Satellite-based model detection of recent climate-driven changes in northern high-latitude vegetation productivity

We applied a satellite remote sensing based production efficiency model (PEM) using an integrated AVHRR and MODIS FPAR/LAI time series with a regionally corrected NCEP/NCAR reanalysis daily surface meteorology and NASA/GEWEX Surface Radiation Budget shortwave solar radiation inputs to assess annual terrestrial net primary productivity (NPP) for the pan-Arctic basin and Alaska from 1983 to 2005. Our results show that low temperature constraints on Boreal-Arctic NPP are decreasing by 0.43% per year (P < 0.001), whereas a positive trend in vegetation moisture constraints of 0.49% per year (P = 0.04) are offsetting the potential benefits of longer growing seasons and contributing to recent disturbances in NPP. The PEM simulations of NPP seasonality, annual anomalies and trends are similar to stand inventory network measurements of boreal aspen stem growth (r = 0.56; P = 0.007) and atmospheric CO₂ measurement based estimates of the timing of growing season onset (r = 0.78; P < 0.001). Our results indicate that summer drought led to marked NPP decreases in much of the boreal forest region after the late-1990s. However, seasonal low temperatures are still a dominant limitation on regional NPP. Despite recent drought events, mean annual NPP for the pan-Arctic region showed a positive growth trend of 0.34% per year (20.27 TgC/a; P = 0.002) from 1983 to 2005. Drought induced NPP decreases may become more frequent and widespread as regional ecosystems adjust to a warmer, drier atmosphere, though the occurrence and severity of drought events will depend on future patterns of plant-available moisture.