Analysis of relationship between forest decline and precipitation changes in Ilam Province

Abstract

Oak decline phenomenon recently led to considerable dieback within Zagros forests. This study was designed to assess the temporal trends of changes in forest cover of the Ilam province in association with precipitation variations during a 14-year period (2000-2013). To this aim remote sensing and GIS techniques were used. The remote sensing data included 16-day NDVI products of MODIS data as well as Google-Earth archive and the related control points. Additionally climatic data from synoptic stations, GPCC long-term data and TRMM rainfall data were applied. The optimal number of land cover and forest classes was defined by means of an unsupervised ISODATA method and mean and minimum divergence separation indexes. The result indicated 35 and 30 land cover classes to be appropriate for annual and long-term assessments, respectively. Spectral graph revealed a growth period from mid-February to early-May within Ilam forests. NDVI trend revealed the first significant reduction in forest cover in 2005 and the second one in 2008, which was followed by a descending trend in forest cover for the last 3-years of the investigated 14-year period. The data from ground stations and rainfall grid point indicated negative annual and monthly rainfall trends. The trend was significant within forests in 90% and 95% confidence levels. The peak of the declining precipitation trend was observed in the beginning of growth season in March. Correlating the NDVI with and precipitation trend revealed significant reduction of annual precipitation and drought stress during the growth season are particularly responsible for considerable amounts of forest dieback and chlorophyll decline within forests of Ilam.

Keywords

• Separability
• oak decline
• GPCC
• satellite TRAA
• Satellite TRMM