Dual Scale Trend Analysis Distinguishes Climatic from Anthropogenic Effects on the Vegetated Land Surface

We present a dual scale trend analysis for characterizing and comparing two contrasting areas of change in Russia and Kazakhstan that lie less than 800 km apart. We selected a global NASA MODIS (moderate resolution imaging spectroradiometer) product (MCD43C4 and MCD43A4) at a 0.05◦ (∼5.6 km) and 500 m spatial resolution and a 16-day temporal resolution from 2000 to 2008. We applied a refinement of the seasonal Kendall trend method to the normalized difference vegetation index (NDVI) image series at both scales. We only incorporated composites during the vegetative growing season which was delineated by start of season and end of season estimates based on analysis of normalized difference infrared index data. Trend patterns on two scales pointed to drought as the proximal cause of significant declines in NDVI in Kazakhstan. In contrast, the area of increasing NDVI trend in Russia was linked through the dual scale analysis with agricultural land cover change. The coarser scale analysis was relevant to atmospheric boundary layer processes, while the finer scale data revealed trends that were more relevant to human decision-making and regional economics