Area-weighted upstream means of several sampled layers

The multi-layer form of wet_upstream_mean() with denom = "covered": for watershed a and each column x of values, sum(x_u * area_u * cover_u) / sum(area_u * cover_u) over the FWA_Upstream set of a, in one pass of wet_upstream_sums(). All columns share one cover (as wet_ws_sample() returns for several layers), so the upstream mean of a linear combination of layers is exactly that combination of their upstream means.

Usage

wet_upstream_means(
  ws,
  values,
  cols = NULL,
  irregular_pairs = NULL,
  extra = NULL
)

Arguments

ws

data.frame(watershed_feature_id, wscode, localcode, area_m2) for a whole basin, e.g. from wet_ws_fetch().

values

data.frame(watershed_feature_id, <cols>, cover). Polygons absent from it count as uncovered. A polygon with cover > 0 must have a value in every column.

cols

Columns of values to average. Default: all but watershed_feature_id and cover.

irregular_pairs

Passed to wet_upstream_sums().

extra

Optional columns of ws to average over the whole upstream area (not only the covered part), e.g. an in-BC indicator.

Value

data.frame(watershed_feature_id, <cols>, <extra>, coverage, upstream_area_m2), one row per row of ws.