Remotely sensed trends in the phenology of northern high latitude terrestrial vegetation, controlling for land cover change and vegetation type C. Jeganathan a, , J. Dash b , P.M. Atkinson b a Department of Remote Sensing, Birla Institute of Technology (BIT), Mesra, Ranchi 835215, Jharkhand, India b Global Environmental Change and Earth Observation Research Group, Geography and Environment, University of Southampton, Southampton SO17 1BJ, United Kingdom abstract article info Article history: Received 1 June 2013 Received in revised form 25 November 2013 Accepted 25 November 2013 Available online xxxx Keywords: NDVI Phenology Start of season End of season Northern latitude Terrestrial vegetation Global warming Remote sensing Trends in the start or end of growing season (SOS, EOS) of terrestrial vegetation reported previously as latitudinal averages limit the ability to investigate the effects of land cover change and species-wise conditioning on the presented vegetation phenology information. The current research provided more reliable estimates of the trends in the annual growth pattern of terrestrial vegetation occurring at latitudes greater than 45°N. 25 years of satellite-derived Normalised Difference Vegetation Index (GIMMS NDVI) was used and reliable vegetated pixels were analysed to derive the SOS and EOS. The rate of change in SOS and EOS over 25 years was estimated, aggregated and scrutinised at different measurement levels: a) vegetation type, b) percentage vegetative cover, c) core area, d) percentage forest cover loss, and e) latitude zones. The research presents renewed and detailed estimates of the trends in these phenology parameters in these strata. In the N 45°N zone, when only reliable pixels were considered, there was an advancement of -0.58 days yr -1 in SOS and a delay of +0.64 days yr -1 in EOS. For homogeneous vegetated areas (91100% cover at 8 km spatial resolution) the 5565°N zone showed the maximum change with -1.07 days yr -1 advancement in SOS for needle leaved deciduous vegetation, and -1.06 days yr -1 delay in EOS for broad leaved deciduous vegetation. Overall, the increasing trend in EOS during senescence (September to November) was greater in magnitude than the decreas- ing trend in SOS during spring (March to May) and the change in EOS was more consistent and greater than that in SOS. © 2014 Elsevier Inc. All rights reserved. 1. Introduction Recent global surface temperature records, after incorporating several corrections, conrm an increasing trend in the rate of global warming, with the 12 month running mean temperature reaching a record high in 2010 (Hansen, Ruedy, Sato, et al., 2010). Such strong- ly increasing trends suggest anthropogenically-induced changes in climate. Although climate is the manifestation of many global process- es occurring due to the coupling of land, ocean and atmospheric phe- nomena, human growth and development in the 20th century has been a major factor in altering the climate system. The human popula- tion is growing enormously at a rate of ~1 billion every 12 years (1.5 - billion in 1900; 3 billion in 1960; 5 billion in 1987; 6.1 billion in 2000; 6.9 billion in mid-2011) with a daily rate of 200,000 people (Weeks, 2012) and, hence, human needs are growing commensurately, inuencing the land, oceans and atmosphere and hence, the global cli- mate. Within the terrestrial biosphere the effects of a changing climate have already been observed through changes in the physiology, distribution, phenology and biodiversity of vegetative species (Cleland, Chuine, Menzel, et al., 2007; Hughes, 2000; Parmesan & Yohe, 2003; Penuelas & Filella, 2001; Thuiller, Lavorel, Araujo, et al., 2005). Hansen, Stehman, and Potapov (2010), using satellite sensor data from 2000 to 2005, revealed a massive loss of gross global forest cover (~1 million km 2 ) due to natural and human-made disturbances, with the boreal biome recording the largest loss. The boreal ecosystem represents one-third of global forest cover with a rich biodiversity and it plays a signicant role in the carbon cycle (Bradshaw, Warkentin, & Sodhi, 2009). Hence, it is important to characterise the spatio- temporal variation in photosynthetic activity driven by vegetation growth in this ecosystem. In addition to the availability of photosyn- thetically active radiation (PAR), the growth cycle of natural vegetation in the tropical regions is driven mainly by rainfall and in the northern latitudes temperature is the dominant driver (Gong & Ho, 2003; Korner & Basler, 2010; Tanja, Berninger, Vesala, et al., 2003). Hence, global warming is expected to increase the photosynthetic activity of higher northern latitude vegetation and this was also predicted through models (Piao, Friedlingstein, Ciais, et al., 2006). Many studies also con- rmed an earlier green-up and longer duration of greening season asso- ciated with increasing temperature, for example, over northern lands (N 50°N) (Xu, Myneni, Chapin, et al., 2013), in western Siberia and Remote Sensing of Environment 143 (2014) 154170 Corresponding author. Tel.: +91 8987630041; fax: +91 651 2275401. E-mail addresses: jegan_iirs@yahoo.com (C. Jeganathan), j.dash@soton.ac.uk (J. Dash), pma@soton.ac.uk (P.M. Atkinson). 0034-4257/$ see front matter © 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.rse.2013.11.020 Contents lists available at ScienceDirect Remote Sensing of Environment journal homepage: www.elsevier.com/locate/rse