Forest list archive: msg00025

[Prev][Next][Index][Thread]

Woodstock (was Re: Yield prediction in reserved forests)



Dear Netters,

I have had several requests for information about Woodstock from readers
of the forest and fmdss-l lists, and Sam Radcliffe <samr@gbcwke.sol.net>
suggested that I make a general informational posting.  So, here goes ... 

(My apologies to those of you on both lists for any duplication, or those 
who have already received this as a result of a personal request.)

Woodstock was developed here at the University of New Brunswick in
cooperation with REMSOFT, a firm specializing in resource management
software (particularly fire weather and fire modeling software).

Woodstock is a forest modeling system, not a growth and yield modeling
system.  It is similar in purpose to FORPLAN, FORMAN and other tools used
for timber harvest scheduling.  Actually, Woodstock is not just limited to
timber harvest scheduling.  It is currently being used: as a vegetation
management tool in Banff National Park (British Columbia); to determine
the effect of changing prices on timber supply (i.e., economic timber
supply); and to model the impacts of climatic change.  Woodstock actually
uses some of the best ideas from a number of earlier tools and combines
them into one system.  The result is, I think, more powerful and more
intuitive than other forest modeling systems. 

The basic premise behind Woodstock is that a modeler can view a problem in
many different ways and may wish to use different analytical techniques. 
So instead of limiting the modeler to one particular analytical technique,
it provides three techniques (with a fourth one being developed).  The
modeler can build a model of the successional patterns in a forest and the
responses to interventions/treatments.  Then with a few simple changes or
additions, any of the three analytical techniques can be applied to the
same forest model.  This provides greater flexibility to address various
issues than other modeling systems, without having to build new models
from scratch every time. 

The first technique is simple inventory projection (what I call
deterministic simulation).  This technique is similar to New Brunswick's
FORMAN.  The second technique is stochastic simulation where the
interventions (eg. harvesting by private woodlot owners, fire occurrence)
or the resulting responses are random variables. The third technique is
linear programming, similar to the US Forest Services's FORPLAN.  The
fourth technique is really just an extension of linear programming - mixed
integer programming - which would allow all-or-none types of decisions
(similar to coordinated allocation choices in FORPLAN). 

Woodstock itself is the strategic planning component of the modeling
system, which means that it is stratum-based and long-term ( > 1
rotation).  Separate components are used for tactical planning of harvest
blocks in the medium term (30-40 years).  These components, called Crystal
and Block, were initially developed as Master's projects at UNB.  Crystal
is a program to aggregate stands into preliminary harvest blocks according
to a strategic harvest schedule (what some people call the spatial
disaggregation process).  Block uses Monte Carlo Integer Programming to
allocate the preliminary blocks to harvest periods in such a way that
restrictions on clearcut size and adjacency (green-up) delay are met.
These two components are now being developed into a single component
called Stanley.

Here are some references:

Walters, K.R. 1994.  Design and development of a generalized forest 
management modeling system: WOODSTOCK. pp.190-196 IN: Paredes, G.L. (ed). 
Proceedings of the International Symposium on Systems Analysis and 
Management Decisions in Forestry, March 9-12, 1993, Valdivia, Chile. 
Universidad Austral Chile.

Jamnick, M.S. and K.R. Walters. 1991. Harvest blocking, adjacency 
constraints and timber harvest volumes. IN: Proceedings of the 1991 
symposium on systems analysis in forest resources, March 3-6, 
Charleston, SC.  Compiled by: M.A. Buford. USDA For. Serv. Gen. Tech. 
Rep. SE-74. pp. 255-261.

Jamnick, M.S. and K.R. Walters. 1993. Spatial and temporal allocation of 
stratum-based harvest schedules. Can. J. For. Res. 23:402-413.

Clements, S.E., P.L. Dallain and M.S. Jamnick. 1990. An operational, 
spatially constrained harvest scheduling model. Can. J. For. Res. 20:
1438-1447.


Please feel free to contact me for more information or copies of the above
papers.  For information about acquiring a copy of Woodstock and the other
programs, contact Ugo Feunekes at REMSOFT: 

	phone: 506-450-4511
	fax:   506-459-7290
	email: remsoft@nbnet.nb.ca
	snail mail:  REMSOFT Inc.
	      	     620 George Street, Suite 5
		     Fredericton, New Brunswick
		     Canada          E3B 1K3


--
Hubert Burger                              T6LV@unb.ca
Forestry and Environmental Management
University of New Brunswick, Fredericton, N.B., Canada






[Metla] [Main Index] [Thread Index]

Mail converted by MHonArc 1.1.0