Skip to contents

Find maximum breakpoint position based on low flow threshold(s)

Source: R/find_nmax.R
find_nmax.Rd

In order to constrain breakpoint analysis on DFI curves (CDCs = Characteristic Delay Curves) the length of 1:N delayed flow separations should be as short as possible. Normally the position N is searched where the CDCs is flattening out (slope = 0). To do this the DFI values of 1:N separations are compared against typical low flow indices (like the MAM/MQ). The low flow indices ranging between 0 and 1 indicating the low flow sensitivity. High index values (e.g. 0.60) indicating more stable flow regimes with larger catchment storages, lower index values (e.g. 0.20) indicating rather flashy flow regimes with faster recession behavior. The maximum breakpoint position is estimated by estimating the DFI value with a separation block size of N that match the index value.

Usage

find_nmax(df, n = 1:180, desc = TRUE, lowflow_index = c("mam_mq"))

Arguments

df

data.frame with two columns, first column must be Date, second must be streamflow data.

n

numeric vector

desc

logical, if TRUE DFI values are converted to be monotonically decreasing with cummin()

lowflow_index

Which Low Flow Index should be used? MAM/MQ (default), Q95/Q50, Q90/Q50. Index names are "mam_mq", "q95_q50" or "q90_q50". Note that especially in highly seasonal regimes the deviation between the indices might be large(r).

Value

Function returns a list with two elements: $index_value and $bp_nmax

index_value

Index value of the used low flow index ranging between 0 and 1. Higher values indicating more stable flow regimes and less low flow sensitivity.

bp_nmax

Position N of breakpoint where absolute deviation between DFI_N and low flow index is minimal (considering all DFI values), min(which.min(abs(cdc$dfi-index)))

Examples

find_nmax(q_data)
#> $index_value
#> [1] 0.1126447
#> 
#> $bp_nmax
#> [1] 65
#>