A global scale land surface phenology reference dataset with 12 years of spot vegetation data
(2012) PhD Student Day ENVITAM — Location: Espace Senghor, Gembloux (8.February.2012)
Files
No attached file found for this publication.
Details
- Authors
- Abstract
- Phenology is the study of the periodic biological events in the living world and of their relation with climatic factors. These events are influenced by the environment, especially the seasonal temperature changes driven by the weather and the climate. The study of vegetation phenology using remote sensing, also referred to as land surface phenology provides aggregated information at moderate to coarse spatial resolution, which relates to the timing of the vegetation growth, senescence, dormancy and associated surface phenomena at seasonal and inter-annual scales. A reliable characterisation of vegetation phenology is of direct relevance to global change science, including vegetation monitoring, net primary production estimation, climate change impact assessment, Carbon accounting and biodiversity monitoring. Identifying current land cover trends and variability creates a baseline of current ecosystem properties against which future change can be measured. The climate modeling community has also identified vegetation phenology as required additional information to the ESA-CCI land cover project. Land cover is one of the Essential Climate variables defined by the United Framework Convention on Climate Change. In order to provide seasonal reference profile to the climate modeling community and to study anomalies and trends of the land surface phenology for different vegetation types in the world, the design of a phenological reference dataset is investigated. The main objective is to measure and analyse the seasonal dynamic and inter-annual variability of the vegetation at a global scale by the means of entire profiles and specific metrics. The data consists of 4380 daily (S1 products) SPOT VEGETATION images from January 1999 to December 2010 at a spatial resolution of 1 km. Daily surface reflectance values are composited into decades to reduce clouds and haze effects using the mean compositing algorithm. Two vegetation indices (VI) are used as indicators of the vegetation growing cycle, the normalized difference vegetation index (NDVI) and the Enhanced vegetation index (EVI). The methodology consists in several steps: (i) The VI values are spatially aggregated on a 3x3 moving average window (ii) Synthetic profiles are produced by computing the median value and the standard deviation for each decades over the 12 years (iii) The synthetic profile and each annual profile are approximated by a series of logistic functions (v) The first and second derivative of these fitted curves are computed in order to extract seasonal metrics. This method is applied only on areas covered by vegetation as defined by the GLOBCOVER land cover map. The reference dataset of phenology that will be built will provide entire profiles and specific metrics. Each pixel identified as vegetation is characterized by a synthetic reference profile, presenting the pattern of variation of NDVI and EVI on a 12-month scale, and a temporal standard deviation, representing the inter-annual variability during the 12 years. Five metrics are associated to the reference profile (minimum, maximum, beginning, end and length of the season). These metrics are also available for each year. The synthetic information about the vegetation phenology will be further explored in order to study anomalies and trends of the vegetation.
- Affiliations
Citations
Verheggen, A., & Defourny, P. (2012). A global scale land surface phenology reference dataset with 12 years of spot vegetation data. In Vanclooster M. (ed.), Proceedings (p. p. 81). https://hdl.handle.net/2078.5/225170