Quantifying the lagged effects of climate factors on vegetation growth in 32 major cities of China. (December 2021)

• Record Type: Journal Article
• Authors: Tang, Wenxi; Liu, Shuguang; Kang, Peng; Peng, Xi; Li, Yuanyuan; Guo, Rui; Jia, Jingni; Liu, Maochou; Zhu, Liangjun
• Is Part Of: Ecological indicators. Volume 132(2021)
• Publication Date: 2021-12
• Subjects: Climate change -- Climate factors -- Time-lag effect -- Vegetation growth -- Enhanced vegetation index; Environmental monitoring -- Periodicals; Environmental management -- Periodicals; Environmental impact analysis -- Periodicals; Environmental risk assessment -- Periodicals; Sustainable development -- Periodicals; 333.71405
• DOI: 10.1016/j.ecolind.2021.108290
• Languages: English
• ISSNs: 1470-160X

Abstract

Highlights: We quantified the time-lag effects of vegetation growth on climate factors. The time-lag effects of vegetation growth on climate factors vary spatially. Applicability of vector autoregressive model (VAR) in time-lag effects are proved. VAR model improves existing models' ease of prediction accuracy. Abstract: Climate change affects vegetation growth around the world. It has been recognized that the effect of climate change on vegetation growth exhibits hysteresis. However, the duration and intensity of time-lag effect of climate factors on vegetation growth is still difficult to quantify. We analyzed the impacts of climate on vegetation growth in 32 major cities of China from 2010 to 2016. Vegetation growth conditions were characterized using enhanced vegetation index (EVI) datasets from Moderate Resolution Imaging Spectrometer (MODIS). The climate data were extracted from the Daily Value Data Set of China Surface Climate Data (V3.0), including precipitation (PRE; mm), air temperature (TEM; o C), sunshine duration (SSD; h), humidity (RHU; %), and evapotranspiration (EVP; mm). We used the vector autoregressive model (VAR) to analyze the lagged effects of climate factors on EVI, predict vegetation responses to future global changes, and validate its accuracy. Results showed that RHU had the longest (6.13 ± 1.96 months) and strongest (median 0.34 EVI per unit RHU in the first lag period) time-lag effect on EVI, while EVP had the shortest (3.45 ± 1.09 months) and weakest (median −0.02 EVI per unit EVP in the first lag period) time-lag effect on EVI. The time-lag effects of PRE and SSD on EVI were stronger in the south than in the north. Meanwhile, the EVI predicted by the VAR model was highly consistent with the observed EVI (root mean squared error, RMSE < 0.08), and the prediction accuracy generally improved by 23.43% compared with the EVI predicted by the multiple linear regression model (MLR). Our study highlights the necessity of considering time-lag effects when exploring vegetation-climate interaction. The methods developed in this study can be used to reveal the lagged effects of climatic factors on vegetation growth and improve prediction of EVI dynamics under climate change.