bubblePlot: Bubble plot for visualization of forecast skill of seasonal...

In SantanderMetGroup/visualizeR: Visualizing climate4R data and Communicating Uncertainty in Seasonal Climate Prediction

Bubble plot for visualization of forecast skill of seasonal climate predictions.

Description

Bubble plot for visualization of forecast skill of seasonal climate predictions. It provides a spatially-explicit representation of the skill, resolution and reliability of a probabilistic predictive system in a single map. This function is prepared to plot the data sets loaded from the ECOMS User Data Gateway (ECOMS-UDG). See the loadeR.ECOMS R package for more details (http://meteo.unican.es/trac/wiki/udg/ecoms/RPackage).

Usage

bubblePlot(
  hindcast,
  obs,
  forecast = NULL,
  year.target = NULL,
  detrend = FALSE,
  score = TRUE,
  size.as.probability = TRUE,
  bubble.size = 1,
  score.range = NULL,
  piechart = FALSE,
  subtitle = NULL,
  t.colors = NULL,
  pie.border = NULL,
  pch.neg.score = NULL,
  pch.obs.constant = NULL,
  pch.data.nan = NULL
)

Arguments

hindcast	A multi-member list with the hindcast for verification. See details.
obs	List with the benchmarking observations for forecast verification.
forecast	A multi-member list with the forecasts. Default is NULL.
year.target	Year within the hindcast period considered as forecast. Default is NULL.
detrend	Logical indicating if the data should be linear detrended. Default is FALSE.
score	Logical indicating if the relative operating characteristic skill score (ROCSS) should be included. See details. Default is TRUE.
size.as.probability	Logical indicating if the tercile probabilities (magnitude proportional to bubble radius) are drawn in the plot. See details. Default is TRUE.
bubble.size	Number for the bubble or pie size. bubble.size=1 by default.
score.range	A vector of length two used to rescale the transparency of the bubbles. For instance, a score.range = c(0.5, 0.8) will turn ROCSS values below 0.5 completely transparent, while values of 0.8 or will have minimum transparency (i.e., opaque). The default to NULL, that will set a transparency range between 0 and 1.
piechart	Logical flag indicating if pie charts should be plot instead of bubbles. Default is FALSE.
subtitle	String to include a subtitle bellow the title. Default is NULL.
t.colors	Three element vector representing the colors for the below, normal and above categories. Default is t.colors=c("blue", "gold", "red")
pie.border	Color for the pie border. Default is pie.border="gray"
pch.neg.score	pch value to highlight the negative score values. Default is NULL. Not available for piecharts.
pch.obs.constant	pch value to highlight those whose score cannot be computed due to constant obs conditions (e.g. always dry). Default is NULL.
pch.data.nan	pch value to highlight those whose score cannot be computed due to time series with all NA values in the observations and/or models. If score=F, highlight those grids with all forecasts NaN.

Details

First daily data are averaged to obtain a single seasonal value. The corresponding terciles for each ensemble member are then computed for the hindcast period. Thus, each particular grid point, member and season, are categorized into three categories (above, between or below), according to their respective climatological terciles. Then, a probabilistic forecast is computed year by year by considering the number of members falling within each category. For instance, probabilities below 1/3 are very low, indicating that a minority of the members falls in the tercile. Conversely, probabilities above 2/3 indicate a high level of member agreement (more than 66% of members falling in the same tercile). Probabilities are also computed for the forecast or the selected year. Color represents the tercile with the highest probability for the forecast or selected year. The bubble size indicates the probability of that tercile. This option is not plotted if the size.as.probability argument is FALSE.

Finally, the ROC Skill Score (ROCSS) is computed for the hindcast period. For each tercile, it provides a quantitative measure of the forecast skill, and it is commonly used to evaluate the performance of probabilistic systems (Joliffe and Stephenson 2003). The value of this score ranges from 1 (perfect forecast system) to -1 (perfectly bad forecast system). A value zero indicates no skill compared with a random prediction. The transparency of the bubble is associated to the ROCSS. By default only positive values are plotted if the score argument is TRUE. The target year is considered as forecast and it is not included in the computation of the score (operational point of view).

Note

The computation of climatological terciles requires a representative period to obtain meaningful results.

Author(s)

M.D. Frias [e-mail] and J. Fernandez based on the original diagram conceived by Slingsby et al (2009).

References

Jolliffe, I. T. and Stephenson, D. B. 2003. Forecast Verification: A Practitioner's Guide in Atmospheric Science, Wiley, NY.

Slingsby A., Lowe R., Dykes J., Stephenson D. B., Wood J. and Jupp T. E. 2009. A pilot study for the collaborative development of new ways of visualising seasonal climate forecasts. Proc. 17th Annu. Conf. of GIS Research UK, Durham, UK, 1-3 April 2009.

See Also

Other visualization functions: cascadePlot(), climagram(), reliabilityCategories(), spreadPlot(), tercileBarplot(), tercilePlot()

Examples

## Not run: 
data(tas.cfs)
data(tas.cfs.operative.2016)
data(tas.ncep)
require(transformeR)
# Select spatial domain
tas.ncep2 <- subsetGrid(tas.ncep, lonLim = c(-80, -35), latLim = c(-12, 12))
tas.cfs2 <- subsetGrid(tas.cfs, lonLim = c(-80, -35), latLim = c(-12, 12))
tas.cfs.operative2.2016 <- subsetGrid(tas.cfs.operative.2016, 
                           lonLim = c(-80, -35), latLim = c(-12, 12))
# Interpolate
tas.ncep2.int <- interpGrid(tas.ncep2, getGrid(tas.cfs2))
# Bubble plot. Only colour of the bubble is plotted indicating the most likely tercile 
bubblePlot(hindcast = tas.cfs2, obs = tas.ncep2.int, forecast = tas.cfs.operative2.2016,
           bubble.size = 1.5, size.as.probability = FALSE, score = FALSE)
# Bubble plot. Added size of the bubble indicating the probability of the most likely tercile 
bubblePlot(hindcast = tas.cfs2, obs = tas.ncep2.int, forecast = tas.cfs.operative2.2016,
           bubble.size = 1.5, score = FALSE)
# Bubble plot. Added transparency of the bubble indicating the ROC skill score (ROCSS)
bubblePlot(hindcast = tas.cfs2, obs = tas.ncep2.int, forecast = tas.cfs.operative2.2016,
           bubble.size = 1.5)
# 3-piece pie chart.
bubblePlot(hindcast = tas.cfs2, obs = tas.ncep2.int, forecast = tas.cfs.operative2.2016,
           bubble.size = 1, piechart = TRUE)

## End(Not run) 

SantanderMetGroup/visualizeR documentation built on Oct. 28, 2023, 6:11 a.m.