Barker, J.R., M. Bollman, M. Fiorella, G. Bradshaw, Paul L. Ringold. 1998. Texture analysis of riparian coniferous forest ADAR imagery to identify potential ecological indications for monitoring In: Proceedings of the First International Geospatial Conference on Agriculture and Forestry: Decision, Technology and Applications, Vol 1. ERIM International, Ann Arbor : pp. 449-456
Digital, multispectral data were used to characterize riparian sites, and to identify possible ecological indicators for long-term monitoring. Research goals were to 1) evaluate riparian forest attributes that can be distinguished with one-meter resolution digital imagery; 2) identify what type of variance texture analysis provide this information; and 3) evaluate if texture analysis can indirectly characterize below-canopy attributes. The ADAR remote sensing system captured one-meter resolution imagery of different riparian sites in Drift Creek Basin of coastal northwestern Oregon. The images were subjected to a variance texture analysis. Field data on tree density and size were collected from these same riparian sites. Data analyses show that the use of the high-resolution, multispectral imagery provided a rich source of information on coniferous riparian attributes. Conifer and deciduous crown cover, total crown cover, and tree-size composition were highly correlated with different texture variables, and therefore can be estimated with regression models. In addition, conifer and deciduous crown cover, and total crown cover were correlated with tree density and tree basal area. These canopy and below-canopy attributes appear to be good indicators or riparian structure because they can be estimated with a known degree of reliability, and provide important information on riparian condition.