Using phase-spaces to characterize land surface phenology in a seasonally snow-covered landscape

Jeffery A Thompson, David J. Paull, Brian G. Lees

Research output: Contribution to journalArticlepeer-review

27 Scopus citations


In seasonally snow-covered environments, snow significantly impacts upon vegetative phenology. As such, there is a need to derive land surface phenological descriptors that are free of the influence of snow-cover. The presence of on-ground snow-cover influences the phenological descriptors produced using some remotely sensed vegetation indices. This study proposes a method that uses both the Normalized Difference Vegetation Index (NDVI) and the Normalized Difference Infrared Index (NDII) to obtain dates for the start and end of the growing season. The method operates in the phase-spaces (sometimes referred to as feature- or band-spaces) created by the intersection of the NDVI-NDII data-space of individual pixels and can be used to derive descriptors for the start and end of the growing period as well as period corresponding with the end of green-up. This paper describes the origins and rationale for the method, which was applied to a MODIS image time-series of Australia's alpine bioregion for 2000-mid-2001. The results were validated against descriptors derived using the method of Delbart et al. (2005). For the start of the growing period, the validation indicated there was moderate correlation (r= 0.51, p≤. 0.001) between the descriptors. Noise in the NDII time-series used resulted in little, or no, correlation for the end of the growing period (r= - 0.13, p≤. 0.001), and the correlation for the end of the green-up period was somewhat limited (r= 0.24, p≤. 0.001). In contrast to traditional NDVI threshold methods, the phase-space algorithm proposed here provided for a clear set of definitions that correspond with biophysical phenomena of the land surface, such as the offset and on-set of seasonal snow-cover.

Original languageEnglish (US)
Pages (from-to)178-190
Number of pages13
JournalRemote Sensing of Environment
StatePublished - Sep 1 2015
Externally publishedYes


  • Alpine
  • Australia
  • Land surface phenology
  • Seasonality


Dive into the research topics of 'Using phase-spaces to characterize land surface phenology in a seasonally snow-covered landscape'. Together they form a unique fingerprint.

Cite this