|
|
|
 |
||||||||||||||||||||||||||||||||||||
  |
|||||||||||||||||||||||||||||||||||||
ACCURATE INVENTORY USING MARVL
MARVL Overview - As seen and used by Interpine Forestry - July 2001
MARVL was designed as
a system to more precisely define the quality and value of forests in terms
of their potential to produce volumes of logs of predefined grades.
A further important design feature is its flexibility. Out-turn
can be estimated not just in terms of a single set of log grades but a whole
range of grades (e.g. either 'export' or 'domestic' log grades).
MARVL is an acronym for "Method
of Assessing Recoverable Volume by Logtype",
and is a 'variable sections' individual
tree quality assessment.
This method employs no preconceived notion of log length and it does not amalgamate
stem
parts
of varying stem quality in arbitrarily determined fixed log length sections.
Each section defined is of variable log length, but contains
only one quality class (having an associated code). At the
point where there is a change in stem quality class, one section is terminated
and another is begun. Although there is a minimum section length, the number
of sections in one tree can be anything from one to 10 or more.
Want to find out more ? click on the links below.....
MARVL is a pre-harvest
inventory system used to assess the potential recoverable volume by log grade
and value or return to the acre / hectare when a stand of trees is harvested.
The single most important feature of
MARVL's design is the independence of cruise data from cutting strategies.
This makes it possible
to reanalyze one set of cruise data many times using different cutting strategies
to determine the sensitivity of product out turn to changes in log specifications
and market demand.
The variable sections method is the most definitive, precise, and flexible
approach to tree quality assessment as opposed to the alternative of 'constant
sections' method using standard log lengths.
The concept of the cutting strategy being independent of tree quality data
is the key to understanding the way MARVL works.
Using pre-harvest inventory cruise data, MARVL allows timberland managers and marketers to accurately predict product mixes and maximize value recovery from a standing resource.
"If
you know what is standing in the resource then you can efficiently align your
markets, customers and processing facilities accordingly"
MARVL
Inventories can be used in a range of forest management situations:
Utilization
Planning - Knowledge of the resource indicates
what sort of processing facilities should be built (e.g. plywood mill, sawmill,
fiber production). This application is most appropriate to long-range planning
of a large scale, relatively mature, and uncommitted resource (an increasingly
rare phenomenon).
Supply
Chain Planning - To meet the short and medium
term requirements for wood supply to existing customers, wood-flow planners
need to know where to go for a required log mix, the kinds of equipment required,
and the anticipated harvesting costs.
Value Recovery
- MARVL indicates the optimal cutting strategies and predicts maximum out-turn
by volume and value. Comparisons between MARVL predictions and actual out-turn
can be used to monitor the efficiency of logging operations and market constraints.
Cutting
Strategies - MARVL can simulate the financial
consequences of various log-bucking strategies, thus allowing the optimal
strategy to be selected.
Marketing
- Sellers can offer different product combinations to prospective buyers based
on different cutting strategies.
Valuation
- MARVL can be used to calculate the highest liquidation value (i.e. return/acre)
for a stand, or in conjunction with GroMARVL, the highest expectation value
(discounted future nett revenue).
In the above
situations it can be seen that MARVL functions
variously as a pre-harvest inventory
method (utilization and harvest planning),
a decision support tool (bucking
strategies and marketing) and a mid-rotation
inventory system (expectation values).
Assessing Stands and
Forests
Before collecting any information on individual trees the sampling methods
are determined based on the type of resource and the desired results and accuracy.
There are a number of steps required to plan out the inventory.
MARVL supports a range of sampling systems. Sampling systems are
available for systematic sampling, simple random sampling, stratified sampling
and double sampling. Plot types include bounded, angle gauge, and horizontal
line.
The wide variety of plot types
and sampling methods supported by MARVL allows the user to decide the best
strategy to assess the stand or forest in question, given a level of accuracy
or limiting budget.
For example, the use of double sampling provides extra plots for assessing
basal area using either a bounded plot or a faster unbounded angle gauge plot
system. The extra plots are collected between the normal bounded
plots and take half the time. The forester must assess if the
increase in precision in the results from these secondary plots is worth the
money and time it takes to collect the extra information.
Normally a pilot survey of plots is collected to determine the required number
of plots needed for a given level of accuracy. After undertaking the pilot
survey the forester can then adjust or modify the strategy to suit the requirements
of the inventory.
