Baranja hill case study

Description

Baranja hill is a 4 by 4 km large study area in the Baranja region, eastern Croatia (corresponds to a size of an aerial photograph). This data set has been extensively used to describe various DEM modelling and analysis steps (see Hengl and Reuter, 2008; Hengl et al., 2010).
barxyz contains 6370 precise observations of elevations (from field survey and digitized from the stereo images); bargrid contains observed probabilities of streams (digitized from the 1:5000 topo map); barstr contains 100 simulated stream networks ("SpatialLines") using barxyz point data as input (see examples below).

Usage

data(bargrid)

Format

The bargrid data frame (regular grid at 30 m intervals) contains the following columns:

p.obs

observed probability of stream (0-1)

x

a numeric vector; x-coordinate (m) in the MGI / Balkans zone 6

y

a numeric vector; y-coordinate (m) in the MGI / Balkans zone 6

Note

Consider using the 30 m resolution grid (see bargrid) as the target resolution (output maps).

Author(s)

Tomislav Hengl

References

• Hengl, T., Reuter, H.I. (eds), (2008) Geomorphometry: Concepts, Software, Applications. Developments in Soil Science, vol. 33, Elsevier, 772 p.

• Hengl, T., Heuvelink, G. B. M., van Loon, E. E., (2010) On the uncertainty of stream networks derived from elevation data: the error propagation approach. Hydrology and Earth System Sciences, 14:1153-1165.

• http://geomorphometry.org/content/baranja-hill

Examples

library(sp)
library(gstat)
# sampled elevations:
data(barxyz)
prj = "+proj=tmerc +lat_0=0 +lon_0=18 +k=0.9999 +x_0=6500000 +y_0=0 +ellps=bessel +units=m 
+towgs84=550.499,164.116,475.142,5.80967,2.07902,-11.62386,0.99999445824"
coordinates(barxyz) <- ~x+y
proj4string(barxyz) <- CRS(prj)
# grids:
data(bargrid)
data(barstr)
coordinates(bargrid) <- ~x+y
gridded(bargrid) <- TRUE
proj4string(bargrid) <- barxyz@proj4string
bargrid@grid
## Not run: # Example with simulated streams:
data(R_pal)
pnt = list("sp.points", barxyz, col="black", pch="+")
spplot(bargrid[1], sp.layout=pnt, col.regions = R_pal[["blue_grey_red"]])
# Deriving stream networks using geostatistical simulations:
Z.ovgm <- vgm(psill=1831, model="Mat", range=1051, nugget=0, kappa=1.2)
vt.gt <- gstat(id=c("Z"), formula=Z~1, data=barxyz, model=Z.ovgm, nmax=30)
# geostatistical simulations:
N.sim <- 10
DEM.sim <- predict.gstat(vt.gt, bargrid, nsim=N.sim, debug.level=-1)
fullgrid(DEM.sim) <- TRUE
stream.list <- list(rep(NA, N.sim))
# derive stream networks in SAGA GIS 2.0.8:
for (i in 1:N.sim) {
  writeGDAL(DEM.sim[i], paste("DEM", i, ".sdat", sep=""), "SAGA", mvFlag = -99999)
  # filter the spurious sinks:
  rsaga.fill.sinks(in.dem=paste("DEM", i, ".sgrd", sep=""), out.dem="DEMflt.sgrd")
  # extract the channel network SAGA GIS 2.0.6:
  rsaga.geoprocessor(lib="ta_channels", module=0, param=list(ELEVATION="DEMflt.sgrd", 
    CHNLNTWRK=paste("channels", i, ".sgrd", sep=""), CHNLROUTE="channel_route.sgrd", 
    SHAPES="channels.shp", INIT_GRID="DEMflt.sgrd", DIV_CELLS=3, MINLEN=40), 
    show.output.on.console=FALSE)
  stream.list[[i]] <- readOGR("channels.shp", "channels", verbose=FALSE)
  proj4string(stream.list[[i]]) <- barxyz@proj4string
}
# plot all derived streams at top of each other:
stream.plot <- as.list(rep(NA, N.sim))
for(i in 1:N.sim){
  streams.plot[[i]] <- list("sp.lines", stream.list[[i]])
}
spplot(DEM.sim[1], col.regions=grey(seq(0.4,1,0.025)), scales=list(draw=T), 
sp.layout=streams.plot)

## End(Not run)