Mark Friedl awarded new grant from NASA
Department of Earth and Environment Professor Mark Friedl was recently awarded funding from NASA for his new project, “Using Three Decades of Landsat Data to Characterize Changes and Vulnerability of Temperature and Boreal Forest Phenology to Climate Change.”
Friedl and his team will use Landsat data collected over the past 30 years to analyze the growing season of temperate and boreal forests in North America. Their goal is to use the data to better understand how these forests have responded to climate change. Working with Friedl as Co-investigators on the project will be Department of Earth and Environment Professor and Chair Curtis Woodcock and current Ph.D. student Eli Melaas.
The project will begin this spring and take place over the next three years.
A detailed abstract of the project is as follows:
“Climate change is creating well-documented impacts on terrestrial ecosystems. Among the best known of these impacts are changes to the growing season of temperate and boreal forests. Changes in phenology provide useful diagnostics of climate change impacts in these biomes, influence coupled biosphere-atmosphere interactions, and also affect regional-to-global carbon budgets. Extreme events and climate variability complicate the response of ecosystems and increase vulnerability by inducing large phenological responses that affect ecosystem function at seasonal (and longer) time scales. Studies using in-situ measurements have suggested that the growing season of temperate and boreal ecosystems is changing, and remote sensing-based research using time series imagery from coarse resolution sensors appear to confirm this trend. Specifically, studies using AVHRR NDVI data have documented changes in growing season NDVI that indicate widespread perturbations to boreal and temperate forests in response to climate change. However, the coarse spatial resolution and other limitations of AVHRR data constrain the types of inferences that can be drawn from these data. Sub-pixel contamination of AVHRR time series by snow and disturbance events introduce sources of variation unrelated to phenology, and challenges associated with instrument calibration, atmospheric correction, and geo-location uncertainty further reduce the utility of these data for long-term phenology studies. In this proposal we describe research to address these challenges using Landsat data. Specifically, we propose to use a new methodology that exploits dense time series of Landsat images to quantify spatio-temporal patterns in North American temperate and boreal forest growing season dynamics. Our proposed methodology uses a sampling strategy designed to capture geographic variation in temperate and boreal forest properties, and focuses on regions of overlap between adjacent Landsat scenes, thereby significantly increasing the temporal sampling of Landsat images. Because temperate and boreal ecosystems are characterized by frequent disturbance and have snow on the ground at times of the year that are especially important for detecting changes in phenology, our methodology will exploit datasets related to fire disturbance such as the Canadian and Alaskan Large Fire databases, and will include strategies to screen and remove snow-contaminated pixels. Results from this research will yield methods and datasets for retrospective characterization of changes to temperate and boreal forest growing seasons spanning 30+ years at 30-meter spatial resolution. In doing so, this research will (1) dramatically improve information about how temperate and boreal forests have changed in response to climate change, and (2) improve understanding regarding the sensitivity and vulnerability of these forests to climate change.”