MONITORING LAND SURFACE PARAMETERS DURING THE TRANSITION SNOW MELT AND GREEN-UP PERIODS USING POLAR ORBITING SATELLITES

摘要

The timing of snow melt and vegetation greenup are important indicators of climate variability as well as being potentially useful for validating ecosystem models. Daily SPOT 4-Vegetation data over Canada are processed using a robust LOESS smoothing algorithm to derive continuous daily estimates of surface reflectance. Snow cover and LAI algorithms are applied to these data for the year 2000. Comparisons between satellite based and in-situ snow cover and phenology indicators are conducted. Continuous snow cover mapping at better that 80% accuracy was achieved. Green-up estimates were more varied although the daily data provided more consistent results than those derived from 10-day maximum NDVI composites.