Large-scale detection of vegetation dynamics and their potential drivers using MODIS images and BFAST: A case study in Quebec, Canada

Monitoring vegetation dynamics at global scale is equally important in the context of terrestrial ecosystem carbon exchange and climate-biosphere, interactions. The Breaks For Additive Seasonal and Trend (BFAST) method and Moderate Resolution Imaging Spectroradiometer (MODIS) 16-day Normalized Difference Vegetation Index (NDVI) at a spatial resolution of 250 m were used to detect vegetation dynamics in Quebec during 2000-2011. The overall agreement between BFAST detected breaks and observed disturbances was about 64% with the highest agreement up to 80% for "Fire" disturbance. The results presented in this study indicated that 25.7% of the total study area experienced NDVI trend changes with one or more breaks during 2000-2011, most of which were detected in the Boreal Shield eco-zone along the coastline of the Gulf of St. Lawrence. Abrupt vegetation changes barely varied under different eco-zones while considerably varied with different land cover types. The abrupt changes areas in 2002 and 2009 were the two greatest, with area percentages of 17.4% and 29.1% of the whole area, respectively. The area percentages of years with abrupt trend changes indicated that abrupt vegetation greening occurred in 2008 and 2009, especially in 2009, with 58.3% of the overall abrupt greening. Abrupt vegetation browning occurred in 2002, 2003, 2005 and 2007, especially in 2002, with 28.2% of the overall abrupt browning. Moreover, our results indicated that the detected vegetation trends varied temporally and spatially. Disturbances from existing field observations or remotely sensed images could only interpret < 40% of the vegetation changes. The impact of climate change on vegetation dynamic is particularly worth being investigated in the future work. To our knowledge, this study is one of the few attempting to explore large-scale detection of vegetation dynamics and their potential drivers in eastern Canada.