Forecast of the forest area in Poland with the use of artificial neural networks

Autorzy

• Alicja Kolasa-Więcek Institute of Innovation of Processes and Products, Faculty of Management and Production Engineering, Opole University of Technology
  Waryńskiego 75, 45-370 Opole, Poland
  phone: +48 77 42 34 039, fax: +48 77 45 38879
  e-mail: a.kolasa-wiecek@po.opole.pl

Abstract

The article presents prognostication concerning the forest area in Poland based on afew changing parameters which influence the condition of forest. An attempt was made to determine which of the parameters revealed the biggest influence. The forest area was forecasted with regard to the area of forest stands damaged by the influence of gas and dust pollutants, the area of forest fires and the area of forested grounds. Based on the data from the years 2002–2006, there was simulated the state of forests in Poland’s individual voivodeships in subsequent five years. The forecasting was conducted with the use of artificial neural networks. The results of forecasting indicated which of the analyzed parameters could negatively or positively affect the forest area in the nearest future. The generated parameters prove high probability of about 95%, which was verified by the average square error.

Keywords

• artificial neural networks
• forecast
• forest area