Publications

Ji, YY; Zhan, WF; Du, HL; Wang, SS; Li, L; Xiao, JF; Liu, ZH; Huang, F; Jin, JX (2023). Urban-rural gradient in vegetation phenology changes of over 1500 cities across China jointly regulated by urbanization and climate change. ISPRS JOURNAL OF PHOTOGRAMMETRY AND REMOTE SENSING, 205, 367-384.

Abstract
A comprehensive investigation on the long-term changes of urban vegetation phenology (AUVP) is critical for a better understanding of changes in urban ecosystems. However, the urban-rural gradient of the AUVP across cities with diverse climate zones and city sizes, as well as the associated drivers, remains largely unclear. Here, using multi-source remote sensing data from 2001 to 2019, we investigated the urban-rural gradient (including six urban-rural subzones, i.e., the old urban core, new urban core, new urban area, suburban area, exurban area, and rural area) of the AUVP across 1500 + cities in China and analyzed the impacts of urbanization level and climate change on the AUVP. First, over the past two decades, the AUVP during spring (i.e., the changes of the start of growing season, termed ASOS) is generally negative (advanced SOS; 73 % +/- 4 %) and that during autumn (i.e., changes of the end of growing season, termed AEOS) is mainly positive (delayed EOS; 63 % +/- 3 %) in all urban-rural subzones. The spatial distribution of the AUVP based on enhanced vegetation index (EVI) data is relatively consistent with that based on solar-induced chlorophyll fluorescence (SIF) data. Along the urban--rural gradient, both the SOS advance and EOS delay exhibit an increasing first, decreasing later, and finally steady curve, with the maximum occurring in the new urban area at -12.4 +/- 18.5 days (for ASOS) and 3.8 +/- 9.9 days (for AEOS). Second, the SOS changes from an advance to a delay from the middle temperate to southern subtropical zones, while the EOS changes from a delay to an advance. The magnitudes of both ASOS and AEOS show a general trend of increasing first and then decreasing with decreasing latitude. Third, the SOS advance and EOS delay generally decrease with city size for the old urban core, increase initially and decrease later with city size for the three suburban subzones, while they show little deviations with city size for the new urban core and new urban area. Finally, in the three urban subzones, the mean AUVP across all the cities is contributed more by climate change (mean: ASOS: 73 %; AEOS: 78 %) than by urbanization level (mean: ASOS: 27 %; AEOS: 22 %). Our findings can improve the understanding of urban vegetation changes in the context of rapid urbanization and climate change.

DOI:
10.1016/j.isprsjprs.2023.10.015

ISSN:
1872-8235