Super typhoon Haiyan makes landfall
Super typhoon Haiyan
Subscribe here to receive special images


Lake Maggiore - Italy, 3-6 April 2000

Mediterranean habitats: a multi-variate analysis of VEGETATION data.

Agustín Lobo1, Jordi Carreras2 and Josep-Maria Ninot2
1Institut de Ciències de la Terra "Jaume Almera" (CSIC)
Lluis Solé Sabarís s/n, 08028 Barcelona, Spain
2Departament de Biologia Vegetal, Facultat de Biologia, Universitat de Barcelona
Diagonal, 645, 08028 Barcelona, Spain

Paper (pdf file, 174 k)

Annual cycles of greenness, as observed from satellite imagery at resolutions ranging from 1 to 8 km, have proven to be a useful proxy of the phenology of light interception. Such data have been applied to calculate primary production, to parameterize dynamic vegetation models and to produce digital land cover charts at continental and global scales. VEGETATION imagery has superior radiometric and geometric specifications compared to the commonly used NOAA-AVHRR images, which should result into a more accurate description of phenolgy. We have analyzed the annual cycles of VEGETATION reflectance quotients for a number of very detailed habitat types in the NW Mediterranean basin. We have processed 36 S10 VEGETATION images and 18 digital maps of habitats at scale 1:50 000, which legend is based on the Directive 92/43 of the European Union with specific improvements for Catalonia (NE Spain). We selected all VEGETATION pixels (1 km2) that were included (at least in a 90%) within one single habitat patch, and extracted their time series of reflectance data and ancillary information. We calculated normalized differences of near-infrarred minus red and of near-infrarred minus medium-infrarred reflectance values. We run a principal component analysis for each set of time series and projected the observations on the planes defined by the first two principal components. We found that individual time series tended to cluster by habitat type. The distribution of habitats on the principal component plane indicated that the first axis is related to the average value of the time series while the second one is related to the position of the maximum value (spring or summer). In few cases some individual series were segregated from their clusters, but we could track back those observations to the 1:50 000 habitat cartography and to the 1:25 000 ortho-imagery and explain their behavior, normally due to land use change or to mixing in the habitat cartography. In two cases our results suggest the need of modifying the legend of the habitats cartography. Our work shows the interest of detailed studies to fully understand the dynamics of multi-temporal VEGETATION imagery and favors the opinion that an analogous sensor with 1 ha resolution would be relevant for the Mediterranean basin.