Detection of interannual variation of vegetation in middle and southern Italy during 1985-1999 with 1 km NOAA AVHRR NDVI data

The potential of NOAA-advanced very high resolution radiometer (AVHRR) time series for environmental studies was investigated at the pixel scale. We a nalyzed a multitemporal set of annual maximum value composite (MVC) of norm alized difference vegetation index. Local area coverage data of middle and southern Italy were processed from measurements taken between 1985 and 1999 of the afternoon viewing of NOAA9, 11, and 14. Significant artificial anom alies were found due to satellite switch, short-wave calibration instabilit y and illumination effects. We removed such systematic errors and achieved a strong reduction of standard deviation values (around 50%). Interesting r esults were obtained from a change detection analysis performed at the pixe l level. Outcomes from satellite-based analysis were compared with independ ent time series data, such as ancillary meteorological data, forest fire ar chives, and results from field surveys. Results showed that the geographica l areas where MVC data indicated a decrease in vegetation activity match we ll with areas affected by forest fires, intense human activity, or rapid de cline of coniferous forests. In contrast, an increase in MVC was found in r egions recently involved in a growth of intensive fanning or invaded by ali en plants which are now recognized as a threat to native species. Our explo ratory results indicate that high-quality AVHRR data can profitably support studies on interannual dynamics of surface parameters. In particular, our parameterization of illumination and calibration adjustments can be directl y applied to Mediterranean-like ecosystems.