CONTROL ID: 1794118

TITLE: Integrating Statistical and Expert Knowledge to Develop Phenoregions for the Continental United States

AUTHORS (FIRST NAME, LAST NAME): Forrest M Hoffman^{1, 2}, Jitendra Kumar², William Walter Hargrove³

INSTITUTIONS (ALL): 1. Earth System Science, University of California, Irvine, CA, United States.
2. Climate Change Science Institute, Oak Ridge National Laboratory, Oak Ridge, TN, United States.
3. Eastern Forest Environmental Threat Assessment Center, USDA Forest Service, Asheville, TN, United States.

ABSTRACT BODY: Vegetated ecosystems typically exhibit unique phenological behavior over the course of a year, suggesting that remotely sensed land surface phenology may be useful for characterizing land cover and ecoregions. However, phenology is also strongly influenced by temperature and water stress; insect, fire, and storm disturbances; and climate change over seasonal, interannual, decadal and longer time scales. Normalized difference vegetation index (NDVI), a remotely sensed measure of greenness, provides a useful proxy for land surface phenology. We used NDVI for the conterminous United States (CONUS) derived from the Moderate Resolution Spectroradiometer (MODIS) at 250 m resolution to develop phenological signatures of emergent ecological regimes called phenoregions. By applying a unsupervised, quantitative data mining technique to NDVI measurements for every eight days over the entire MODIS record, annual maps of phenoregions were developed. This technique produces a prescribed number of prototypical phenological states to which every location belongs in any year. To reduce the impact of short-term disturbances, we derived a single map of the mode of annual phenological states for the CONUS, assigning each map cell to the state with the largest integrated NDVI in cases where multiple states tie for the highest frequency. Since the data mining technique is unsupervised, individual phenoregions are not associated with an ecologically understandable label. To add automated supervision to the process, we applied the method of Mapcurves, developed by Hargrove and Hoffman, to associate individual phenoregions with labeled polygons in expert-derived maps of biomes, land cover, and ecoregions. Utilizing spatial overlays with multiple expert-derived maps, this "label-stealing" technique exploits the knowledge contained in a collection of maps to identify biome characteristics of our statistically derived phenoregions. Generalized land cover maps were produced by combining phenoregions according to their degree of spatial coincidence with expert-developed land cover or biome regions. Goodness-of-fit maps, which show the strength the spatial correspondence, were also generated.

INDEX TERMS: 0476 BIOGEOSCIENCES Plant ecology, 0480 BIOGEOSCIENCES Remote sensing, 0434 BIOGEOSCIENCES Data sets, 1914 INFORMATICS Data mining.
(No Image Selected)
(No Table Selected)

Additional Details

Previously Presented Material: Only the concept was previously presented, not the results.

Contact Details

CONTACT (NAME ONLY): Forrest Hoffman
CONTACT (E-MAIL ONLY): forrest@climatemodeling.org
TITLE OF TEAM: