Document: JIO-3-89-7

Seasonal patterns of NOAA-AVHRR derived NDVI in Arctic tundra vegetations of the North Slope of Alaska

JIA, J.* ¹, H.E.EPSTEIN ¹ and D.A.WALKER ²

University of Virginia, Charlottesville, VA 22903 USA ¹
University of Alaska FairbanksFairbanks, AK 99775-7000 USA ²

Abstract:
Biweekly NOAA Very High Resolution Radiometer (AVHRR) images were used to examine the seasonal patterns of normalized difference vegetation index (NDVI) and their relationships with climate gradients within four major vegetation types (moist acidic tundra, MAT; moist nonacidic tundra, MNT; wet tundra, WT; and shrub tundra, ST) on the North Slope of Alaska. NDVI was derived from a 4-year time series (1995 -1998) of biweekly AVHRR data for Alaska. A series of simulated digital climate maps and a vegetation map were processed and overlaid with the NDVI grid. Homogeneous vegetation patches for each of the four vegetation types were defined using aerial photos, TM images and vegetation map, with mean size of 30 pixels. Mean and variance statistics of NDVI for each 2-week period were compared among the vegetation types. Throughout the North Slope, there was an obvious spatial variations of peak NDVI along latitude and elevation gradients, i.e., lower peak NDVI in the coastal plain in the north, and higher values in foothill in the south. In most cases, wet tundra had the lowest NDVI values throughout the year, while shrub tundra had the highest values. The peak NDVI appeared in the period of July 22 - August 4 (203/216 Julian days) for all four vegetation types, with the value of 0.57 for ST, 0.53 for MAT, 0.41 for MNT and 0.32 for WT. The earliest onset of greenness occurred in ST, followed by MAT and MNT, while WT had the latest onset. There was no obvious linear relationship between peak NDVI and monthly maximum air temperature in the region. It is suggested that more detailed and accurate climate data are necessary for further analysis.

Keywords: arctic tundra, seasonal pattern, NOAA-AVHRR, NDVI, vegetation, North Slope, onset

This abstract is being presented at: 4:45 PM in session:
Oral Session #64: Remote Sensing.