Integrated Framework and Technical Path for Multi-level Nested Assessment of Landscape Character

  • Yuncai WANG ,
  • Qizhen DONG
Expand
  • College of Architecture and Urban Planning (CAUP), Tongji University

WANG Yuncai, Ph.D., is a tenured professor and doctoral supervisor in the College of Architecture and Urban Planning (CAUP), Tongji University, and an editorial board member of this journal. His research focuses on pattern language, and teaching, research and engineering practice of landscape ecological planning and design

DONG Qizhen is a master student in the College of Architecture and Urban Planning (CAUP), Tongji University. Her research focuses on landscape ecological planning

Received date: 2024-04-13

  Revised date: 2024-11-13

  Online published: 2025-12-07

Copyright

Copyright © 2025 Landscape Architecture. All rights reserved.

Abstract

[Objective] The landscape character assessment system is an effective tool to help people understand the history and current status of landscapes. Its results are widely used in land decision-making and spatial planning control. Landscape character assessment (LCA) and landscape personality assessment (LPA) are two different perspectives and systems. LCA has a certain research and practical foundation, forming a relatively mature methodology system, emphasizing the characterization of the current status of landscape. In contrast, there is relatively little research and practice related to LPA, although attention has been paid to the trend of landscape changes. In addition, there is a nested relationship in terms of value implications, landscape spatial carrier, evaluation index system, and practical application of results, and there is a certain degree of complementarity between LPA and LCA. Integrating LCA and LPA and constructing a multi-level nested framework for landscape character assessment can sort out the multi-level relationships of landscape character representation, better meet practical needs at different scales, and help people comprehensively understand the past, present, and future of landscape.
[Methods] Taking the landscape character assessment system as the research object, this research analyzes the concepts and connotations related to landscape character, and evaluates the shortcomings of existing LCA and LPA research and practice from the aspects of value dimension, indicator system, process characteristic, classification method, etc. There is considerable research and practical experience in the LCA research field, and scholars have also explored LPA. This research analyzes, evaluates, and horizontally compares relevant typical research and practical projects both domestically and internationally. The overall landscape carries the overall humanistic ecosystem, whose structurality and decomposability determine that landscape space is a complex composed of multiple relatively independent spaces, concatenation and nesting determine the multi-level nesting of the overall landscape space, and perceptibility and symbolism determine that identifying the characteristics of a landscape is a way to recognize the unique value of the landscape. Based on existing research content, the research summarizes the overall characteristics and future development trends of the two assessment methods, analyzes the differences and underlying connections between them, and explores possible integration methods for them.
[Results] LCA can describe what a landscape is, while LPA can explore why it is. Comparing the methodology of LCA with that of LPA, LPA and LCA have certain complementarity in research perspective, indicator system, classification method, and other aspects. LCA focuses on the objective description of elements and their combination level features, which can depict local landscape differences and is supported by quantitative analysis techniques; LPA focuses on the comprehensive effects in the dynamic process of resource combination, which can characterize the value and personalized characteristics of the overall landscape, but lacks quantitative classification techniques. Landscape personality is formed by highly condensed landscape character with inherent attributes that are perceived by humans, and LCA is the foundation of LPA formation. LCA and LPA have certain complementarity in different stages of landscape character assessment, and are two sets of local character representation ideas suitable for different scenes in the landscape character assessment system. The integration of the two can help optimize the landscape character assessment system in both theory and practice. Based on the overall characteristics of landscape and the nesting of landscape spatial units, integrating the perspectives and systems of LCA and LPA, a multi-level nested framework for landscape character assessment is proposed as a reference for understanding the multi-scale local characteristics and spatial systems of landscape. By combining multi-scale segmentation and spatial clustering techniques of deep learning, a technical path for multi-level nested landscape character assessment is constructed as a new idea for characterizing the local characteristics of landscape at multiple scales.
[Conclusion] In the process of developing landscape character assessment systems, there have been numerous methodological systems. The multi-level nested assessment framework and technical path integrating LCA and LPA can accurately grasp the local characteristics of landscape through quantitative assessment and spatial mapping using artificial intelligence technology at different scales of practical needs. This framework serves as a comprehensive framework in the landscape character assessment system, providing a holistic perspective for exploring the unique value of landscape. In addition, combining the integrated framework with digital technology analysis, a landscape character assessment approach that adapts to multi-scale practical needs is proposed, providing a technical path for analyzing the local characteristics of landscape based on different levels of landscape space. The landscape spatial unit system can be used as a carrier to characterize the characteristics of different levels of landscape places, and the assessment results can be integrated with applications at different scales, thus assisting in the practice of landscape planning and design, zoning control, and resource protection and utilization in different fields. The integrated framework and technical path for LCA and LPA can help future landscape character assessment research comprehensively understand the past, present, and future of landscape, systematically understand the local characteristics of landscape from both the local and overall perspectives, and combine practical needs at different scales of results.

Cite this article

Yuncai WANG , Qizhen DONG . Integrated Framework and Technical Path for Multi-level Nested Assessment of Landscape Character[J]. Landscape Architecture, 2025 , 32(1) : 87 -95 . DOI: 10.3724/j.fjyl.202404130209

