Student:	Özkan BİNGÖL
Supervisor:	Assist. Prof. Dr. Murat EKİNCİ
Department:	Bilgisayar Mühendisliği
Institution:	Graduate School of Natural and Applied Sciences
University:	Karadeniz Technical University, Turkey

Title of the Thesis:	Modeling of Forest Ecosystem and Determination of Silviculturel Prescription Alternatives Using Heuristics Techniques
Level:	M.Sc.
Acceptance Date:	20/8/2009
Number of Pages:	124
Registration Number:	i2091

Summary:

In this study, Modular based a forest ecosystem model (FEM) was developed for

applicability of Ecosystem Based Multiple Use Forest Management (EBMUFM). Unified

modeling language (UML) was used in the beginning stage of model design. Thanks to UML

that supplies standard design diagrams, FEM can be used by not only the designers but also

other users who want to add new modules or users who want to change existing modules. For

this characteristic, model has extensible structure.

Silvicultural interfere options were defined on structure of modeled forest area and

showed their effects to forest. Basal Area Rate Simulation Method (BARSM) growth model

was used in order to analyze of changes of forest structure (forest dynamics) versus time. FEM

can be used on decision stage in addition to set out data used for stage of forest management

planning by decision maker. For this purpose, Simulation and optimization modulus were

developed and set to run over FEM.

In simulation model, decision maker desire can monitorize forest dynamics by applying

operation alternatives and planning target. Initially only stand parameters were used in this

module. In next stage, spatial planning was done by integrated spatial parameters model. In

optimization model, types of operation alternatives can be decided for desired forest level.

Operation alternatives that can be applied in each forest area was produced by using planning

strategies determined by decision maker. Restriction equation, suitable for linear programming

structure was determined from operation alternatives, produced from decision variables. These

equations were solved by using Simplex Method (SP) and determined the best operation

alternatives

Even thought chosen prescription alternatives are the best alternative, it is impracticable

for spatial parameters due to not included into model. Spatial parameters is very difficult to

define with linear equation, mostly it is impossible. Simulated Annealing method used for to

satisfy spatial restriction with minimum penalty value of the result of LP.

Key Words: Forest Ecosystem Model, Ecosystem Based Multiple Use Forest Management,

Simulation, Growth Model, Optimization, Linear Programming, Simplex

Method, Heuristic Search Techniques, Simulated Annealing, Penalty Functions,

Spatial Planning