You are here
Harvard Forest REU Symposium Abstract 2011
- Title: Observing Phenological Events in Vegetation through Technological Methods
- Author: Rachel L Norman (University of North Carolina at Chapel Hill)
- Abstract:
Phenology is commonly referred to as the rhythm of the seasons, or the study of life cycle events. These cycles are responsive to various environmental factors, such as climate change. Studying phenology and the changes resulting from response to environmental factors are important because they affect a range of ecosystem functions- including global carbon cycles, plant and animal survival, air and soil temperatures, and soil moisture. Up until recently the most commonly used method has been field site observations. Although effective, field observations have proven to be time consuming and discrepancy amongst observers. This has prompted urgency for more timely and efficient methods through the use of technology. This project looks into the potential of these technological methods by extracting phonological dates (bud burst, peak canopy, start of senescence, max coloration, and end of senescence) from visual inspection of a digital image archive created by the Phenocam Network. The Phenocam Network consists of 24 core sites and 63 affiliated sites based primarily in the United States, with a few international sites. The network creates an archive of digital images from cameras that take multiple images daily. These different sites boast various altitudes, continental locations, climates, and resulting vegetation types. The goal of this project was to compare our human eye observations and the green chromatic coordinate (GCC) of the digital images to the MODIS (Moderate Resolution Imaging Spectroradiometer) on the Tera and Aqua satellites. Both MODIS and GCC track canopy changes and provides phenocam time series. Through the comparisons of the data sets generated from each method, we seek to develop computer models that will further our understanding of phenological transitions and their responses to climate change at regional and continental scale. We have found a variance within each set of results, indicating the bias of patterns in each program. For example, MODIS usually extracts earlier dates for phenological events. These findings confirm previous studies indicating consistent biases.
- Research Category: Biodiversity Studies
