Research project T4/DD/10 (Research action T4)
A major challenge facing ecologists studying the earth as a dynamic system is the mapping of vegetation quantities overtime as over large areas. Maps of these quantities, such as leaf area index (LAI) or absorbed photosynthetically active radiation (APAR) are needed to parameterise biogeochemical cycle and climate models. The role of the satellite remote sensing as key source of quantitative information for the regional and the global scale is no longer discussed.
However, there is none retrieval algorithm of these variables of interest from the remotely sensed signal. Today, the data interpretation relies on a fitting of a linear or polynomial semi-empirical models based on the relationships between the field-measured variable and the sensor signal. This simplistic calibration between observations at non-compatible spatial resolutions can no longer be supported. According this approach a point measurement is directly related to a pixel signal corresponding to an area as large as one square kilometre. This research proposes to tackle the upscaling issue between observation level using the recent evolution of geostatistics.
The overall objective is to develop methods for estimating LAI and FPAR variables derived from satellite data based on geostatistics concepts. The latter would provide the basic approach to take advantage of the spatial autocorrelation of these variables.
The goal is a functional characterisation of vegetation cover by biophysical variables for various aggregation levels from local to regional scale. At the same time this study would also develop a method of quantitative assessment of seasonality and spatial variability of the vegetation cover. Moreover, the proposed approach could give a way for new upscaling methods between different observation levels in order to enhance the semi-empirical modelling based on relationships between measured and remotely sensed variables.
The research consists in statistical and geostatistical analysis of a unique and original data set on the tropical forest biome. This well-known biome for its global impact is poorly documented. Several spatio-temporal series recently acquired on the field and from space (SPOT XS, AVHRR) over a dense humid tropical forest in Central African Republic.
Geostatistics will first be used to describe and quantify the variability of the field-measured and remotely sensed variables in both space and time. Based on these results, the development of a space-time kriging approach will allow to predict at unsampled locations and times the value of the field-measured variable using field-measured and remotely sensed data available at different spatial scales and temporal resolutions. This approach will then provide an assessment of the uncertainty attached to geostatistical prediction through non-linear geostatistics and conditional simulations. Finally, the use of the model of space-time uncertainty to investigate the impact of different scenarii.