20世纪60年代,随着欧洲城市化进程的加快和环境保护主义的崛起,景观特征评价(landscape character assessment, LCA)应运而生,成为城市规划和相关决策的重要参考。1991年,在英国乡村委员会(Countryside Commission)的华威郡低地景观特征评估项目中,景观特征(landscape character, LC)这一概念首次正式获得应用,用以形容一个地区区别于其他地区的相对价值[1]。20世纪90年代,Lewis[2]在《设计明天:可持续的区域设计过程》(Tomorrow by Design: A Regional Design Process for Sustainability)一书中首次提出景观性格(landscape personality, LP)的概念,并使用景观性格评价(landscape personality assessment, LPA)结果指导美国伊利诺伊州地方规划设计和资源管控与发展,进一步丰富了LCA体系。
在LCA体系中,景观特点(landscape feature, LF)、景观特征、景观性格等概念反映了景观不同层次的地方特性。从定义来看,在英国乡村事务局(Countryside Agency)和苏格兰自然遗产协会(Scottish Natural Heritage)联合出版的《英格兰与苏格兰景观特征评价指导手册》(Landscape Character Assessment: Guidance for England and Scotland)中,景观特点指特别突出或者引人注目的景观要素[3]。景观特征指景观中具有独特性、可识别性和一致性的要素集合模式[3]。Lewis[2]用景观性格指代使景观区别于其他景观的颜色、质地、格局和空间特性。从研究内容来看,景观特点多用来指代有区分作用的景观元素或元素特征[4-5],如土地覆盖多样性、距停车场的距离等[6]。景观特征创造了不同地区的场所精神,帮助人们理解景观的现状和未来[1]。景观性格强调人们从多种景观特征中经提炼得到的整体认知结果。从景观特点、景观特征到景观性格是景观地方特性从局部特征要素识别走向整体特质综合提炼的过程(图1),三者表征了从整体到局部不同景观空间层级的景观地方特性,对景观性格进行分析有助于人们认知景观地方价值,对景观特征进行分类可为因地制宜地进行景观管理提供依据,对景观特点进行认知可为场地尺度的景观优化提升提供抓手。
图1 景观地方特性相关概念的多层级嵌套关系

Fig. 1 Multi-level nested relationship of concepts related to the local characteristics of landscape

LCA和LPA作为LCA体系中评价不同层级景观地方特性的方法,在价值内涵、景观空间载体、指标体系、实践应用等方面各有特点。LCA强调对景观现状的客观特征进行认知与表征,能够帮助人们理解景观现状、协助景观规划设计、辅助景观保护和管理。LCA强调景观客观物理属性,并将景观要素特征进行排列组合,其较强的可操作性和清晰的研究范式使该方法一经提出便受到广泛认可,并被广泛应用于国内外不同区域多种尺度的景观实践项目中[1, 5],可将该方法与不同前沿技术方法相结合以实现定量的空间测度[6]。相较于LCA,LPA更加关注景观发展演变过程中呈现出的整体性特质[7]。LPA基于整体景观空间的系统性、综合性分析,提炼景观的本源性特色,揭示了景观空间特征组合的空间效应[2, 7]。然而,LCA与LPA目前仍存在局限性:LCA缺少对景观变化历程和发展潜力的关注,尚未把人类感知直接纳入分类体系中[8];LPA的相关概念存在混淆,缺少量化分析的技术路径;受限于单元划分和分类分区的方式,LCA和LPA都存在评价结果难以与实践直接对接、不同尺度评价结果的嵌套关系不明确、定量测度结果不能直接形成分区等问题。在城乡建设过程中,景观地方特性亟须保护,不同景观要素需要得到系统性统筹。因此,有必要回顾LCA与LPA的发展历程,将二者整合并构建景观特征多级嵌套评价整合框架,结合前沿技术完善景观地方特性的定量评价路径,从而更好地满足不同尺度的实践需求,为人们全面客观地认知景观的过去、现在和未来提供系统性工具。基于此,本研究提出3个方面的研究内容:1)基于LCA和LPA的特点,分析LCA与LPA的互补性;2)构建LCA与LPA一体化的景观特征多级嵌套评价整合框架;3)提出景观特征多级嵌套评价技术路径。

1 LCA的特点:强调景观局部特征的客观认知与表征

LCA发展至今已经有了一定的研究与实践基础,形成了相对成熟的方法体系。目前相关研究聚焦于评价指标体系的构建[8-10]、评价结果的应用[11-16]等方面。在LCA的视角和体系中,价值维度是理解景观特征价值内涵的切入角度,指标体系为刻画景观特征提供了量化工具,而分类方法则决定了评价流程的科学性和结果应用的有效性。因此,本研究从4个方面对LCA研究与实践的现状与发展趋势进行评述。

1.1 LCA强调景观特征的客观认知

现有LCA研究强调对景观特征的客观认知,主观感知对景观特征分类只起到补充或修正的作用。从价值维度来看,景观的根本属性体现在主客体的相互作用之中[17]。LCA主要考虑自然物理、社会经济、人类美学和政治4个维度:英国LCA采用地形地貌、生境、人文3类因素;英国威尔士景观信息地图涉及自然、人文、公众认知3类属性[18];美国克利夫兰生态区涉及包含地表覆盖、行政管理、地方认同等维度[19]。早期LCA中主观感知情况的获取方式以问卷调查和访谈为主,结果以文本形式附加在对应景观特征类型上[20-22]。然而,该方式不适用于中大尺度的LCA[23],公众参与式地理信息系统(public participation geographic information system, PPGIS)提供了不同尺度下获取公众景观感知情况的途径[24-25],但PPGIS的参与者难以代表全部感知主体,且参与度甚至比传统问卷调查更低[26]
主观感知获取成本高、结果空间精度低,且时效性和准确性不足,难以和客观特征共同主导景观特征类型的划分,因此目前LCA主要围绕客观维度展开。随着大数据研究不断发展,已有学者在主观感知量化评价方面进行了一定的探索[27]。Koblet等[28]通过网络文本获取第一人称感知,将文本与文本描述所对应的景观特征区域以及单个景观要素相关联。兼顾客观认知和主观感知的LCA将会成为未来发展方向之一。

1.2 LCA指标体系结构扁平且基础指标冗余

扁平化指标体系的评价结果难以表征多维度的综合景观特征,动辄将十几个同维度的基础指标叠加,导致LCA结果往往存在图像分辨率较低、信息冗余等问题[29]。仅增加同维度的基础指标数量难以获得全面、具体的评价结果,还可能影响结果的可靠性和可读性。在物质环境维度的指标中,气候、土壤、基岩、地质、地形等数据源均具有较高空间分辨率,人类社会维度的指标中只有土地利用类型能作为空间制图的数据源。
在长期的研究与实践过程中,历史景观特征评估(historic landscape characterisation, HLC)、城镇景观特征评估、海洋景观特征评估等特定类型的LCA方式不断丰富了LCA体系。其中,最有影响力的是HLC,该方法基于历史地图、考古资料等数据[30],强调历史维度指标的主导性[31],在地方尺度上的应用具有灵活性和多样性[32],在国家和区域尺度上的应用弥补了LCA的缺陷[33]。未来,如何构建具有主导性的多层级评价指标体系将会成为LCA的重点研究方向。

