <-SpatialGridDataFrame(gt,data=gridVals)
# Vector to classify data
catMethod=seq(from=0,to=4,by=.33)
# plotting the map and getting params for legend
mapParams <- mapGriddedData( sGDF, nameColumnToPlot='att',catMethod=catMethod,colourPalette=colourPalette,addLegend=FALSE)

Answer 1

I finally figured it out. rworldmap wants the data organized from the upper left of the map(Northwest corner), that is Long = -180, Lat=90. The NetCDF data starts at Long=0 and Lat=-90(the middle of the map and south edge). So we have to reverse the values in the North-South direction:
#
# Flip the Latitude values so south is last
ncMatrix2 <- ncMatrix[,dim(Lat):1]
Then switch the values for east longitude and west longitude:
#
#Longitude values need to be from -180 to 0 then 0 to 180
# So we divide into East and West, then recombine with rbind
East_Long_values <-ncMatrix2[1:180,]
West_Long_Values <-ncMatrix2[181:360,]
ncMatrix3 <- rbind(West_Long_Values,East_Long_values)
Then everything else works.