Response of the normalized di erence vegetation index to varying cloud conditions in Arctic tundra environments

Frequent cloud cover is a major environmental constraint to optical remote sensing studies in Arctic locations, including studies based on ground or aircraft observations. The objective of this study was to determine how cloud induced variations in solar illumination a€ ect the normalized di€ erence vegetation index (NDVI) of representative vegetation types on the North Slope of Alaska. Illumination conditions were quanti® ed using a cloud index (incident shortwave radiation at the surface divided by radiation at the top of the atmosphere). The results indicated that the NDVI was stable across a wide range of cloud index values, particularly when the value exceeded 0.5. It is concluded that a cloud index threshold may be used to select unbiased NDVI values from a data set collected under varying illumination conditions.