1.3 LCA关注静态时间断面的景观特征

景观是长期人地作用动态演绎的结果,已有LCA研究关注特定时间断面下的景观空间特征[34-36],评价结果可为规划设计提供依据。如Martin等[37]的马德里高速公路景观特征研究为道路景观规划提供了直接指导,Shannon等[38]将视觉景观特征的评价结果作为圣劳伦斯河谷绿道规划的基础。还有一些LCA研究开始关注时间维度特征,包括在指标体系中体现景观的历史价值和发展潜力[39-40],用动态指标表征景观变化[27],以历史信息定义景观特征类型[40],对景观的长时序趋势进行探索[41-42]等,但这些方法尚未作为整体性视角的一部分融入LCA体系中。
总体而言,大量的LCA集中于静态时间断面的景观特征刻画,但已经有学者将动态视角引入LCA。在城市快速扩张的压力下,长时序动态视角下的LCA可以用于分析都市区景观的增长和变化模式[41, 43],帮助政府在可持续发展和生态保护方面做出更明智的决策[42-45],但仍未能综合景观的历史变化和资源潜力,对变化驱动力和发展趋势进行探索。早期的LCA只能对过去和现在的情况进行定性描述[46],如今数据源的更新效率不断提升、定量分析方法不断丰富,未来人工智能模型可以为景观特征定量分析和动态模拟提供支持[28]

1.4 LCA的分类方法存在不足且表征结果难以直接应用于实践

目前,LCA景观分类方法从原理上可分为自上而下和自下而上2种。传统叠图法以区域景观专题图作为数据源,基于ArcGIS平台叠加不同主题的图层,由专家学者基于专业知识和经验将具有相似特征的景观区域划为一类,如英国国家特征区、中国香港景观特征[47]、欧洲景观地图[48-49]等。这种方法的分类结果易被公众理解,但是主观性较强[9],受从业者价值观点制约较大[50]。随着科学技术的不断发展,多光谱图像分割技术和聚类算法等方法的使用提升了景观分类的科学性[51-53],但是目前仍存在分类结果图像分辨率低、不同数据源和算法产生的结果差异大、尺度嵌套不清等不足[32]。有学者综合使用多种聚类算法来提升结果的可靠性[54],如Lu等[29]使用K均值聚类算法(K-means clustering algorithm, K-means)、高斯混合模型(gaussian mixture model, GMM)和自组织映射(self-organizing map, SOM)3种算法建立了多聚类框架,并分析了街区尺度的城市景观特征(表1)。
表1 LCA景观分类方法[29]

Tab. 1 LCA landscape classification methods[29]

类型 方法 原理及特点 示意图
自上而下 叠图法 早期常用的景观分类方法,对不同类别的指标进行空间制图并赋权叠加后,对计算结果进行自上而下分类
双向指示种分析(two-way indicator species analysis, TWINSPAN) 遵循自上而下的分拆逻辑,从总体开始逐步一分为二,直到根据终止原则不能再分拆,广泛应用于植被群落研究中
自下而上 K-means 是一种基于距离的聚类算法,该算法设定好簇的个数K,随机创建K个质心,不断迭代收敛找出最优质心,将离质心最近的数据分配到对应簇中[29]
K中心聚类算法
(K-medoids)
类似K-means,两者的不同之处在于K-medoids选择数据集的观测值为质心,而不是随机生成的点,该算法适用于处理异常数据,更适合空间对象
SOM 是一种基于神经网络的聚类算法,依靠神经元之间的竞争学习机制适应输入数据的拓扑结构,从而将高维数据映射到低维空间中并进行可视化[29]
近邻传播算法
(affinity propagation, AP)
将全部数据点作为潜在聚类中心,通过计算数据点之间的相似度构建相似度矩阵,通过算法收敛得到有代表性的数据点作为簇中心
GMM 假设数据点服从多个高斯分布,对初始参数选择敏感,计算复杂度高[29]
LCA评价单元的选取方式限制了它在不同场景中的对接与应用[37-38]。基于行政单元的评价结果边界清晰且利于相关部门的权责划分,但与实体空间的景观地物要素无法形成对应关系[31]。基于流域等地理单元的评价结果能够承载丰富的生态过程信息,适用于大尺度自然生态区域而非城市区域[55]。基于网格单元的评价尺度灵活、操作简单,但单元边界没有实际意义,评价结果的精度受网格大小直接影响。对于中小尺度的景观特征研究而言[56],基于土地利用单元的评价结果边界明确且与实体空间对应,但斑块往往细碎零散,亟须进一步整合。未来,LCA评价单元的划分方式需要能够与不同尺度的景观实践应用相对接。

2 LPA的特点:提供景观整体性描述的视角和体系

LPA能够准确地把握景观的本底特质,其评价结果可以直接指导区域尺度的景观分区管控,作为资源多样性保护与利用策略提出的基础。然而,Lewis[2]提出的LPA未被广泛应用于研究与实践,尚未形成完善成熟的技术流程,目前仅有美国伊利诺伊州[2]和中国黑龙江伊春市[7]2个区域尺度的实证研究。因此,本研究结合上述2个实证研究对LPA体系进行分析,探究LPA如何对景观进行整体性描述,以及未得到广泛应用的原因。

2.1 LPA侧重于主观感知下的景观特质挖掘

LPA将主观感知应用于景观性格分类和结果描述环节,提炼景观的人格化属性,反映了景观形成中不可缺少的人地感应过程。在对伊春市景观性格分类的研究中,王云才等[7]将伊春市的空间划分为景观特质鲜明区、景观特质较突出区和景观特质平淡区。在对伊利诺伊州景观性格的描述中,Lewis[2]用“使人能够探索悬崖、深峡、山泉等崎岖地 貌”(represents as opportunity to study rugged terrain with limestone bluffs, deep ravines, springs)形容伊利诺伊州西北地区极具特色且引人入胜的景观性格。LPA通过人格化的表达直接表征对该区域景观的感知体验,帮助人们理解景观的本底价值。

