Abstract

The study of vegetation phenology is of great significance for understanding global climate change. The Yellow River basin has a wide spatial range and a complex ecological environment. The phenological characteristics of forest and grassland need further clarification. Based on the MODIS-EVI data from 2000 to 2018, we extracted the phenology of forest and grassland in the Yellow River basin using piecewise logistic and double logistic phenological models with the corresponding curvature change extremum method and derivative method, respectively. The temporal and spatial variations of phenological parameters were analyzed. The start of growing season (SOS) was at 90-165 day of year (DOY), and gradually delayed from southeast to northwest. The increase of 100 m elevation delayed SOS 0.94 d, and the SOS of forest was earlier than that of grassland. The end of growing season (EOS) was at 270-315 DOY, which delayed from west to southeast. For every 100 m increase in altitude, the EOS advanced 0.63 d, with EOS of forest being later than that of grassland. The length of growing season (LOS) was 110-230 d, which shortened gradually from southeast to northwest. The LOS of forest was larger than that of grassland. During the study, SOS showed an advance trend from 2000 to 2018 with a rate of 4.1 d·(10 a)-1, and the proportion of spatial advance area was 73.2%. There was an obvious advance in the central part of the basin. EOS generally showed a significant postponement trend with a rate of 2.3 d·(10 a)-1, and the proportion of spatially delayed area was 63.4%, the phenological advance and delay of forest was less stronger than that of grassland. LOS showed a significant prolongation trend with a rate of 6.4 d·(10 a)-1, and the proportion of spatial extension was 71.8%. The piecewise Logistic and double Logistic phenological models and the corresponding curvature extremum method and derivative method were suitable for the extraction of natural vegetation in the Yellow River Basin. The overall LOS of forest and grassland showed a prolonging trend, which was shortened with the increases of altitude. The LOS of forest was longer than that of grassland in the study area.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call