作者: Kirsten M. de Beurs , Braden C. Owsley , Jason P. Julian
DOI: 10.1016/J.JAG.2015.10.009
关键词:
摘要: Abstract In this study we present, evaluate and validate an approach to monitor forest grassland disturbance. We apply the both Landsat MODIS imagery for North Island of New Zealand results based on high resolution OrbView Ikonos imagery. found overall accuracy disturbance index 98% two studied land cover types. The kappa value was 0.770 indicating a 77% better agreement than what would have occurred by chance. that there is difference between received areas compared areas, with outperforming (Kappa 0.855 vs. 0.656). split validation soil type also effect different types respect grazing pressures. behaved consistently all available orders. approximately 36.2% exotic forests, resulting in annual clearing rate 2.6% over period. Lastly present close-up changes one intensely used catchment. demonstrate December/January rates increased from about 6% 2000 16% 2012.