2.2 LPA指标体系强调整体性和主导性

LPA指标体系强调景观表征维度的整体性和结构的主导性,采用多层次的表征体系和复合指标。Lewis[2]通过资源清单制图的方式,以自然资源丰富度、人类活动强度等作为定量表征景观性格的复合指标。王云才等[7]将景观特征要素总结为5类,构建了影响因子和特征要素二级表征结构,采用了生态敏感度、发展潜力度等复合指标。在后续的理论研究中,王云才等[57]从生态本底、景观风貌、人居建设和综合发展4个横向维度建立了表征体系,构建了景观要素、景观特征和景观性格三级纵向表征结构。

2.3 LPA考虑长时序下的人地作用综合效应

景观是一个整体概念,是人与环境的整体表征[58],它包含塑造土地的物质、自然和社会文化,以及人们与土地互动的方式。LPA关注景观的过程特征,在指标选取方面考虑人地作用的综合效应,对景观演变的历史过程和未来发展趋势进行刻画。Lewis[2]选取人类干扰强度、景观变化强度、人类活动潜在威胁程度等指标,王云才等[7]选取高质量空间资源分布、文化传说发源地、未来可开发建设用地分布等指标,这些指标的背后是人地作用过程的长时序积累和对未来景观发展演变的可能性预测。

2.4 分类方法限制了LPA体系的推广

LPA能够挖掘出空间资源禀赋和地方特色,可以应用于空间识别、保护与发展,并为制定相应的资源利用和空间发展方案提供依据。Lewis[2]基于LPA结果选择了详细设计的场地,制定了基础设施、空间模式和资源利用的发展策略。王云才等[7]基于LPA结果,针对不同分区提出了规划管控建议。然而在已有LPA研究中,定量的指标计算结果只能为景观性格分类、分区提供参考。景观性格分类、分区仍需要主观划定,分类、分区方法的科学性存在提升空间。此外,已有LPA研究采用网格单元作为评价单元,但单元边界与现实边界无法对应,不利于景观管理,限制了LPA体系在研究与实践过程中的广泛应用。

3 LCA与LPA一体化:景观特征多级嵌套评价整合框架与技术路径

3.1 LCA与LPA的互补性

比较LCA与LPA的方法体系,发现LPA和LCA二者在研究视角、指标体系、分类方法等方面存在一定互补性(图2)。LCA侧重于对要素及其组合层面特征的客观描述,能够刻画出景观局部差异性;关注资源组合动态过程中的综合效应,能够表征整体景观价值和人格化特征。景观性格由不同的景观特征高度凝练形成,具有易被人感知的内在属性,而LCA是LPA的基础。如何全面地表征景观独特价值并将评价结果更好地对接实践是LCA目前面临的挑战。LCA关注景观在某一状态下的静态特征,通过客观描述和罗列来刻画景观“是什么”,多应用于中小尺度的研究与实践;LPA则强调景观在发展演变中的动态特征,结合对景观性格的感知和概括来揭示景观“为什么”,有助于区域尺度上的景观独特价值认知。从评价结果来看,LCA在国家和区域尺度上常被用于区分大范围人地作用影响下的地质和地形,在地方尺度上用于区分更加细致的景观特征差异。LCA的命名往往采用特征编码与排列组合的方式,英格兰LCA将景观类型分为浅色山峰地区和深色山峰地区两大类,每类再依据地质、地形进行细分[1],该分类结果有助于规划和管理各尺度的景观和环境资源,指出了具有退化特征的景观区域,推动了英格兰国家森林计划和社区森林计划等项目的提出。在区域尺度上(图3),LPA通过要素在人地作用和人地感应方面的整体效应建立景观性格分区,LPA的命名方式融入了人们对该区域的感知体验,如Lewis[2]在伊利诺伊州的景观性格分类中用“草原核心滋养哺育众多河流”(numerous rivers draining the prairie core)“多元丰富的性格”(the rich features)等拟人化的用语来描述不同景观的性格类型并识别其地方价值[2],分类、分区结果与资源清单可以共同为设计地点的选择提供依据,在区域规划设计过程中可以为具体发展方案的制定提供参考 。
图2 LCA与LPA方法体系及特点

Fig. 2 Methodological systems and their respective characteristics of LCA and LPA

图3 美国伊利诺伊州LPA实践案例评价结果[2]

Fig. 3 Evaluation result of LPA practice case in Illinois, USA[2]

3.2 景观特征多级嵌套评价整合框架

已经有学者提出将不同评价方法相结合对已有LCA体系进行优化[8, 33],如Zhao等[59]将西方的LCA和HLC相结合,基于中国山水哲学提出民族景观遗产的山水特征识别路径。LCA和LPA作为LCA体系中2种不同景观地方特性评价的方法,对于全面理解景观的价值和发展潜力、制定空间管控策略、景观规划设计等方面都具有重要意义。基于LCA与LPA的互补性与整体性视角,本研究构建了景观特征多级嵌套评价整合框架(图4),为理解相关概念体系内部的层级逻辑和体系之间的对应关系提供参考,为后续景观特性的多级表征提供参考。首先,就景观的本质内涵而言,人类社会与自然环境的长期作用形成了整体景观,深入理解景观内涵可以为挖掘景观地方特性的价值提供理论框架。其次,从景观特征的形成逻辑来看,整体景观承载着整体人文生态系统,整体人文生态系统的结构性和解构性决定了景观空间是由多个相对独立的空间构成的复合体;拼接性和嵌套性决定了整体景观空间的多级嵌套性;感知性与表意性决定了景观地方特性识别是认知景观独特价值的途径[60],这为景观地方特性的量化评价提供了理论依据。最后,景观地方特性以景观空间为载体,景观地方特性的多级表征与景观空间单元的多级嵌套关系相对应,为评价单元的选取和评价结果的实践应用提供了可操作的空间载体。
图4 景观特征多级嵌套评价整合框架

