Location, Time, and Scale in GIS-Based Multi-Criteria Decision Analysis - Advancing the Agenda

基于地理信息系统的多准则决策分析(MCDA)被用于环境、交通和土地利用规划、废物管理、水文、农业和林业等领域的决策。然而，关于地理信息系统-MCDA的空间、时间和尺度组成部分的重要概念性问题在很大程度上仍然没有得到回答。拟议的研究计划旨在扩大我目前由NSERC资助的这一领域的研究。在GIS-MCDA中，输入数据的位置分量往往没有得到充分利用。在最近的工作中，一些研究人员试验了空间显式MCDA技术，如局部加权线性组合(LWLC)。LWLC技术通过局部权重表示标准结果的局部差异。因此，LWLC创建了一个更加分散的MCDA分数模式，它突出了局部差异，而不是一般的空间模式。这一方法的有效性仍有待在实际应用中加以证明，这是本提案的目标之一。地理信息系统MCDA中的时间部分指的是随着时间的推移而重复进行的评价。由于MCDA分数是无单位的，不能直接进行比较，当综合指数被用来跟踪一段时间以来的社会经济或环境系统表现时，就会出现问题。在最近由NSERC资助的工作中，我提出了“联合标准化”，使不同时间点的分数具有可比性。与全局加权与局部加权类似，联合标准化可以与全局(时间)方法相关联，而单独(时间特定的)标准化代表局部方法。因此，单独的标准化产生更大的结果差异，而联合标准化产生关于一般时间模式的信息。这两种方法将在本研究中得到进一步检验，并探讨它们在实际应用中的有效性。最后，尽管本研究的重点将放在空间尺度上，但GIS-MCDA中的尺度成分既涉及空间要素又涉及时间要素。比例尺与地理空间分析中众所周知的可修改面积单位问题(MAUP)有关。MAUP的尺度效应和区划效应都会影响到GIS-MCDA的结果。在不同的分析尺度上进行地理信息系统-MCDA案例研究，并使用不同的分区配置，将有助于更好地了解这些影响。拟议的研究旨在为决策者提供严格的指导方针，使分析规模与行动规模保持一致。总体而言，拟议的研究旨在探索影响GIS-MCDA结果有效性的基本问题。将探讨这些影响的范围，并制定处理地理信息系统-MCDA的地点、时间和比例组成部分的最佳做法。最终，这项研究计划旨在为空间和时间上明确的、特定规模的决策支持开发一个概念性框架。

GIS-based multi-criteria decision analysis (MCDA) is used for decision-making in fields such as environmental, transportation, and landuse planning, waste management, hydrology, agriculture, and forestry. Yet important conceptual questions regarding the spatial, temporal, and scale components of GIS-MCDA remain largely unanswered. The proposed research program aims to expand my current NSERC-funded research in this field. The location component of input data in GIS-MCDA is often under-used. In recent work, some researchers have experimented with spatially explicit MCDA techniques such as the locally weighted linear combination (LWLC). The LWLC technique represents local variations in criterion outcomes through local weights. As a result, LWLC creates a more dispersed pattern of MCDA scores, which highlights local differences rather than general spatial pattern. The usefulness of this approach remains to be demonstrated in practical applications, which is one objective of this proposal. The time component in GIS-MCDA refers to repeated evaluations over time. Since MCDA scores are unitless and not directly comparable, issues arise when composite indices are used for tracking socio-economic or environmental system performance over time. In recent NSERC-funded work, I proposed "joint standardization" to make scores comparable when they are calculated for different points in time. Similar to global vs. local weighting, joint standardization can be associated with a global (in time) approach, while individual (time-specific) standardization represents a local approach. Consequently, individual standardization yields greater variation in results while joint standardization yields information about general temporal pattern. These two approaches will be further examined in this research, and their usefulness explored in practical applications. Finally, the scale component in GIS-MCDA involves both spatial and temporal elements, although the focus of this research will be on spatial scales. Scale is connected to the well-known Modifiable Areal Unit Problem (MAUP) in geospatial analysis. Both the scale and zoning effects of the MAUP influence the results of GIS-MCDA. Conducting GIS-MCDA case studies at different scales of analysis and using different zoning configurations will contribute to a better understanding of these effects. The proposed research aims to give decision-makers stringent guidelines for aligning the scale of analysis with the scale of action. Overall, the proposed research serves to explore fundamental questions that impact the validity of GIS-MCDA results. The scope of these effects will be explored and best-practice for handling the location, time, and scale components of GIS-MCDA developed. Ultimately, this research program aims to develop a conceptual framework for spatially and temporally explicit, and scale-specific, decision support.