PARENT SESSION
Oral Session # 72: GIS and Remote Sensing I.
Presiding: J Drake
Thursday, August 7. 8:00 AM to 11:30 AM, SITCC Meeting Room 203.

Land surface phenology, ecoregional expectations, and change analysis: Toward ecological forecasting with image time series.

Henebry, Geoffrey^*,1, Goodin, Douglas², ¹ University of Nebraska, Lincoln, NE² Kansas State University, Manhattan, KS

ABSTRACT- To make ecological forecasting an operational possibility, we need the capability to establish and to update complex spatio-temporal baselines that will enable prediction of the usual and identification, quantification, and assessment of the unusual. For the analysis of image time series, it is necessary to move beyond viewpoints that center on pixels or patches and characterize the dynamic of pattern and process that sequences portray. The phenology of the land surface provides a basic ecological phenomenon fruitful for comparative analyses. The USGS CONUS AVHRR maximum NDVI biweekly composites for 1990-2000 were partitioned by Omernik's 84 Level III ecoregions. For each ecoregion at each composite date, a random walk resampling procedure was used to estimate measures of spatial dependence, spatial heterogeneity, and aspatial intensity. Expectations of land surface phenology were calculated for each ecoregion. Comparisons of ecoregional expectations reveal broad patterns in the relationships between spatial dependence, spatial heterogeneity, and aspatial intensity. Analysis of seasonal trajectories in spatial structure reveals time periods (watchpoints) with high convergence or divergence in interannual variation. Patterns of these watchpoints provide a basis for forecasting anomalous land surface behavior.

Key words: Omernik Level III ecoregions, AVHRR NDVI, phenology, ecological forecasting