Land surface dynamics by remote sensing | Lifewatch regional portal

U bent hier

Land surface dynamics by remote sensing

Screenshot UCL vegetation Screenshot UCL snow Screenshot UCL sun Screenshot UCL fire

Vegetation

Snow

Solar energy

Fire

Vegetation

The vegetation cycle is approximated using the Normalized Difference Vegetation Index (NDVI), which is sensitive to the vegetation cover and its photosynthetic activity. The NDVI time series is based on the work of A. Verhegghen (UCL), who used SPOT VGT weekly composites at 1000m of spatial resolution. Different metrics have been derived to characterize those time series. On the webGIS the long term average (week by week) NDVI can be viewed, as well as the minimum and maximum NDVI. Anomalies in NDVI will become available soon.

Snow

The snow indices are derived from 13 years of Earth observation using MODIS sensors. We therefore used the eight day maximum snow extent product of the NASA, which has a spatial resolution of 500m. The input data are filtered and the proportion of snow observations is used to estimate the snow probability. A set of temporal metrics is then derived from this average time series. Eventually, unusual snow and non-snow events are identified for each time period. On the webGIS the long term average (per date) of snow probability can be viewed, as well as the start, the end and the average duration of the snow period, and snow anomalies per date.

Solar energy

The solar energy is converted by the vegetation or dissipates as heat. Because of the orientation of the terrain, it may vary locally and hence create micro-climatic conditions. Weekly sunshine data has been produced using an astronomical model (90 m) combined with snow and clouds observations from daily satellite data. On the webGIS the long term average (per week) of solar power (W/m²) can be viewed. Anomalies in solar energy will become available soon.

Fire

The fire indices are based on the MODIS burnt area products since 2000. Because fires rarely cover large areas, the yearly burnt areas have been spatially aggregated using the merged UTM grids of the European Bird Census Council. The average burnt area and the frequency of at least one fire per cell have then been derived. A statistical test was then performed to identify unusual burnt area extents. On the webGIS the long term average (year by year) of the extent of burnt area (%) can be viewed, as well as the average extent of burnt area (%), the probability of fire (frequency of fires) and unusual fire events per year.

LifeWatch Wallonia-Brussels publishes a yearly Bulletin of Land Surface Dynamics. The pdf's are available here.