Assessment of Forest Damage in Croatia using Landsat-8 OLI Images

Background and Purpose: Rapid assessments of forest damage caused by natural disasters such as ice-break, wind, flooding, hurricane, or forest fires are necessary for mitigation and forest management. Forest damage directly impacts carbon uptake and biogeochemical cycles, and thus, has an impact on climate change. It intensifies erosion and flooding, and influences socio-economic well-being of population. Quantification of forest cover change represents a challenge for the scientific community as damaged areas are often in the mountainous and remote regions. Forested area in the western Croatia was considerably damaged by ice-breaking and flooding in 2014. Satellite remote sensing technology has opened up new possibilities for detecting and quantifying forest damage. Several remote sensing tools are available for rapid assessment of forest damage. These include aerial photographic interpretation, and airborne and satellite imagery. This study evaluates the capability of Landsat-8 optical data and a vegetation index for mapping forest damage in Croatia that occurred during the winter of 2014. Materials and Methods: The change detection analysis in this study was based on the Normalized Difference Vegetation Index (NDVI) difference approach, where preand postevent Landsat-8 images were employed in the ENVI image change workflow. The validation was done by comparing the satellite-generated change detection map with the ground truth data based on field observations and spatial data of forest management units and plans. Results: The overall damage assessment from this study suggests that the total damaged area covers 45,265.32 ha of forest. It is 19.20% less than estimated by Vuletić et al. [3] who found that 56,021.86 ha of forest were affected. Most damage was observed in the mixed, broadleaf and coniferous forest. The change errors of commission and omission were calculated to be 35.73% and 31.60%, respectively. Conclusions: Landsat-8 optical bands are reliable when detecting the changes based on the NDVI difference approach. The advantage of Landsat-8 data is its availability to acquire data and detect Abstract 1 Bowling Green State University, School of Earth, Environment and Society, 190 Overman Hall, Bowling Green, OH-43403, Ohio, USA 2 G-ECO Research, Toronto, ON M5K 1P2, Ontario, Canada 3 Croatian Forest Research Institute, Division for Forest Management and Forestry Economics, Trnjanska cesta 35, HR-10000 Zagreb, Croatia

[1]  Jean-Pierre Wigneron,et al.  Object-based change detection in wind storm-damaged forest using high-resolution multispectral images , 2014 .

[2]  C. Steinmeier,et al.  Detection of storm losses in the Alpine forest areas by different methodical approaches using high-resolution satellite data , 2002 .

[3]  A. B. Miller,et al.  An analysis of land cover changes in the Northern Forest of New England using multitemporal Landsat MSS data , 1998 .

[4]  Change Detection of Vegetation Cover, using Multi- Temporal Remote Sensing Data and GIS Techniques. , 2008 .

[5]  Global variability of terrestrial surface properties derived from MODIS visible to thermal-infrared measurements , 2005, Proceedings. 2005 IEEE International Geoscience and Remote Sensing Symposium, 2005. IGARSS '05..

[6]  M. Bauer,et al.  Digital change detection in forest ecosystems with remote sensing imagery , 1996 .

[7]  Ian Olthof,et al.  Modelling and Mapping Damage to Forests from an Ice Storm Using Remote Sensing and Environmental Data , 2005 .

[8]  S. Clandillon,et al.  Benefits of SPOT 5 HR and VHR data for forest management and windfall damage mapping , 2003, IGARSS 2003. 2003 IEEE International Geoscience and Remote Sensing Symposium. Proceedings (IEEE Cat. No.03CH37477).

[9]  Hui Qing Liu,et al.  An error and sensitivity analysis of the atmospheric- and soil-correcting variants of the NDVI for the MODIS-EOS , 1994, IEEE Trans. Geosci. Remote. Sens..

[10]  Yeqiao Wang,et al.  Remote sensing change detection tools for natural resource managers: Understanding concepts and tradeoffs in the design of landscape monitoring projects , 2009 .

[11]  A. Strahler,et al.  Indicators of land-cover change for change-vector analysis in multitemporal space at coarse spatial scales , 1994 .

[12]  Hui Qing Liu,et al.  A feedback based modification of the NDVI to minimize canopy background and atmospheric noise , 1995, IEEE Transactions on Geoscience and Remote Sensing.

[13]  Ian Olthof,et al.  Mapping deciduous forest ice storm damage using Landsat and environmental data , 2004 .

[14]  Ivan Balenović,et al.  Assessment of Forest Damage in Croatia Caused by Natural Hazards in 2014 , 2014 .

[15]  Ronald E. McRoberts,et al.  Using satellite image-based maps and ground inventory data to estimate the area of the remaining Atlantic forest in the Brazilian state of Santa Catarina , 2013 .

[16]  Fugui Wang,et al.  Comparison of remote sensing change detection techniques for assessing hurricane damage to forests , 2010, Environmental monitoring and assessment.

[17]  Xianjun Hao,et al.  Post-hurricane forest damage assessment using satellite remote sensing , 2010 .

[18]  W. Cohen,et al.  Aerial and satellite sensor detection and classification of western spruce budworm defoliation in a subalpine forest , 1995 .

[19]  Eric Kwabena Forkuo,et al.  Analysis of Forest Cover Change Detection , 2012 .

[20]  Robert H. Fraser,et al.  A method for detecting large-scale forest cover change using coarse spatial resolution imagery , 2005 .

[21]  Warren B. Cohen,et al.  Spatial, spectral and temporal patterns of tropical forest cover change as observed with multiple scales of optical satellite data , 2007 .