Placing height measurements in perspective   John Eichholz
  Apr 29, 2004 21:37 PDT 

Bob, Howard:

material deleted....
Using several sets of instruments can improve our prediction of tree
height. When we want to know the absolute height of a specific tall
tree, we can use many methods to cross reference and hone in on the
height. However, I believe the evidence exists that the height
measurement has a natural variation due to rangefinder variability and
human factors.

Some of the large scale sampling we will be doing at MTSF will likely be
simplified by classing trees into 5' height ranges anyway. We may be
measuring large numbers or all of the trees in a limited area of high
growth. This will readily tell us the maximum height and cbh, and the
average. Grouping trees into height classes allows us to also know the
modal or most common height. We locate the 5' height range that has the
most trees, comparing that to the top height attained and the high to
low range. We don't know what kind of distributions we will find,
(except they will be unusually tall). Comparing old stands of trees to
young ones nearby will happen first, to see how trees evolve in height
and mass. Later we can compare stand to stand, species to species and
correlate that with differences in site. Then we might be ready to
compare with sites outside MTSF. Or, maybe we could make a Rucker index
of the modal values.

Looking at the length of some of the bibliographies that appear on this
list you would think this has been done before. That may be true, but I
think it is safe to say it has not been done at MTSF until now!

John
RE: Placing height measurements in perspective   John Eichholz
  Apr 30, 2004 06:10 PDT 
Ed:

Yes, and I am particularly interested in the relationship between the
modal value and the maximum. The traditional Rucker index , relying on
the single tallest individual of 10 species, is quick and easy to
determine, relative to the modal index. The modal index I think might
represent more accurately the volume potential of a site. It would
reflect the average canopy height, not its ceiling. How well would the
distribution of tree heights reflect the age diversity of the site?
Also, how would modal values/ height distributions by species vary with
the prevalence of that species on the site?

If I am interpreting correctly, the Reineke stand density index (SDI)
draws conclusions about diameter distributions, saying that basal area
of any site can be converted to a basal area/tree count at 10" diameter,
enabling stocking rates of sites of different ages to be compared. Has
this been debunked? Can we do something similar with heights? Will
sites with a high SDI also have high heights? Will sites with a high
"height index" continue volume growth for a longer time than shorter
sites? How about sites with a high SDI?

I am probably way over my head when it comes to traditional forest
mensuration methods. On the other hand, it makes sense to me that we
have a recent technological advantage in being able to quickly and
accurately measure heights, especially in hardwoods. Are any of the
foresters out there aware of height measures similar to the SDI?

John
RE: Placing height measurements in perspective   Robert Leverett
  Apr 30, 2004 10:13 PDT 

John:

With respect to existing indices, I expect that the only way we're
going to find out what works and what doesn't for the natural stands of
100 years in age and older is to gather the data ourselves and then
develop a variety of statistical distributions, even if we stumble a few
times in the process. We've got Lee Frelich on board to keep us going in
the right direction. He won't let us stray too far from a scientifically
defensible direction.

Maybe there are stand-based growth relationships that could be dug out
of books and applied directly to the habitats and species mixes we're
studying. However, the rates, averages, and indexes developed for timber
management purposes have built in biases that make them of questionable
applicability for our purposes - if we're looking to achieve high levels
of accuracy. I'm reminded of what Karl Davies discovered about the
development of volume-based growth models from a woefully inadequate
database. Inadequate or not, the FS routinely applies it.

     Several things we may assume about the silvicultural data reflected
in the tables that wood producers use that were developed using data
from the past.

    1. Much of the data were gathered in relatively even-aged stands
and/or in highly controlled environments. It is unclear how much data
have been gathered from the kinds of places we study, e.g. extensive
boulder fields with forests of all ages.

    2. Whole tree volume was unimportant.

    3. Where used, whole tree height determinations using clinometer and
baseline for volume determinations would have been off, sometimes by
significant amounts, but an average of 5% to 7% is probable.

    4. Volumes for logs based on standard lengths and diameters assumed
circular cross-sections. I think that Bob Van Pelt has discussed in the
past the size errors that can result from invalid assumptions of
circularity. I think some wood producers attempt to get around the
problem by using calipers to try to arrive at an average diameter.
However, this method still has problems. As an example, if the
cross-sectional area of a tree is elliptical with the major axis being
25% greater than the minor axis, then averaging the lengths of the two
axes and computing the area of the circle based on the resulting average
yields a circle with an area about 1.2% greater than the area of the
ellipse. However, as we have also seen, assumptions of ellipticity are
risky. My guess is that deviations from circularity generally produce
area errors of 3% or less. Maybe someone out there has a feel for this.
I expect BVP and Lee Frelich do.

In terms of our own volume calculations, I'm unsure if we need to
worry about departure from circularity, unless visual inspection of a
tree suggests radical departure. At the least, we should get several
sets of calipers.


Bob