Design Efficacy Evaluation of a LIM-SD-based Approach for Ecological Engineered Landscaping Design: A Case Study of An Urban River Wetland

Abstract

Abstract Ecological engineered landscaping (EEL) is a type of design work which incorporates various ecological principles and aims to provide aesthetic and ecological characteristics to semi-natural environments. Digital landscape techniques, such as Landscape Information Modelling (LIM) and Stable Diffusion (SD) -based approach, has significant potential to support EEL projects for both environmental engineers and landscape architects. This study aims to quantitatively evaluate the design efficacy and feasibility of the LIM-SD-based approach applied for an urban river wetland EEL design task. An experiment was used to quantify the EEL design efficacy within the digital approach for students from different backgrounds, i.e., industrial design postgraduates (ID), landscape architecture undergraduates (UG), landscape architecture postgraduates (PG). Experts were invited to evaluate the EEL design works of the participants according to 4 criteria, i.e., "Site adequacy", "Aesthetic", "Spatial arrangement" and "Eco-engineering", while the vegetation coverage rate (VCR) and proportion of riparian areas (PRA) of each design is automatically calculate by a parametric program. Participants’ objective experiences are also evaluated by questionnaires. It indicates that LIM-SD-based approach can improve most designers' productivity in EEL design tasks. The ranking of scores for most of the quantitative indicators is PG > UG > ID. Explanation of some phenomena, i.e., the mismatch between LIM and SD generated renderings and the different objective experiences of the participants, are provided. Future research can further investigate the impact of digital EEL workflows on landscape architects’ and environmental engineers’ performance.