tsMake: Create time series from water quality data

Description

Creates a matrix time series object from an object of class "WqData", either all variables for a single site or all sites for a single variable.

Usage

"tsMake"(object, focus, layer, type = c("ts.mon", "zoo"), qprob = NULL)

Arguments

object

Object of class "WqData".

focus

Name of a site or water quality variable.

layer

Number specifying a single depth; a numeric vector of length 2 specifying top and bottom depths of layer; a list specifying multiple depths and/or layers; or just the string "max.depths".

type

ts.mon to get a monthly time series, zoo to get an object of class "zoo" with individual observation dates.

qprob

quantile probability, a number between 0 and 1.

Value

A matrix of class "mts" or "zoo".

Details

When qprob = NULL, the function averages all included depths for each day, the implicit assumption being that the layer is well-mixed and/or the samples are evenly distributed with depth in the layer. If layer = "max.depths", then only the value at the maximum depth for each time, site and variable combination will be used. If no layer is specified, all depths will be used.

The function produces a matrix time series of all variables for the specified site or all sites for the specified variable. If type = "ts.mon", available daily data are averaged to produce a monthly time series, from which a quarterly or annual series can be created if needed. If you want values for the actual dates of observation, then set type = "zoo".

When qprob is a number from 0 to 1, it is interpreted as a probability and the corresponding quantile is used to aggregate observations within the specified layer. So to get the maximum, for example, use qprob = 1. If type = "ts.mon", the same quantile is used to aggregate all the available daily values.

Examples

Run this code

# Create new WqData object
sfb <- wqData(sfbay, c(1, 3:4), 5:12, site.order = TRUE,
         time.format = "%m/%d/%Y", type = "wide")

# Find means in the 0-10 m layer
y <- tsMake(sfb, focus = 's27', layer = c(0, 10))
plot(y, main = 'Station 27')
# Or select medians in the same layer
y1 <- tsMake(sfb, focus = 's27', layer = c(0, 10), qprob = 0.5)
plot(y1, main = 'Station 27')
# Compare means:medians
apply(y/y1, 2, mean, na.rm=TRUE)

# Combine a layer with a single additional depth
y <- tsMake(sfb, focus = 'chl', layer = list(c(0, 2), 5))
plot(y, main = 'Chlorophyll a, ug/L')

# Use values from the deepest samples
y <- tsMake(sfb, focus = 'dox', layer = "max.depths", type = 'zoo')
head(y)
plot(y, type="h", main = "'Bottom' DO, mg/L")