This report describes a scientific investigation of the effects of eight different cleaning techniques on the Berkshire Lee marble component of the facade of the East Center Pavilion at Philadelphia City Hall; the study was commissioned by the city of Philadelphia. The eight cleaning techniques evaluated in this study were power wash (proprietary gel detergent followed by water rinse under pressure), misting (treatment with potable, nebulized water for 24-36 hours), gommage (proprietary Thomann-Hanry low-pressure, air-driven, small-particle, dry abrasion), combination (gommage followed by misting), Armax (sodium bicarbonate delivered under pressure in a water wash), JOS (dolomite powder delivered in a low-pressure, rotary-vortex water wash), laser (thermal ablation), and dry ice (powdered-dry-ice abrasion delivered under pressure). In our study approximately 160 cores were removed from the building for laboratory analysis. We developed a computer program to analyze scanning-electron-micrograph images for the microscale surface roughness and other morphologic parameters of the stone surface, including the near-surface fracture density of the stone. An analysis of more than 1,100 samples cut from the cores provided a statistical basis for crafting the essential elements of a reduced-form, mixed-kinetics conceptual model that represents the deterioration of calcareous stone in terms of self-organized soiling and erosion patterns. This model, in turn, provided a basis for identifying the variables that are affected by the cleaning techniques and for evaluating the extent to which such variables influence the stability of the stone. The model recognizes three classes of variables that may influence the soiling load on the stone, including such exogenous environmental variables as airborne moisture, pollutant concentrations, and local aerodynamics, and such endogenous stone variables as surface chemistry and microstructure (fracturing, roughness, and so on). This study showed that morphologic variables on the mesoscale to macroscale are not generally affected by the choice of a cleaning technique. The long-term soiling pattern on the building is independent of the cleaning technique applied. This study also showed that soluble salts do not play a significant role in the deterioration of Berkshire Lee marble. Although salts were evident in cracks and fissures of the heavily soiled stone, such salts did not penetrate the surface to a depth of more than a few hundred micrometers. The criteria used to differentiate the cleaning techniques were ultimately based on the ability of each technique to remove soiling without altering the texture of the stone surface. This study identified both the gommage and JOS techniques as appropriate for cleaning ashlar surfaces and the combination technique as appropriate for cleaning highly carved surfaces at the entablatures, cornices, and column capitals.