Skip to contents

Cell values for each fundamental watershed

Source: R/wet_ws_sample.R
wet_ws_sample.Rd

Two methods:

Usage

wet_ws_sample(r, ws, method = c("centroid", "area"))

Arguments

r

terra::SpatRaster, one or more layers.

ws

terra::SpatVector of polygons with a watershed_feature_id attribute; or, for "centroid" only, a data.frame(watershed_feature_id, lon, lat) of centroids already computed (e.g. by wet_ws_fetch()), which avoids transferring geometry for a whole basin.

method

"centroid" or "area".

Value

For one layer, data.frame(watershed_feature_id, value, cover). For several, data.frame(watershed_feature_id, <layer names>, cover).

Details

  • "centroid": the value of the cell containing each polygon's centroid (centroid taken in the polygons' own CRS, then reprojected). This is what fwapg does with ST_Value(), and the method to use for a parity check. cover is 1 where the value is present and 0 where it is NA.

  • "area": the mean of every cell the polygon overlaps, weighted by the overlapping fraction, over cells that have a value. cover is the share of the polygon's overlap that falls on cells with a value, so a polygon half off the model domain reports cover = 0.5 rather than a silently halved value downstream.

With several layers, a cell counts only where every layer has a value, and all layers are averaged over those same cells. There is then one cover per polygon, so upstream means of every layer share one denominator and a linear combination of layers is exactly the same combination of their means.