What is Density-Weighted Proportion (DWP) and why does it matter for fatality estimates?

Density-Weighted Proportion (DWP) is the fraction of a turbine’s carcass fall zone that is actually walked during a search, weighted by where carcasses are most likely to land. It exists because real fall zones are larger and more irregular than the search plot. Carcasses land in dense vegetation outside the boundary, on inaccessible terrain, behind hazards, and in areas the search team is not allowed to enter. Without a DWP correction, every one of those carcasses is invisible to the fatality estimate. The IFC PCFM Handbook is direct that this correction is non-optional:

To produce unbiased fatality estimates, it is important to account for carcasses that land in these unsearched and unsearchable areas of the fall zone, especially if such areas cover a large proportion of the fall zone (e.g., in RAP search plots).

— IFC PCFM Handbook, §3.3.7.1

Why DWP is more than fraction-of-area-searched

A naïve correction would just take the searched area divided by the fall-zone area. That is wrong, because carcass fall density is not uniform inside the fall zone. Bat carcasses cluster close to the tower; large birds spread further out. A search plot covering the inner 80 m may catch most bats but a much smaller share of large birds. DWP weights the area correction by where carcasses actually land per size class, so the correction reflects the carcasses you have any chance of finding rather than just the geometry on the map.

Why this is the piece programs outsource

Calculating DWP correctly requires spatial analysis (carcass fall density, plot geometry, terrain, vegetation, size class), and the IFC PCFM Handbook acknowledges directly that this is the bottleneck most programs run into:

Wind wildlife ecologists in emerging market countries frequently mention the lack of a simple to understand method for estimating the DWP as a principal barrier to estimating fatality rates at WEFs.

— IFC PCFM Handbook, Box 3.4

In practice this is where most PCFM programs hit the wall at year-end. Search and trial data is sitting in folders, but the DWP file is not. A contractor is hired, scope is negotiated, the GenEst submission slips by weeks, and the spatial correction that determines the entire fatality estimate is built once, by someone outside the program, against geometry someone else owns.