Assessing
Tree Quality
In order to estimate the log products from standing trees it is necessary
to make a visual assessment of the quality of the stem progressively from
bottom to top. This requires setting up a quality code dictionary. The dictionary
is a range of codes used to describe the stem characteristics in terms of
qualities such as branch size, straightness, and roundness. Trees are coded
into sections according to changes in these quality codes.
"The main advantage of systems such as MARVL over other inventory methods is the ability to separate out the tree qualities from the limiting log qualities"
MARVL dictionaries
are able to be uniquely defined for each forest management area or inventory.
Once determined, dictionaries are usually not altered unless it is necessary
to improve the stem description to cope with changing log quality specifications.
Building
a new dictionary
Although stem qualities and log types (dictionary
and cutting strategy) are considered independent of one another there is a
need to consider the range of log types when first defining a new dictionary.
Log types are normally defined according to written specification sheets,
which are highly detailed. Typically the main stem qualities that are assessed
using MARVL are branch size, internode length and sweep.
The quality code associated with each stem section is selected from an alphabetical
list of codes. The codes are in terms of quality characteristics such as sweep,
forking, bark damage, rot, branch size, etc, and not in terms of log dimensions
such as log length, small end diameter, and large end diameter. Initially
it is best to create a dictionary, which is as detailed as possible, without
compromising ease of use in the field.
Reanalyzing
data with the same dictionary
Once the cruise data has been collected then it can be easily reanalyzed using
different grade spec tolerances with cutting strategies that reflect these
changes. Hence if grade specifications are tightened then the full consequences
on grade out-turn can be quantified.
Assessing
Tree Structure
As well as describing tree qualities such as branching and sweep it is also
necessary to describe the structure of the stem, such as forking or dead tops.
Cruising
Trees
At the fieldwork stage trees are directly measured for diameter at breast
height (DBH) and height. The stem is then visually assessed and classified
according to a set of quality and structural codes. A sample
is taken over a given forest area using traditional bounded, variable and
count plot methods, where trees within the plot are assessed for quality as
described above.
Data
Analysis and Processing - the MARVL Program
Once data from the variable sections quality assessment (forest inventory)
has been collected, it will be processed using the MARVL computer program
within the friendly WindowsOS.
The program essentially creates a mathematical
image of each plot tree based on the forest inventory data.
This involves subjecting trees measured for height (and diameter) to regression
analysis to create a diameter/height relationship. This relationship (Petterson
curve) is then used to allocate a height to each of the plot trees without
height measurements. The standing volume of each plot tree is then calculated
from the measured diameter and estimated or measured height using individual
tree volume and taper equations (these can be inserted into the program or
adjusted by program users). The volume, and taper equations within the program
are then used to produce a profile of the individual trees.
Following the creation of tree profiles,
there are two simulations undertaken. The
first is a felling simulation, incorporating a breakage equation that is added
by the user. This determines the point at which trees will break
during felling. The second simulation bucks the stems
into allowable log types. This is done using a cutting instruction
(strategy) applied to find the economically optimum way of making stems (between
stump and break point) into logs - taking account of stem quality codes, small
and large end diameter constrains, and relative log values set by the cutting
instruction.
Data
outputs and reporting
The MARVL output function / reporting system provides a wealth of information.
The reporting function can be adjusted to provide outputs
in database, worksheet or readable form, and reports can be constructed by
the user to suit unique requirements.
All data collected is stored within easily accessible database tables that
can be reported on using any database reporting programs such as Crystal Reports
or FoxPro. Simple raw inventory data reporting capabilities are also available
within the MARVL report function.
"The power of the MARVL program is evident
during the data analysis stage. Cutting instructions can be altered any number
of times and plot data reanalyzed against the different strategies enabling
sensitivity analysis of different bucking and marketing decisions. The value
of any number of scenarios and options can be calculated and reported on"
MARVL is a licensed product of New Zealand Forest Research Institute (NZFR). This overview is an introductory outline on behalf of Interpine Forestry Ltd and may not reflect directly on the views of NZFR.
Further information can be gathered directly from NZFR by following this link*
* due to the nature of the Internet we can not guarantee links outside of our website - we will however attempt to ensure they are kept up to date. You can notify us of link problems via our feedback form.
