Submitted to: Society for Range Management
Publication Type: Abstract Only
Publication Acceptance Date: 12/2/2004
Publication Date: 2/11/2005
Citation: Clark, P.E., and Hardegree, S.P., 2005. Quantifying vegetation change with repent landscape photography. (abstract). 58th Annual Society for Range Management Meeting, February 5-11, 2005. Fort Worth, TX. CD-ROM Abstract. Interpretive Summary: Sound natural resource management requires rigorous evaluation of the successes and failures past management efforts. Researchers and managers lack of the tools and techniques, however, to extensively assess very long-term vegetation changes that may have occurred on rangelands in response to past management. While aerial photography of rangelands may only extend back to the early 1930s, landscape photography documents the condition of many rangelands back prior to1900 and often to the 1860s. Unfortunately, contrasts between historic and repeat landscape photography have only been qualitative in nature until now. On a mountain big sagebrush rangeland in southwestern Idaho, we developed and tested digital image preparation, sampling, and analysis techniques for quantifying vegetation changes based on repeated landscape photography. This research now allows scientists and managers to effectively utilize landscape photography to assess very detailed vegetation changes on extensive rangelands at time-spans potentially inclusive of the past 150 years. These techniques have applicability in all land types but the limitations of oblique photography must be considered in each application.
Technical Abstract: Quantitative assessment of vegetation change is often conducted using digital analysis of aerial or vertical photography time-series. Use of repeat landscape photography for change detection, however, has been limited merely to qualitative assessments. The purpose of this study was to develop sampling and analysis techniques for using digitized, landscape photography time-series to quantify vegetation change on rangeland landscapes. Digital images were created from black and white, landscape photographs acquired in 1917, 1962, and 2000 near Whiskey Mountain in the Reynolds Creek Experimental Watershed in southwestern Idaho. Images were spatially registered to each other using control points and a polynomial transformation algorithm. Thirty random pixels along each of 30 random image lines were selected as point samples (n = 900) from within each image. Landscape features represented in each selected pixel were classified into 15 cover types. Cover type classification accuracy was estimated to be 92.2% for the 2000 image based on ground-truth data collected in the field. Classification accuracy was increased to 98.9% by combining rare or poorly separable cover type classes. Image cover of vegetation cover types was quantified for each photography acquisition date. Changes in image cover of each cover type and direction of cover type conversions were determined for each intervening time period. Analysis of cover using repeated landscape photography is constrained by limitations imposed by oblique view angles and variable image quality. Repeat landscape photography, however, can be used to quantitatively assess long-term dynamics of vegetation cover on rangeland landscapes with visually distinct vegetation types.