基于代谢理论的公路物质消耗时空演化规律与生命周期环境效应研究

The materials stored in the large-scale construction phase of highway will gradually exceed the design life, and thus huge and continuous material investments are needed in the future to maintain the existing highway network. However, many areas are facing the shortage of road resources currently; especially, the high quality aggregate are nearly exhausted. The sustainable development of highway is facing huge resource pressure and environmental concern. This project regards highway as "life body". By analogy with the natural metabolic process, it reveals the metabolic mechanism of material and energy exchange between highway and external environment, and provides new idea and method for solving the above-mentioned problems. The project first constructs the spatio-temporal model of highway material metabolism by fusing multi-source data such as high resolution RS image. Then it calculates the environmental burden within the life cycle and tracks the path, scale, and efficiency of the material metabolic process. Furthermore, it integrates all information on the GIS platform in order to measure the evolution trend of highway material consumption on the time axis, to analyze the metabolic types of typical highways on the spatial axis, and finally to predict the demand of metabolic substances in the future. The achievements of the project provide theoretical and methodological supports for evaluation of the future resource consumption, potential environmental risk and improvement of the highway metabolic patterns.

公路大规模建设阶段贮存的物质将逐步超过设计年限，因此未来需要持续性投入大量物质以维持现有路网。然而目前许多地区已然面临筑路资源短缺，尤其是优质集料接近枯竭，公路的可持续性发展面临巨大的资源压力和环境问题。本项目视公路为“生命体”，通过类比自然代谢过程，揭示公路同外界环境物质能量交换的代谢机理，为解决上述问题提供新的思路和方法。项目通过融合提取高清遥感影像等多源数据，构建公路物质代谢时空模型；计算生命周期内的环境负荷，追踪物质代谢的路径、规模和效率；集成信息于GIS平台，度量时间轴上公路物质消耗演变趋势，解析空间轴上典型公路代谢类型，预测物质代谢需求，为评估未来资源占用、潜在环境风险以及改进公路代谢模式等提供理论与方法支撑。

道路可持续发展研究缺乏对道路系统层面资源需求与环境影响的定量化分析，特别是其物质存量的规模、构成、时空分布格局、演化趋势以及生态环境影响等。本研究从代谢的视角揭示区域范围内道路系统生命周期内物质消耗的时空分布格局与演化规律并对资源投入量进行预测。.本研究以2014-2021年南京市城市道路系统为依托开展系列研究。1）基于全卷积神经网络建立了由高分遥感图像提取道路信息的网络模型，模型训练准确率95%以上。2）采用“由下及上”的MFA方法建立城市道路系统物质存量模型，并进行城市道路系统物质存量时空演变规律分析。时间上，南京市道路物质总存量总体上呈现随时间不断增长的趋势；在空间上，其呈现自中心城区向城市郊区由大变小的扩散规律。3）敏感性分析表明影响最大的参数分别是主干道道路宽度、主干道道路长度、快速路道路宽度与快速路道路长度。4）基于MFA-LCA建立了生命周期视角下的城市道路代谢-碳排放分析模型，碳排放分析结果显示材料生产阶段碳排放占比最多，超过了70%。5）基于GIS和PyEcharts建立了道路系统物质存量“4D-GIS”可视化模型，有助于城市道路物质存量时空规律对比分析。6）基于GBDT算法建立了的城市道路系统物质存量预测回归模型，训练集及测试集准确率均达到了98%。.本研究成果可为评估未来资源占用、环境风险以及改进道路代谢模式等提供思路和方法。