Fig. 4 Integrated framework for multi-level nested assessment of landscape character

景观特征多级嵌套评价整合框架揭示了景观地方特性、景观空间载体和实践应用方面存在的层级间的嵌套关系。景观要素是景观空间最基础的构成,为景观特点提供了载体;景观基本空间单元是在构成与形态方面与邻近空间具有分异性的最小单元,为景观特征的精确表征提供了载体;具有形态、功能、界面等相似特征的基本空间单元组合形成复合空间单元,景观性格揭示了景观的整体性特征,反映了局部特征复合形成的综合效应,复合空间单元为景观性格提供了空间载体[61]。以景观空间单元体系为载体可以表征不同层级的景观地方特性,使评价结果能够对接不同尺度的实践,助力景观规划设计、分区管控、资源保护及利用等。

3.3 景观特征多级嵌套评价技术路径

以景观特征多级嵌套评价整合框架为基础,本研究结合前沿方法技术,制定了一套灵活且可操作的技术路径(图5),明确各环节采用的方法,形成标准化流程,以指导实际景观特征评价的开展。
图5 景观特征多级嵌套评价技术路径

Fig. 5 Technical path for multi-level nested assessment of landscape character

1)结合具体的研究问题和现实需要,确定研究目的,包括宏观层面上对景观性格进行认知、中观层面上对景观特征进行描述、微观层面上对场地景观特点进行刻画。
2)构建包括生态环境、社会经济、文化美学、遥感卫星影像不同数据类型在内的数据库。
3)结合景观特征多级嵌套评价整合框架,划定不同尺度的评价单元。基于深度学习中的U形卷积网络(U-net convolutional network, U-Net)与双边门控空洞卷积网络(bi-directional cascade network, BDCN)模型对景观要素与单元边界进行图像识别,能够实现景观要素与景观基本单元的划分[61]。语义分割是深度学习中常用的高分辨率影像解译预处理方法,能够将景观基本单元作为基础栅格,结合不同指标计算结果划定景观复合单元。具体过程为:①根据评价对象特征和专家打分结果选取波段类型、确定波段权重;②根据相应尺度下景观指数随粒度大小变化的特征,推断不同尺度景观研究的适宜粒度[62],从而确定最优分割尺度;③以异质性最小为原则,基于eCognition Developer平台对评价单元进行分级[52]
4)在评价指标体系构建与指标计算方面,多级嵌套评价表征体系是对景观地方特性内涵的结构化组织。对于景观特点,基于同种景观要素的差异性和要素与周边环境的关系提取基础指标[63];对于景观特征,从要素、格局、过程、感知等角度选取基础指标并形成复合指数;对于景观性格,依托整体景观的形成逻辑确定表征维度,选取综合性指数作为目标层,结合景观特征和景观特点的表征,选取准则层的复合指数和方案层的基础指标。
5)基于R语言平台,运用SOM算法对指标分类结果初次聚类,再运用K-means二次聚类,得到景观分类结果。SOM与K-means存在互补性[64],SOM算法不需要预先设定聚类类别数量,自主性强但结果具有一定模糊性;K-means计算效率高,但需要预先确定聚类类别数量和聚类中心。二者相结合的算法具有自主性强且运行效率高的特点。
景观特征多级嵌套评价技术路径展现了整合框架的科学性和可行性,全面地考虑了人地关系,鲜明地表征了景观地方特性,能够满足不同尺度的研究需要。多尺度分割得到的评价单元具有景观空间的嵌套性特征,使评价结果可以更好地对接实践。多层次的指标体系避免了扁平化的指标体系导致的评价结果分辨率低,提升了结果的可读性;分类方法综合了不同聚类算法,增强了结果的稳定性和可靠性。

4 结论

在长期的研究与实践过程中,LCA体系在景观变化管理、社区服务、文化教育等领域得到了广泛应用[16]。本研究从价值判断、指标体系、时序特征、分类方法和实践应用等方面对LCA与LPA进行评述,主要得到3点结论。
1)LCA强调对景观要素特征的客观认知,能够识别要素的组合模式并进行罗列表征,方法具有普适性,被广泛应用于不同地区。目前LCA存在主观感知分析结果难以直接指导分类、指标体系结构扁平且指标冗余、只能刻画静态时间断面的特征、分类方法难以重复、分类结果难以对接实践等方面的不足。
2)LPA强调对整体性特征感知的提炼,能够识别不同人地作用下景观的空间异质性和本底属性,多层次的指标体系具有主导性且在指标选取过程中考虑了景观演进的长时序过程特征。目前LPA存在只能在较大尺度实践中识别出空间异质性,指标定量测度结果与景观性格类型划定脱节等方面的不足。
3)LCA能够描述景观“是什么”,LPA 能够探究“为什么”,二者存在一定互补性,是LCA体系中2套适用于不同景观地方特性的评价方法,二者的整合有助于LCA体系在理论和实践方面的优化与提升。
本研究整合了LCA和LPA的视角和方法,基于景观地方特性表征体系和景观空间单元体系的多级嵌套关系,建立了景观特征多级嵌套评价整合框架。该框架作为LCA体系中的综合性框架,为景观独特价值的挖掘提供了整体性视角。此外,本研究将整合框架与数字化技术相结合,提出了适应不同尺度实践需求的LCA思路,为不同层级景观空间的景观地方特性分析提供了技术路径。LCA与LPA一体化景观特征多级嵌套评价整合框架和技术路径能够帮助人们全面地理解景观的过去、现在和未来,可以为从局部到整体对景观地方特性进行系统性理解提供方法,为不同尺度下的景观实践需求提供直接参考。

文中图表均由作者绘制,其中图3由作者根据参考文献[2]绘制,表1由作者根据参考文献[29]绘制。

[1]
司万维克, 高枫, 邓位. 英国景观特征评估[J]. 世界建筑, 2006 7): 23-27

SWANWICK C, GAO F, DENG W. Landscape Character Assessment in Britain[J]. World Architecture, 2006 7): 23-27.

