Spatial interpolation

The irregularly-spaced data are interpolated onto regular latitude-longitude grids by weighting each station according to its distance or angle from the center of the search radius cdd.

spInterp(
  points,
  dat,
  range,
  res = 1,
  fun.weight = c("cal_weight", "cal_weight_sf"),
  wFUN = c("wFUN_adw", "wFUN_idw", "wFUN_thiessen", "wFUN_mean"),
  .parallel = FALSE,
  ...,
  Z = NULL,
  weight = NULL
)

spInterp_adw(
  points,
  dat,
  range,
  res = 1,
  fun.weight = c("cal_weight", "cal_weight_sf"),
  ...
)

Arguments

points

A matrix (N,2) with longitude and latitude of points of data observed

dat

matrix, [npoint, ntime], the observed data used to interpolate grid

range

[xmin, xmax, ymin, ymax]

res

the grid resolution (degree)

fun.weight

function to calculate weight, one of c("cal_weight", "cal_weight_sf").

wFUN

wFUN_* functions, see wFUN() for details

...

other parameters to plyr::ldply

Z

covariates, not used

weight

predefined weight to speed-up calculation, which is returned by spInterp() itself.

References

1. Xavier, A. C., King, C. W., & Scanlon, B. R. (2016). Daily gridded meteorological variables in Brazil (1980-2013). International Journal of Climatology, 36(6), 2644-2659. doi:10.1002/joc.4518

See also

wFUN(), cal_weight()

Author

Dongdong Kong and Heyang Song

Examples

library(ggplot2)

data(TempBrazil) # Temperature for some poins of Brazil

loc <- TempBrazil[, 1:2] %>% set_names(c("lon", "lat"))
dat <- TempBrazil[, 3] %>% as.matrix()  # Vector with observations in points

range <- c(-78, -34, -36, 5)
res = 1
# weight <- weight_adw(loc, range = range, res = 1)
# first example: 
r = spInterp_adw(loc, dat, range, res = res, cdd = 450)
print(str(r))
#> List of 3
#>  $ weight   :Classes ‘data.table’ and 'data.frame':	6349 obs. of  6 variables:
#>   ..$ lon  : num [1:6349] -55.5 -55.5 -55.5 -54.5 -54.5 -54.5 -53.5 -53.5 -53.5 -52.5 ...
#>   ..$ lat  : num [1:6349] -34.5 -34.5 -34.5 -34.5 -34.5 -34.5 -34.5 -34.5 -34.5 -34.5 ...
#>   ..$ I    : int [1:6349] 220 21 213 220 213 21 220 213 21 220 ...
#>   ..$ dist : num [1:6349] 224 379 421 151 357 ...
#>   ..$ angle: num [1:6349] 60.7 20.6 49.6 42.7 39.4 ...
#>   ..$ w    : num [1:6349] 0.1573 0.0418 0.0246 0.2879 0.0429 ...
#>   ..- attr(*, ".internal.selfref")=<externalptr> 
#>  $ coord    :Classes ‘data.table’ and 'data.frame':	932 obs. of  2 variables:
#>   ..$ lon: num [1:932] -73.5 -73.5 -73.5 -73.5 -72.5 -72.5 -72.5 -72.5 -72.5 -72.5 ...
#>   ..$ lat: num [1:932] -8.5 -7.5 -6.5 -5.5 -9.5 -8.5 -7.5 -6.5 -5.5 -4.5 ...
#>   ..- attr(*, ".internal.selfref")=<externalptr> 
#>   ..- attr(*, "sorted")= chr [1:2] "lon" "lat"
#>  $ predicted: num [1:932, 1] 25 25 25 25 25 ...
#>   ..- attr(*, "dimnames")=List of 2
#>   .. ..$ : NULL
#>   .. ..$ : chr "1"
#>  - attr(*, "class")= chr "spInterp"
#> NULL
df = r %$% cbind(coord, value = predicted[, 1])
ggplot(df, aes(lon, lat)) +
  geom_raster(aes(fill = value)) +
  geom_point(data = loc, size = 0.5, shape = 3, color = "red") +
  lims(x = range[1:2], y = range[3:4])


# second example: use sf as backend, but of low computing efficiency, 100 times slower
r_sf = spInterp_adw(loc, dat, range, res = res, cdd = 450, fun.weight = "cal_weight_sf")
df_sf = r_sf %$% cbind(coord, value = predicted[, 1])
ggplot(df_sf, aes(lon, lat)) +
  geom_raster(aes(fill = value)) +
  geom_point(data = loc, size = 0.5, shape = 3, color = "red") +
  lims(x = range[1:2], y = range[3:4])


# third example: use a large `cdd`, set `cdd = 1000`
r = spInterp_adw(loc, dat, range, res = res, cdd = 1000)
df = r %$% cbind(coord, value = predicted[, 1])

ggplot(df, aes(lon, lat)) +
  geom_raster(aes(fill = value)) +
  geom_point(data = loc, size = 2.5, shape = 3, color = "red") + 
  lims(x = range[1:2], y = range[3:4])