SCALE-DEPENDENT ANALYSIS OF SATELLITE IMAGERY FOR CHARACTERIZATION OFGLACIER SURFACES IN THE KARAKORAM HIMALAYA

Information regarding process-structure relationships and change in th e Karakoram Himalaya is of great importance in studying glacier hydrol ogy, mass balance, and dynamic environmental change. Such information is not readily available. Detailed spatio-temporal assessment requires field investigation coupled with quantitative remote sensing studies. We conducted an investigation of the large Batura Glacier in Pakistan to determine if spectral variability can be quantified and used to ch aracterize glacier surfaces. Specifically, SPOT Panchromatic satellite data were evaluated for differentiating features of glacier structure resulting from ice movement, ablation, and supraglacial fluvial actio n. Image semivariogram analysis was conducted for assessing spectral v ariability patterns and fractal analysis was used to examine scale-dep endent variation in the data. Results indicate that spectral variabili ty from fields of ice seracs can exhibit fractal characteristics, alth ough most surface features on the glacier exhibit a change in the frac tal dimension over different ranges in scale. The fractal dimension wa s found to be useful for differentiating between glacier surfaces such as white ice and debris-covered ice. Characteristics of the debris-lo ad and the scale-dependent nature of calculating the fractal dimension ultimately determined the potential of class separability. (C) 1998 E lsevier Science B.V.