Two-weekly AVHRR images were used to examine spatial patterns of the normalized difference vegetation index (NDVI) and their relationships with environmental variables for moist acidic tundra (MAT) and moist non-acidic tundra (MNT) along two latitudinal transects in northern Alaska. The NDVI database was derived from a 5-yr time series (1995–1999) of two-weekly AVHRR composites for Alaska. A digital climate map, digital elevation map and vegetation map were processed and overlain with the NDVI grid. Homogeneous vegetation patches for both MAT and MNT were defined as sample sites using infrared aerial photos, MSS images and the vegetation map along the transects. Linear and non-linear regression modeling were performed between NDVI indices and environmental variables, total summer warmth (TSW) and elevation. It was demonstrated that along both western and eastern transects, there were obvious latitudinal trends of peak NDVI (AP-NDVI), average growing season NDVI (GS-NDVI), and early June NDVI (EJ-NDVI). In most cases, MNT had lower NDVI values than MAT throughout the year. There were significant (p < 0.01) relations between NDVI (AP-NDVI, GS-NDVI and EJ-NDVI) and total summer warmth (TSW) and elevation in the region. EJ-NDVI showed the strongest correlation with TSW or elevation, making it the most sensitive NDVI indicator along environmental gradients in northern Alaska. NDVI was likely controlled by TSW and elevation, with the former being dominant.
Nomenclature: Yurtsev (1994).
Abbreviations: AK = Alaska; AP-NDVI = Annual peak NDVI; AVHRR = Advanced Very High Resolution Radiometer; DEM = Digital elevation model; EJ-NDVI = Early June NDVI; GS-NDVI = Growing season NDVI; LAI = Leaf-area index; MAT = Moist acidic tundra; MNT = Moist non-acidic tundra; NDVI = Normalized difference vegetation index; TSW = Total summer warmth.