DOI: 10.1140/epjst/e2008-00839-y
Identifying spatial pattern of NDVI series dynamics using recurrence quantification analysis
A case study in the region around Beijing, China
S.C. Li1, Z.Q. Zhao1, 2 and F.Y. Liu11 College of Urban and Environmental Sciences, Peking University, 100871 Beijing, China
2 The Key Laboratory for Environmental and Urban Sciences, Shenzhen Graduate School, Peking University, 518055 Shenzhen, China
scli@urban.pku.edu.cn
Abstract
Ecosystem is a prototypical complex system, exhibiting a
non-stationary temporal dynamics and complicated spatial patterns,
the characterization and description of which is riddled with
challenges. In this paper recurrence quantification analysis (RQA),
an extended branch of recurrence plots (RPs), was used to measure
the determinism and predictability of Normalized Difference
Vegetation Index (NDVI) series and its spatial patterns. After
introducing the theoretical background of RPs and RQA indices, the
implementation of this methodology was demonstrated using NDVI data
of region around Beijing, China. The results show that the RQA
indices can efficiently capture the nonlinear features of NDVI
series and explicitly identify the spatial patterns. The temporal
variation and dynamics of NDVI series shows significant spatial
differences with the change of landuse and landcover types,
characterizing by higher determinism and predictability in natural
ecosystem and lower determinism and predictability in agricultural
ecosystem. The research work indicates that combination of recurrence quantification analysis and geographical information
system can offer an alternative approach to identifying the spatial
pattern of temporal NDVI series dynamics.
© EDP Sciences, Springer-Verlag 2008