Mark A. Friedl

Learn More
This paper presents a new approach to identifying and eliminating mislabeled training instances for supervised learning. The goal of this approach is to improve classiication accuracies produced by learning algorithms by improving the quality of the training data. Our approach uses a set of learning algorithms to create classiiers that serve as noise lters(More)
a Department of Geography and Environment, Boston University, 675 Commonwealth Avenue, Boston, MA 02215, USA b Earth Resources Technology, Inc., NASA Goddard Space Flight Center, Code 614.5, Greenbelt, MD 20771, USA c Center for Sustainability and the Global Environment, University of Wisconsin-Madison, 1710 University Avenue, Room 264, Madison, Wisconsin(More)
Accurate measurements of regional to global scale vegetation dynamics (phenology) are required to improve models and understanding of inter-annual variability in terrestrial ecosystem carbon exchange and climate–biosphere interactions. Since the mid-1980s, satellite data have been used to study these processes. In this paper, a new methodology to monitor(More)
This paper presents a new approach to identifying and eliminating mislabeled training instances. The goal of this technique is to improve classiication accuracies produced by learning algorithms by improving the quality of the training data. The approach employs an ensemble of clas-siiers that serve as a lter for the training data. Using an n-fold cross(More)
We use eddy covariance measurements of net ecosystem productivity (NEP) from 21 FLUXNET sites (153 site-years of data) to investigate relationships between phenology and productivity (in terms of both NEP and gross ecosystem photosynthesis, GEP) in temperate and boreal forests. Results are used to evaluate the plausibility of four different conceptual(More)
Despite early speculation to the contrary, all tropical forests studied to date display seasonal variations in the presence of new leaves, flowers, and fruits. Past studies were focused on the timing of phenological events and their cues but not on the accompanying changes in leaf area that regulate vegetation-atmosphere exchanges of energy, momentum, and(More)
Land cover and vegetation classification systems are generally designed for ecological or land use applications that are independent of remote sensing considerations. As a result, the classes of interest are often poorly separable in the feature space provided by remotely sensed data. In many cases, ancillary data sources can provide useful information to(More)
a r t i c l e i n f o Green leaf phenology is known to be sensitive to climate variation. Phenology is also important because it exerts significant control on terrestrial carbon cycling and sequestration. High-quality measurements of green leaf phenology are therefore increasingly important for understanding the effects of climate change on ecosystem(More)