nnNewMask: grid-based nearest neighbour interplation into new mask
In garciapintado/rdafCESM: Setup and Assimilation of CESM

Description Usage Arguments Details Value Author(s) Examples

Get a nearest neighbour interpolation of a 2D map from a current mask into a new one

1	nnNewMask(z2D, imask, omask, newdef=0)

`z2D`	REAL (:,:), input matrix
`imask`	LOGICAL (:,:), input mask
`omask`	LOGICAL (:,:), output mask
`newdef`	REAL, default value for cropped areas. That is, areas where imask is TRUE but omask is FALSE

This functions is to ease adaptation to new masks in CESM where both the input and the output share the grid. For example this has been used for interpolation from a preindustrial simulation to be used as initial conditions for a LGM climate situation, where the lans-sea mask has been modified. z2D, imask and omask are matrices or the same size.

A 2D matrix

Javier Garcia-Pintado

  library(rdafCESM)
  nx  <- 100
  ny  <- 200
  oo  <- c(50,100)
  sig <- 30
  xmat <- matrix(1:nx,nx,ny)
  ymat <- matrix(rep(1:ny,each=nx),nx,ny)
  dis2oosq <- (xmat - oo[1])^2 + (ymat - oo[2])^2
  A     <- matrix(exp(-dis2oosq/(2*sig^2)),nx,ny)
  imask <- matrix(FALSE,nx,ny)
  omask <- matrix(FALSE,nx,ny)
  imask[50:90,50:100] <- TRUE
  omask[40:80,60:120] <- TRUE
  A[!imask] <- 0
  Ao <- nnNewMask(A, imask, omask)
  if (1 > 2) {
    rdafPlot::imageF(A)
    rdafPlot::imageF(Ao)
    rdafPlot::imageF(Ao-A)
  }