LONG-TERM GOALS My long term goals are to develop, test, and clearly present new quantitative methods for evaluating ' stresses in the earth's crust and to through new field observations and new mechanical analyses to contribute to a better understanding of geologic fracture phenomena, especially faulting, landsliding, joint formation, and dike intrusion. OBJECTIVES The main scientific objectives of this project are to identify and better understand the factors controlling where submarine landslide failure surfaces nucleate, how they propagate, how deformation accumulates in the incipient stages of landsliding, and to develop methods for analyzing these phenomena. A second objective is to reconcile predictions of fracture mechanics theory with observations of secondary fractures around faults. The landslide and faulting studies are linked because they both involve shear fracture, albeitunder different environmental conditions. The work also is undertaken with the objective of developing my graduate students as well-grounded research scientists. APPROACH This study is primarily theoretical and utilizes numerical stress analyses to understand sliding processes. Landslide failure surfaces and faults are modeled as fractures in elastic media using displacement discontinuity boundary element codes (e.g., Crouch and Starfield, 1983; Thomas, 1993). Fleming and Johnson (1989) proposed viewing landslide failure surfaces as fractures, and this concept is tested quantitatively here. The mechanical analyses for landslides have been conducted in both two-and three-dimensions and account for topography and stresses due to gravity. The stresses in a slope without a failure surface are examined to see where failure might nucleate. Stresses and displacements within a slope containing different failure surface geometries are then examined to understand how a failure surface might propagate and how the slope deforms in response. The model results are compared with observations made by other investigators to test the model predictions. For the faults, stress analyses have been conducted in three-dimensions to indicate the location, orientation, and size of secondary fractures. The mechanical analyses are then tested against my field observations of faults, collected as part of another project. Development of mechanical analysis methods is a major component of this research. IJTIG QUALITY JEZr%&Wm 4 20000627 025