[2]
LEWIS P H. Tomorrow by Design: A Regional Design Process for Sustainability[M]. New York: John Wiley&Sons Ltd, 1996.

[3]
SWANWICK C. Landscape Character Assessment: Guidance for England and Scotland[R]. Edinburgh: The Countryside Agency, 2002.

[4]
EVEILLARD-BUCHOUX M, BENINGER P G, CHADENAS C, et al. Small-Scale Natural Landscape Features and Seabird Nesting Sites: The Importance of Geodiversity for Conservation[J]. Landscape Ecology, 2019, 34(10): 2295-2306.

[5]
ISSAYEVA A, YESHIBAYEV A, LESKA B, et al. Comparative Assessment of Geomorphological and Landscape Character Around the Small Aral Sea[J]. Journal of Ecological Engineering, 2021, 22(10): 73-84

DOI

[6]
SCHMIDT K, MARTÍN-LÓPEZ B, PHILLIPS P M, et al. Key Landscape Features in the Provision of Ecosystem Services: Insights for Management[J]. Land Use Policy, 2019, 82 353-366

DOI

[7]
王云才, 马玥莹, 申佳可. 景观性格评价在国土空间规划和管控中的应用[J]. 风景园林, 2020, 27(1): 35-40

WANG Y C, MA Y Y, SHEN J K. Application of Landscape Personality Assessment in National Territory Spatial Planning and Control[J]. Landscape Architecture, 2020, 27(1): 35-40.

[8]
FAIRCLOUGH G, HERRING P. Lens, Mirror, Window: Interactions Between Historic Landscape Characterisation and Landscape Character Assessment[J]. Landscape Research, 2016, 41(2): 186-198

DOI

[9]
BUTLER A. Dynamics of Integrating Landscape Values in Landscape Character Assessment: The Hidden Dominance of the Objective Outsider[J]. Landscape Research, 2016, 41(2): 239-252

DOI

[10]
GALIANA-MARTIN L, HERRERO G, SOLANA J. A Wildland-Urban Interface Typology for Forest Fire Risk Management in Mediterranean Areas[J]. Landscape Research, 2011, 36(2): 151-171

DOI

[11]
BUTLER A, BERGLUND U. Landscape Character Assessment as an Approach to Understanding Public Interests Within the European Landscape Convention[J]. Landscape Research, 2014, 39(3): 219-236

DOI

[12]
ATIK M, IŞIKLI R C, ORTAÇEŞME V, et al. Definition of Landscape Character Areas and Types in Side Region, Antalya-Turkey with Regard to Land Use Planning[J]. Land Use Policy, 2015, 44 90-100.

[13]
CASPERSEN O H. Public Participation in Strengthening Cultural Heritage: The Role of Landscape Character Assessment in Denmark[J]. Geografisk Tidsskrift-Danish Journal of Geography, 2009, 109(1): 33-45

DOI

[14]
KIM K H, PAULEIT S. Landscape Character, Biodiversity and Land Use Planning: The Case of Kwangju City Region, South Korea[J]. Land Use Policy, 2007, 24(1): 264-274

DOI

[15]
BARTLETT D, GOMEZ-MARTIN E, MILLIKEN S, et al. Introducing Landscape Character Assessment and the Ecosystem Service Approach to India: A Case Study[J]. Landscape and Urban Planning, 2017, 167 257-266

DOI

[16]
张慧, 张媛. 景观特征评估(LCA)的应用价值[J]. 中国城市林业, 2018, 16(2): 74-79

DOI

ZHANG H, ZHANG Y. Application Value of Landscape Character Assessment[J]. Journal of Chinese Urban Forestry, 2018, 16(2): 74-79

DOI

[17]
黄昕珮. 论“景观”的本质: 从概念分裂到内涵统一[J]. 中国园林, 2009, 25(4): 26-29

DOI

HUANG X P. Discussion on the Essence of Landscape: From Disrupt Concepts to Unified Connotation[J]. Chinese Landscape Architecture, 2009, 25(4): 26-29

DOI

[18]
宋峰, 宋蕾蕾. 英美景观评估方法评述及借鉴[J]. 开发研究, 2016 5): 69-74

SONG F, SONG L L. Review and Reference of Landscape Assessment Methods in the UK and the US[J]. Research on Development, 2016 5): 69-74.

[19]
BEACH D, FLAMM B. Citizens Bioregional Plan for Northeast Ohio[J]. EcoCity Cleveland, 1999, 6 4-6.

[20]
GKOLTSIOU A, PARASKEVOPOULOU A. Landscape Character Assessment, Perception Surveys of Stakeholders and SWOT Analysis: A Holistic Approach to Historical Public Park Management[J]. Journal of Outdoor Recreation and Tourism, 2021, 35 100418

DOI

[21]
WARNOCK S, GRIFFITHS G. Landscape Characterisation: The Living Landscapes Approach in the UK[J]. Landscape Research, 2015, 40(3): 261-278

DOI

[22]
ATIK M, IŞIKLI R C, ORTAÇEŞME V, et al. Exploring a Combination of Objective and Subjective Assessment in Landscape Classification: Side Case from Turkey[J]. Applied Geography, 2017, 83 130-140

DOI

[23]
TERKENLI T, GKOLTSIOU A, KAVROUDAKIS D. The Interplay of Objectivity and Subjectivity in Landscape Character Assessment: Qualitative and Quantitative Approaches and Challenges[J]. Land, 2021, 10(1): 53

[24]
BROWN G, BRABYN L. An Analysis of the Relationships Between Multiple Values and Physical Landscapes at a Regional Scale Using Public Participation GIS and Landscape Character Classification[J]. Landscape and Urban Planning, 2012, 107(3): 317-331

DOI

[25]
BROWN G, BRABYN L. The Extrapolation of Social Landscape Values to a National Level in New Zealand Using Landscape Character Classification[J]. Applied Geography, 2012, 35(1/2): 84-94

DOI

[26]
GREG B, JESSICA M M, KATIE L. Public Participation GIS: A Method for Identifying Ecosystem Services[J]. Society & Natural Resources, 2012, 25(7): 633-651.

