Exploring the response of net primary productivity variations to urban expansion and climate change: A scenario analysis for Guangdong Province in China

Urban land development alters landscapes and carbon cycle, especially net primary productivity (NPP). Despite projections that NPP is often reduced by urbanization, little is known about NPP changes under future urban expansion and climate change conditions. In this paper, terrestrial NPP was calculated by using Biome-BGC model. However, this model does not explicitly address urban lands. Hence, we proposed a method of NPP-fraction to detect future urban NPP, assuming that the ratio of real NPP to potential NPP for urban cells remains constant for decades. Furthermore, NPP dynamics were explored by integrating the Biome-BGC and the cellular automata (CA), a widely used method for modeling urban growth. Consequently, urban expansion, climate change and their associated effects on the NPP were analyzed for the period of 2010-2039 using Guangdong Province in China as a case study. In addition, four scenarios were designed to reflect future conditions, namely baseline, climate change, urban expansion and comprehensive scenarios. Our analyses indicate that vegetation NPP in urban cells may increase (17.63 gC m(-2) year(-1)-2335 gC m(-2) year(-1)) in the climate change scenario. However, future urban expansion may cause some NPP losses of 241.61 gC m(-2) year(-1), decupling the NPP increase of the climate change factor. Taking into account both climate change and urban expansion, vegetation NPP in urban area may decrease, minimally at a rate of 228.54 gC m(-2) year(-1) to 231.74 gC m(-2) year(-1). Nevertheless, they may account for an overall NPP increase of 0.78 TgC year-1 to 1.28 TgC year(-1) in the whole province. All these show that the provincial NPP increase from climate change may offset the NPP decrease from urban expansion. Despite these results, it is of great significance to regulate reasonable expansion of urban lands to maintain carbon balance. (C) 2014 Elsevier Ltd. All rights reserved.