Trend Test
trend_test.Rd
Test for period of record and user-specified ranges. The default trends are calculated for 10 year and the full period of record.
Usage
trend_test(
gw_level_dv,
gwl_data,
n_years = 10,
parameter_cd = NA,
date_col = NA,
value_col = NA,
approved_col = NA,
stat_cd = NA,
pctComplete = 0.5,
days_required_per_month = 14,
POR_trend = TRUE
)
Arguments
- gw_level_dv
daily groundwater level data frame. Often obtained from from
readNWISdv
- gwl_data
data frame returned from dataRetrieval::readNWISgwl, or data frame with mandatory columns lev_dt (representing date), lev_age_cd (representing approval code), and a column representing the measured value (either lev_va, sl_lev_va, or value).
- n_years
integer. This is the number of years to calculate the trend on. Default is 10. This can be a vector of years.
- parameter_cd
If data in gw_level_dv comes from NWIS, the parameter_cd can be used to define the value_col. If the data doesn't come directly from NWIS services, this can be set to
NA
,and this argument will be ignored.- date_col
the heading of the date column. The default is
NA
, which the code will try to get the column name automatically.- value_col
name of value column. The default is
NA
, which the code will try to get the column name automatically.- approved_col
name of column to get provisional/approved status.
- stat_cd
If data in gw_level_dv comes from NWIS, the stat_cd can be used to help define the value_col.
- pctComplete
number percentage complete. This is a fraction that represents the amount of data that must be included overall in order to calculate a trend. The default is 0.5, which means if gaps in the data span more than 50 total record, a trend will not be calculated.
- days_required_per_month
integer. Number of days required per month to include in the trend test. Default is 14.
- POR_trend
a logical indicating whether to include a trend test for the full period of record. Default is
TRUE
.
Details
For data that is at least on a daily interval, the rkt function is used. For periodic data, the kendallTrendTest is used.
Examples
# site <- "263819081585801"
# gw_level_data <- dataRetrieval::readNWISgwl(site)
# Using package example data:
gwl_data <- L2701_example_data$Discrete
gw_level_dv <- L2701_example_data$Daily
trend_test(gw_level_dv,
gwl_data,
parameter_cd = "62610")
#> test tau pValue slope intercept trend
#> 1 10-year trend 0.6144578 0 1.4755392 -2998.4482 Up
#> 2 Period of record -0.3191964 0 -0.2836721 545.4421 Down
trend_test(gw_level_dv,
gwl_data,
POR_trend = FALSE,
parameter_cd = "62610")
#> test tau pValue slope intercept trend
#> 1 10-year trend 0.6144578 0 1.475539 -2998.448 Up
trend_test(gw_level_dv,
gwl_data,
parameter_cd = "62610",
n_years = 5)
#> test tau pValue slope intercept trend
#> 1 5-year trend -0.01333333 0.9514722 NA NA Not significant
#> 2 Period of record -0.31919643 0.0000000 -0.2836721 545.4421 Down
trend_test(gw_level_dv,
gwl_data,
parameter_cd = "62610",
n_years = c(5, 10, 20))
#> test tau pValue slope intercept
#> 1 5-year trend -0.01333333 9.514722e-01 NA NA
#> 2 10-year trend 0.61445783 0.000000e+00 1.4755392 -2998.4482
#> 3 20-year trend 0.34461806 1.807443e-13 0.6532362 -1340.4724
#> 4 Period of record -0.31919643 0.000000e+00 -0.2836721 545.4421
#> trend
#> 1 Not significant
#> 2 Up
#> 3 Up
#> 4 Down
# Only periodic data:
trend_test(NULL,
gwl_data,
parameter_cd = "62610")
#> test tau pValue slope intercept trend
#> 1 10-year trend 0.5182648 9.079091e-11 1.5966253 -3243.1985 Up
#> 2 Period of record -0.1628813 1.078165e-05 -0.2200138 414.9972 Down