IKONOS image object-oriented analyses of 2011 Northern Michigan cherry orchard yields with advanced geospatial techniques
Geography & Planning Department
College of Liberal Arts and Sciences
Social and Behavioral Sciences
The northern Michigan region of Grand Traverse Bay has the highest concentrations of tart cherries in the Midwestern United States. Cherry growers and researchers visually inspect their crops for various diseases, though it is difficult to observe/monitor /estimate yields for entire orchards in an efficient manner. High-resolution IKONOS multispectral images in the 3 visible and near-infrared channels were used to identify individual tree crowns, and possibly determine the health of cherry orchard trees. IKONOS images were acquired in mid-July 2011 during the cherry ripening period. Red ripened cherries have a distinctive reflectance when compared with the orchard leaves and tree limbs. Corresponding fieldwork acquired GPS locations and photographs of 250+ cherry trees across 25 orchards. Densiometer readings and visual disease evaluations were recorded to ascertain tree crown densities and tree health. Image classification and advanced analysis techniques such as vegetation indices, Object-oriented classification, sub-pixel analyses were used to develop a model to estimate the cherry orchards yields. Object-oriented classification delineated medium/sparsely dense cherry orchards accurately for this 150,000 acres. For 8-10% of the study area, dense deciduous forest stands were mis-classified as very dense cherry orchards. Additional modeling techniques were used to distinguish the highly dense cherry orchards and dense forest stands in order to improve USDA's Federal Marketing Orders and refine National Agricultural Statistics Service yield reports. These advanced analyses will enhance remote monitoring for cherry growers/researchers, and enhance knowledge of cherry orchard growth and promote the cherry market.
111th Annual Geographers Conference
Ma, Kin M., "IKONOS image object-oriented analyses of 2011 Northern Michigan cherry orchard yields with advanced geospatial techniques" (2015). Faculty Scholarly Dissemination Grants. 525.
This document is currently not available here.