Add observations to Aeme object

Usage

add_obs(aeme, lake = NULL, level = NULL, depth_ref = c("surface", "bottom"))

Arguments

aeme: Aeme object.
lake: data frame with required columns "Date", "var_aeme", "depth" and "value", and optional columns "depth_to" (bottom of an integrated sample), "sd" (measurement standard deviation, in the variable's units) and "depth_ref" (see depth_ref). The legacy "depth_from" / "depth_to" column pair is accepted and collapsed to "depth" (with a one-time deprecation warning). "Date" may be a Date, POSIXct or date/datetime string; it is stored as a UTC POSIXct with daily rows anchored at 12:00:00. If NULL, no observations are added.
level: data frame with columns "Date", "var_aeme" and "value". "Date" is handled as for lake. If NULL, no observations are added.
depth_ref: character; what "depth" in lake is measured from. "surface" (default): metres below the water surface. "bottom": metres above the lake bed, for observations recorded as a height off the bottom (a logger 1 m above the sediment). A "depth_ref" column in lake, with values "surface" or "bottom", sets this per row and takes precedence; depth_ref applies to rows without one.

Value

Aeme object with observations added

Details

Stored observations always have depth in metres below the surface, so everything that reads it keeps working. For a bottom-referenced row the height you give is kept in height_above_bed (with depth_ref = "bottom"), and depth is worked out as level - height, where level is the height of the water surface above the bed on that date: from the lake level observations (those in level, or already on the object), interpolated in time; failing that, the full depth of the basin, with a warning because it ignores how far the real level was from full. That depth is an estimate. get_var() with use_obs = TRUE and ensemble_summary() compare bottom-referenced observations with the model at height_above_bed using the model's own water level, which is what "1 m above the bed" means.