Changes in vegetation cover and snowmelt timing in the Noatak National Preserve of Northwestern Alaska estimated from MODIS and Landsat satellite image analysis

Abstract

Trends and transitions in the growing season Normalized Difference Vegetation Index (NDVI) from the MODerate resolution Imaging Spectroradiometer (MODIS) satellite sensor at 250-m resolution were analyzed for the period from 2000 to 2018 to understand recent patterns of vegetation change in ecosystems of the Noatak National Preserve in northwestern Alaska. Statistical analysis of changes in the NDVI time series was conducted using the “Breaks for Additive Seasonal and Trend” method (BFAST). This structural change analysis indicated that at least one NDVI abrupt change breakpoint was detected in 25% of the MODIS pixels covering the NOAT. All of the large wildfires mapped by Landsat burn severity classifications within the NOAT since the year 2000 coincided with multiple negative NDVI breakpoints. Results showed that six large drainage basins, all within the Eli River and Kugururok River systems of the western NOAT, had significant correlations between early spring snowmelt and annual area burned. Later-than-average snowmelt dates were strongly associated with a high number of abrupt shifts in greenness cover detected by BFAST. Negative (browning) NDVI trends in the de-seasonalized residuals were detected as significant (p < 0.05) at 15% of the total MODIS 250-m pixel coverage of the NOAT.