Identification of Optimal Satellite Compositing Length Using GLOBE Budburst Measurements

Phenology, the study of recurring biological cycles and their connection to climate, is a critical and growing field of global change research. In particular, scientists now recognize that regular satellite monitoring of the timing and length of the terrestrial growing season is a valuable metric of biospheric responses to short- and long-term climate variability. While many methodologies exist with which to detect growing season dynamics, most have a poorly understood relationship to actual ground vegetation conditions. GLOBE schools, through participation in the budburst protocols, are helping to bridge this gap between satellite observations and ground conditions. In this research we show how GLOBE budburst data can be used to select the optimal satellite compositing length (a technique used to reduce cloud, snow, and atmospheric contamination). One- and two-week compositing lengths produced similar results, both of which were superior to monthly compositing. The longer compositing length, contrary to popular remote sensing lore, tended to predict an earlier initiation of growth due to removal of inflection points in the satellite greenness time series. Overall, the GLOBE budburst data were extremely useful but also contained several troubling artifacts probably relating to infrequent observation, errors in date reporting, and use of exotic species