Find pseudoreplicate/duplicate data in Water Tracker output

Find pseudoreplicate/duplicate data in Water Tracker output

Usage

find_duplicates(df)

Arguments

df

Input data frame from water tracker

Value

tibble with added logical field "dup"

Details

Internal to format_watertracker(). Not meant to be called separately.

An observation is considered to be a likely duplicate of a previous observation (i.e. representing the same individual scene) if: (1) the date range of the mosaic starts on or before previous mosaic's end date, (2) the PercentWater and ObservedAreaHa stats match, and (3) PercentWater is either NA or greater than zero (because difficult to distinguish multiple dry images from duplicates).