Time series visualization

Function temporalPlot is a wrapper for the lattice (trellis) plot methods in lattice::xyplot. Therefore, many different aspects of the plot can be controlled passing the relevant arguments to xyplot. Fine control of graphical parameters for the trellis display can be also controlled using trellis.par.set from package lattice.

Additionally, includes its own parameters:

• aggr.spatial: Controls the function (and related arguments) to perform spatial aggregation.
• cols: Controls the colors of the lines in each series.
• lwd: Controls the width of the lines of the series.
• lty: Controls the type of the lines of the series.
• show.na: Paints with gray the background space corresponding to missing values.

Next we compare two time series, the first corresponds to a multimember grid (seasonal forecast, CFS_Iberia_pr) and the second to a grid without members (observational data, EOBS_Iberia_pr). Before plotting we apply function subsetGrid from package transformeR to choose the desired time interval.

library(transformeR)
data("CFS_Iberia_tas")
data("EOBS_Iberia_tas")

# Combine grids with members (CFS) and without members (EOBS)
a <- subsetGrid(CFS_Iberia_tas, years = 1985:1992)
b <- subsetGrid(EOBS_Iberia_tas, years = 1985:1992)
temporalPlot("EOBS" = b, "CFS" = a)

Note that in ``CFS'' the multimember spread (range) is indicated with a shadow of the same color. Also note that, since this datasets contain only winter data, there is a blank space for the rest of the stations.

We can introduce, and thus compare, as many grids (or station data) as we want in a single call. In addition, different time intervals can also be combined. For instance, next we add the time series of VALUE station data (VALUE_Iberia_pr), but considering a narrower time interval. In this case, we compute the spatial aggregation using function min. We also use parameter xyplot.custom to add several xyplot parameters.

data("VALUE_Iberia_tas")
v <- subsetGrid(VALUE_Iberia_tas, years = 1988:1990)
temporalPlot("EOBS" = b, "CFS" = a, "VALUE" = v, lwd = 0.9,
              aggr.spatial = list(FUN = min, na.rm = TRUE),
              xyplot.custom = list(main = "winter temperature",
                                  ylab = "Celsius", ylim = c(-20, 10)))

We can use subsetGrid to obtain and plot a single location (no spatial aggregation):

 a1 <- subsetGrid(a, lonLim = 2, latLim = 42)
 b1 <- subsetGrid(b, lonLim = 2, latLim = 42)

In the following example, we change the default colors of the series with parameter cols. Option show.na = TRUE (see above) is also used:

 temporalPlot("EOBS" = b1, "CFS" = a1, 
             cols = c("green", "deeppink"), show.na = TRUE,
            xyplot.custom = list(main = "winter temperature", ylab = "Celsius"))