[27]
MARTÍN B, ORTEGA E, MARTINO P, et al. Inferring Landscape Change from Differences in Landscape Character Between the Current and a Reference Situation[J]. Ecological Indicators, 2018, 90 584-593

DOI

[28]
KOBLET O, PURVES R S. From Online Texts to Landscape Character Assessment: Collecting and Analysing First-Person Landscape Perception Computationally[J]. Landscape and Urban Planning, 2020, 197 103757

[29]
LU Y, XU S, LIU S, et al. An Approach to Urban Landscape Character Assessment: Linking Urban Big Data and Machine Learning[J]. Sustainable Cities and Society, 2022, 83 103983

DOI

[30]
李和平, 杨宁. 城市历史景观的管理工具: 城镇历史景观特征评估方法研究[J]. 中国园林, 2019, 35(5): 54-58

LI H P, YANG N. Research on the Management Tools of Historic Urban Landscape: Historic Townscape Characterisation[J]. Chinese Landscape Architecture, 2019, 35(5): 54-58.

[31]
王志芳, 周瑶瑾, 徐敏, 等. 县域景观特征管理单元划分方法: 以武胜县为例[J]. 北京大学学报(自然科学版), 2020, 56(3): 553-560

WANG Z F, ZHOU Y J, XU M, et al. Classification of County Natural Landscape Governance Units for Wusheng[J]. Acta Scientiarum Naturalium Universitatis Pekinensis, 2020, 56(3): 553-560.

[32]
周剑云, 蒋定哲, 鲍梓婷, 等. 历史景观特征评估: 链接保护区与城市历史景观[J]. 国际城市规划, 2023, 38(2): 124-132

ZHOU J Y, JIANG D Z, BAO Z T, et al. Historic Landscape Characterisation: Linking “Conservation Areas” and “Historic Urban Landscape”[J]. Urban Planning International, 2023, 38(2): 124-132.

[33]
汪伦, 张斌. 景观特征评估: LCA体系与HLC体系比较研究与启示[J]. 风景园林, 2018, 25(5): 87-92

WANG L, ZHANG B. Landscape Feature Assessment: Comparative Study and Inspiration Between LCA System and HLC System[J]. Landscape Architecture, 2018, 25(5): 87-92.

[34]
VAN E V, ANTROP M. A Stepwise Multi-scaled Landscape Typology and Characterisation for Trans-regional Integration, Applied on the Federal State of Belgium[J]. Landscape and Urban Planning, 2009, 91(3): 160-170.

[35]
YANG D C, GAO C, LI L Y, et al. Multi-scaled Identification of Landscape Character Types and Areas in Lushan National Park and Its Fringes, China[J]. Landscape and Urban Planning, 2020, 201 103844

DOI

[36]
KONKOLY-GYURÓ É. Conceptualisation and Perception of the Landscape and Its Changes in a Transboundary Area. A Case Study of the Southern German-French Borderland[J]. Land Use Policy, 2018, 79 556-574

DOI

[37]
MARTÍN B, ORTEGA E, OTERO I, et al. Landscape Character Assessment with GIS Using Map-Based Indicators and Photographs in the Relationship Between Landscape and Roads[J]. Journal of Environmental Management, 2016, 180 324-334

DOI

[38]
SHANNON S, SMARDON R, KNUDSON M. Using Visual Assessment as a Foundation for Greenway Planning in the St. Lawrence River Valley[J]. Landscape and Urban Planning, 1995, 33(1-3): 357-371

DOI

[39]
HERRING P C. Framing Perceptions of the Historic Landscape: Historic Landscape Characterisation (HLC) and Historic Land-Use Assessment (HLA)[J]. Scottish Geographical Journal, 2009, 125 61-77

DOI

[40]
TURNER S. Historic Landscape Characterisation: A Landscape Archaeology for Research, Management and planning[J]. Landscape Research, 2006, 31(4): 385-398.

[41]
TORREGGIANI D, LUDWICZAK Z, DALL’ARA E, et al. TRuLAn: A High-Resolution Method for Multi-time Analysis of Traditional Rural Landscapes and Its Application in Emilia-Romagna, Italy[J]. Landscape and Urban Planning, 2014, 124 93-103

DOI

[42]
MOU J, CHEN Z, HUANG J. Predicting Urban Expansion to Assess the Change of Landscape Character Types and Its Driving Factors in the Mountain City[J]. Land, 2023, 12(4): 928

DOI

[43]
CATALÁN B, SAURÍ D, SERRA P. Urban Sprawl in the Mediterranean?: Patterns of Growth and Change in the Barcelona Metropolitan Region 1993–2000[J]. Landscape and Urban Planning, 2008, 85(3-4): 174-184

DOI

[44]
MEDEIROS A, FERNANDES C, GONÇALVES J F, et al. A Diagnostic Framework for Assessing Land-Use Change Impacts on Landscape Pattern and Character: A Case-Study from the Douro Region, Portugal[J]. Landscape and Urban Planning, 2022, 228 104580

DOI

[45]
VAN EETVELDE V, ANTROP M. Indicators for Assessing Changing Landscape Character of Cultural Landscapes in Flanders (Belgium)[J]. Land Use Policy, 2009, 26(4): 901-910

DOI

[46]
BASTIAN O, RÖDER M. Assessment of Landscape Change by Land Evaluation of Past and Present Situation[J]. Landscape and Urban Planning, 1998, 41(3-4): 171-182.

[47]
鲍梓婷, 周剑云. 香港景观特征评估(LCA)的实践与经验[J]. 中国园林, 2015, 31(9): 100-104

DOI

BAO Z T, ZHOU J Y. The Practice and Experience of Landscape Character Assessment in Hong Kong[J]. Chinese Landscape Architecture, 2015, 31(9): 100-104

DOI

[48]
MÜCHER C A, KLIJN J A, WASCHER D M, et al. A New European Landscape Classification (LANMAP): A Transparent, Flexible and User-Oriented Methodology to Distinguish Landscapes[J]. Ecological Indicators, 2010, 10(1): 87-103.

