Multi-condition Plots

[gregjohnsonbiometrics]

We need to understand how to best handle FIA plots with multiple conditions. The following three graphs from the diameter growth subsets illustrate the reason for concern. HT40 appears to be different for plots with multiple conditions (tends to be higher). This hints that plot-level statistics like HT40 (e.g., BA, CCF,TPA) may not be appropriate for the combined sub-plots.

[gregjohnsonbiometrics]

After some thought, I think this moves us to compute per-area statistics at the sub-plot level. That way we can drop non-forest plots easily. It will likely introduce some imprecision¹. When re-processing occurs we should decide whether we want to go that direction.

What do y'all (@holoros, @DMtree-DougFir) think?

Footnotes

1. Hann, David. (1991). Growth Model Predictions as Affected by Alternative Sampling-Unit Designs. Forest Science. 37. 1641-1655. 10.1093/forestscience/37.6.1641. ↩

[holoros]

Sorry, I didn't think of this before. But I generally do all my work with FIA data at the sub-plot. I think it is more consistent and gives better results, despite the potential "smaller" plot effects. Fortunately, an easy fix in the R code for data generation.

[gregjohnsonbiometrics]

from @DMtree-DougFir: _

GrowthModelingAndFIAPlotConditions_20260201.pdf

_