Comparison of phenology trends by land cover class: a case study in the Great Basin, USA


Bethany Bradley, tel. +401 440 9660, fax +609 258 0390, e-mail:


Direct impacts of human land use and indirect impacts of anthropogenic climate change may alter land cover and associated ecosystem function, affecting ecological goods and services. Considerable work has been done to identify long-term global trends in vegetation greenness, which is associated with primary productivity, using remote sensing. Trend analysis of satellite observations is subject to error, and ecosystem change can be confused with interannual variability. However, the relative trends of land cover classes may hold clues about differential ecosystem response to environmental forcing. Our aim was to identify phenological variability and 10-year trends for the major land cover classes in the Great Basin. This case study involved two steps: a regional, phenology-based land cover classification and an identification of phenological variability and 10-year trends stratified by land cover class. The analysis used a 10-year time series of Advanced Very High Resolution Radiometer satellite data to assess regional scale land cover variability and identify change. The phenology-based regional classification was more detailed and accurate than national or global products. Phenological variability over the 10-year period was high, with substantial shifts in timing of start of season of up to 9 weeks. The mean long-term trends of montane land cover classes were significantly different from valley land cover classes due to a poor response of montane shrubland and pinyon-juniper woodland to the early 1990s drought. The differential response during the 1990s suggests that valley ecosystems may be more resilient and montane ecosystems more susceptible to prolonged drought. This type of regional-scale land cover analysis is necessary to characterize current patterns of land cover phenology, distinguish between anthropogenically driven land cover change and interannual variability, and identify ecosystems potentially susceptible to regional and global change.