ciCalculations

Function to calculate confidence bands for flow normalized concentration or flow normalized flux.It returns the data frame CIAnnualResults, which is used as input to the functions plotConcHistBoot, and plotFluxHistBoot which produce the graphical output.

Usage

ciCalculations(eList, startSeed = 494817, verbose = TRUE,
  jitterOn = FALSE, V = 0.2, nBoot = 100, blockLength = 200,
  widthCI = 90)

Arguments

eList

named list with at least the Daily, Sample, and INFO dataframes. Created from the EGRET package, after running modelEstimation.

startSeed

setSeed value. Defaults to 494817. This is used to make repeatable output.

verbose

logical specifying whether or not to display progress message, default = TRUE

jitterOn

logical, if TRUE, adds "jitter" to the data in an attempt to avoid some numerical problems. Default = FALSE. See Details below.

V

numeric a multiplier for addition of jitter to the data, default = 0.2. See Details below.

nBoot

number of times the bootstrap resampling and model estimating is done. Default is 100, but that will take a long time. Testing should initially be done using a smaller number like 10.

blockLength

integer size of subset, expressed in days. 200 days has been found to be a good choice.

widthCI

numeric, the width of the confidence intervals. 0.9 means the confidence intervals will be calculated with 90%.

Value

CIAnnualResults a data frame with the following columns:

Year	mean decYear value for the year being reported
FNConcLow	the lower confidence limit for flow normalized concentration, in mg/L
FNConcHigh	the upper confidence limit for flow normalized concentration, in mg/L
FNFluxLow	the lower confidence limit for flow normalized flux, in kg/day
FNFluxLow	the lower confidence limit for flow normalized flux, in kg/day

Details

In some situations numerical problems are encountered in the bootstrap process, resulting in highly unreasonable spikes in the confidence intervals. The use of "jitter" can often prevent these problems, but should only be used when it is clearly needed. It adds a small amount of random "jitter" to the explanatory variables of the WRTDS model. The V parameter sets the scale of variation in the log discharge values. The standard deviation of the added jitter is V * standard deviation of Log Q. The default for V is 0.2. Larger values should generally be avoided, and smaller values may be sufficient.

Argument values suggested. To test the code nBoot = 10 is sufficient, but for meaningful results nBoot = 100 or even nBoot = 500 are more appropriate. blockLength = 200 widthCI = 90 (90% confidence interval)

Examples

library(EGRET)
eList <- Choptank_eList
if (FALSE) {
CIAnnualResults <- ciCalculations(eList,
                                  nBoot = 10)
plotConcHistBoot(eList, CIAnnualResults)

# run in batch mode, using non-stationary flow normalization
# In this example nBoot is set very small, useful for an initial trial run.
# A meaningful application would use nBoot values such as 100 or even 500. 
seriesOut_2 <- runSeries(eList, windowSide = 11)
CIAnnualResults <- ciCalculations(seriesOut_2, 
                     nBoot = 10,
                     blockLength = 200,
                     widthCI = 90)
                     
 plotConcHistBoot(seriesOut_2, CIAnnualResults)

}