[49]
CHUMAN T, ROMPORTL D. Multivariate Classification Analysis of Cultural Landscapes: An Example from the Czech Republic[J]. Landscape and Urban Planning, 2010, 98(3/4): 200-209.

[50]
SIMENSEN T, HALVORSEN R, ERIKSTAD L. Methods for Landscape Characterisation and Mapping: A Systematic Review[J]. Land Use Policy, 2018, 75 557-569.

[51]
刘文平, 宇振荣. 北京市海淀区景观特征类型识别及评价[J]. 生态学杂志, 2016, 35(5): 1338-1344

LIU W P, YU Z R. Identification and Assessment of Landscape Character of Haidian District, Beijing[J]. Chinese Journal of Ecology, 2016, 35(5): 1338-1344.

[52]
盛硕, 王云才. 基于生态特征与问题聚类识别的平原农业区乡村生态修复策略: 以辽宁省黑山县为例[J]. 中国园林, 2022, 38(12): 20-25

SHENG S, WANG Y C. Rural Ecological Restoration Strategies in Plain Agricultural Areas Based on the Identification of Ecological Characteristics and Issues Through Clustering: A Case Study of Heishan County, Liaoning Province[J]. Chinese Landscape Architecture, 2022, 38(12): 20-25.

[53]
高飞, 郭沁, 段广德. 蒙东农牧交错区乡村聚落文化景观区划研究[J]. 城市建筑, 2023, 20(5): 27-29

GAO F, GUO Q, DUAN G D. Research on the Zoning of Cultural Landscape of Rural Settlements in the Agro-Pastoral Area of Eastern Inner Mongolia[J]. Urbanism and Architecture, 2023, 20(5): 27-29.

[54]
李哲, 卢馨逸, 施佳颖, 等. 基于小圩形态指数的宣芜平原圩田景观肌理量化研究: 以固城湖永丰圩为例[J]. 中国园林, 2023, 39(1): 41-46

LI Z, LU X Y, SHI J Y, et al. Quantitative Research of Landscape Texture for Polders in Xuanwu Plain Based on Morphological Index: A Case Study of Yongfeng Polder, Gucheng Lake[J]. Chinese Landscape Architecture, 2023, 39(1): 41-46.

[55]
王金亮, 谢德体, 倪九派, 等. 基于源汇景观单元的流域土壤侵蚀风险格局识别[J]. 生态学报, 2017, 37(24): 8216-8226

WANG J L, XIE D T, NI J P, et al. Identification of Soil Erosion Risk Patterns in a Watershed Based on Source-sink Landscape Units[J]. Acta Ecologica Sinica, 2017, 37(24): 8216-8226.

[56]
陈慕婷, 周剑云, 鲍梓婷, 等. 省域湿地退化风险评价研究: 基于地理格网与景观特征单元的多尺度分析[J]. 中国园林, 2023, 39(4): 52-58

CHEN M T, ZHOU J Y, BAO Z T, et al. Study on Risk Assessment of Provincial Wetland Degradation-Based on Multi-scale Analysis of Geographical Grids and Landscape Character Units[J]. Chinese Landscape Architecture, 2023, 39(4): 52-58.

[57]
王云才, 陈照方, 成玉宁. 新时期乡村景观特征与景观性格的表征体系构建[J]. 风景园林, 2021, 28(7): 107-113

WANG Y C, CHEN Z F, CHENG Y N. The Representation System of Rural Landscape Character and Personality in the New Era[J]. Landscape Architecture, 2021, 28(7): 107-113.

[58]
鲍梓婷, 周剑云. 英国景观特征评估概述: 管理景观变化的新工具[J]. 中国园林, 2015 3): 46-50

BAO Z T, ZHOU J Y. Landscape Character Assessment in UK: A New Tool to Manage Landscape Change[J]. Chinese Landscape Architecture, 2015 3): 46-50.

[59]
ZHAO Y, HARVEY D C, GAO C. Identifying Shan-Shui Characteristics for National Landscape Heritage: Reconciling Western and Chinese Landscape Characterisation from a Trans-cultural Perspective[J]. The Geographical Journal, 2020, 186(3): 300-313

DOI

[60]
王云才. 基于空间生态特性的景观图式语言研究方法与方法论[J]. 风景园林, 2018, 25(1): 28-32

WANG Y C. The Research Method and Methodology of Pattern Language Based on Ecological Nature of Landscape Space[J]. Landscape Architecture, 2018, 25(1): 28-32.

[61]
陈照方, 王云才. 乡村景观单元识别与认知的空间生态智慧: 以南京市牌坊社区为例[J]. 风景园林, 2022, 29(7): 30-36

CHEN Z F, WANG Y C. Spatial Ecological Wisdom for Identification and Cognition of Rural Landscape Units: A Case Study of Nanjing Paifang Rural Community[J]. Landscape Architecture, 2022, 29(7): 30-36.

[62]
华琳, 黄志霖, 马良, 等. 三峡库区低山丘陵区多尺度景观指数响应及适宜粒度[J]. 生态学报, 2022, 42(11): 4703-4717

HUA L, HUANG Z L, MA L, et al. Suitable Granularity and Response of Multi-scale Landscape in Low Mountain and Hilly Area of the Three Gorges Reservoir[J]. Acta Ecologica Sinica, 2022, 42(11): 4703-4717.

[63]
华琳, 黄志霖, 马良, 等. 三峡库区低山丘陵区多尺度景观指数响应及适宜粒度[J]. 生态学报, 2022, 42(11): 4703-4717

HUA L, HUANG Z L, MA L, et al. Suitable Granularity and Response of Multi-scale Landscape in Low Mountain and Hilly Area of the Three Gorges Reservoir[J]. Acta Ecologica Sinica, 2022, 42(11): 4703-4717.

[64]
朱建宁. 展现地域自然景观特征的风景园林文化[J]. 中国园林, 2011, 27(11): 1-4

DOI

ZHU J N. Landscape Culture Reproducing Regional Natural Landscape Characteristics[J]. Chinese Landscape Architecture, 2011, 27(11): 1-4

DOI

Outlines

/