Review and Comparative Study of Decision Support Tools for the Mitigation of Urban Heat Stress

: Over the last few decades, Urban Heat Stress (UHS) has become a crucial concern of scientists and policy-makers. Many projects have been implemented to mitigate Urban Heat Island (UHI) effects using nature-based solutions. However, decision-making and selecting an adequate framework are difﬁcult because of complex interactions between natural, social, economic and built environments. This paper contributes to the UHI issue by: (i) identifying the most important key factors of a Decision Support Tool (DST) used for urban heat mitigation, (ii) presenting multi-criteria methods applied to urban heat resilience, (iii) reviewing existing spatial and non-spatial DSTs, (iv) and analyzing, classifying and ranking DSTs. It aims to help decision-makers through an overview of the pros and cons of existing DSTs and indicate which tool is providing maximum support for choosing and planning heat resilience measures from the designing phase to the heat mitigation phase. This review shows that Multi-Criteria Decision Analysis (MCDA) can be used for any pilot site and the criteria can be adapted to the given location accordingly. It also highlights that GIS-based spatial tools have an effective decision support system (DSS) because they offer a quick assessment of interventions and predict long-term effects of urban heat. Through a comparative study using speciﬁc chosen criteria, we conclude that the DSS tool is well suited and fulﬁls many prerequisites to support new policies and interventions to mitigate UHS.


Introduction
Urbanization and an exponential increase in population have brought the concept of Urban Heat Island (UHI) and heat stress into the limelight. The world has seen adverse effects, particularly a rise in air temperature, a higher mortality rate, and changes in weather patterns [1]. Most studies have focused on the UHI in densely populated capital cities and there is insufficient literature available for smaller cities [2]. Different authors explained that UHI has severe effects on the most vulnerable populations, especially during the summer season. This phenomenon indeed highly raises the consumption of cooling energy as well as the corresponding peak electricity demand of cities. Therefore, the UHI can be linked with a significant increase in urban pollutant concentrations and is concerned with the city's carbon footprint as well as ground-level ozone. Urban Heat Stress (UHS) severely affects health, comfort, and increases mortality problems [3]. In current times, urban planners and policy-makers are keen to address issues such as increased urban heat due to climate change triggered by human activities.
Europe, Australia and North America are the major continents among those working to mitigate UHS in different ways, for example, by increasing urban forestry or by using green and blue interventions. On the other hand, Asia has also worked on thermal comfort but their focal point is grey and blue infrastructure. Accommodating heat stress measures in urban areas is not the easiest task as it encounters issues such as water scarcity, high cost and unsuitable environments for green infrastructure.
It is the responsibility of the decision-makers to evaluate multiple possible solutions to resolve the issues by considering specific criteria. Urban planners are still perplexed due to the severity of changes that have taken place in different zones.
Sometimes, alternative decisions are to be taken in order to combat the complex situation by considering some criteria [4]. It is observed in previous studies that every location has unique characteristics and parameters and the decision-makers have concerns about criteria such as cost, efficiency, and materials. For every location change, the mitigation measure should be modified. To solve these issues, a proper decision support system (DSS) is required to help decision-makers.
A DSS is an information system that requires judgment, determination, and a sequence of actions. It assists the mid-and high-level management of an organization by analyzing huge volumes of unstructured data and information. It is either human-powered, automated or a combination of both and it can be used in any domain due to its versatility.
There are many studies related to climate vulnerabilities-some have used economic or mechanistic modeling [5][6][7] and other researchers have used outranking approaches that later have been criticized due to axiomatic violations [8,9]. The number of characteristics required for the evaluation of UHS management similarly challenges is not constant. There is a need for a tool that allows one to work in consideration of all parameters simultaneously and helps to identify negative trends of urban heat and eventually allow better adaptation measures.
This paper presents a comprehensive review of DSTs in the essence of UHI, climate change adaptation, and heat stress. In Section 2, the methodology of the paper is discussed, Multi-Criteria Decision Analysis (MCDA) approaches are reviewed in Section 3, and DSTs (toolkits and spatial tools) are discussed in Section 4 of the research paper. All tools are critically analyzed by 15 important criteria in Section 5 and, finally, the conclusion is presented in Section 6.

Review Strategy
In this review article, we have used a qualitative and exploratory approach. Peerreviewed research papers were gathered from Google Scholar. The research papers were selected by using keywords such as multi-criteria decision, UHI mitigation, heat resilience and UHS DST. Tools that are developed for urban heat resilience under the banner of different projects were searched by using the same keywords. The survey is presented in two tables. In the first table, we reviewed 9 academic studies in which different MCDA approaches were applied for developing DSTs for UHS mitigation. In the second table, we performed a review on different DSTs which deal with the UHI, climate change risks, extreme heat events, heat resilience adaptation and mitigation measures.

Inclusion and Exclusion Criteria
In this paper, we analyze 12 DSTs with the principle aspects to analyze the support of the decision-making tool, such as: (i) experts' assistance in the development of a support system; (ii) social culture factors, for example, number of population and their age, their activities, health data and the local environment; (iii) adaptive capacity of the tool which allows the indication of the suspect areas, informs where intervention is needed and when to schedule outdoor activities; (iv) good integration with other domains, which is correlated to a rise in UHS, can make the tool more advanced and gives a possibility to use the tool universally; (v) input requirements from the user, which means the decision results depend on the input data; (vi) indicator showing the vulnerability, heat events and effectiveness of the intervention; (vii) political and administrative support for developing the tool; (viii) vegetation, which is a basic and natural intervention that helps to reduce heat stress; (ix) graphical interface and heat stress visualization by mapping; (x) spatial coverage, which helps to indicate the suspect areas in a city on a GIS map; (xi) cost assessment of the measure; (xii) quick assessment of the intervention's effectiveness in real-time; (xiii) user-friendliness, which shows how easy and difficult it is to use the tool; (xiv) uncertainty risk analysis, which gives trustworthy results; and (xv) plus points, which are when the tool provides a long-term effect of heat stress or considers other interventions apart from vegetation. These selected criteria were obtained after going through the literature and serve as a methodology, as shown in Figure 1.
coverage, which helps to indicate the suspect areas in a city on a GIS map; (xi) cost assessment of the measure; (xii) quick assessment of the intervention's effectiveness in real-time; (xiii) user-friendliness, which shows how easy and difficult it is to use the tool; (xiv) uncertainty risk analysis, which gives trustworthy results; and (xv) plus points, which are when the tool provides a long-term effect of heat stress or considers other interventions apart from vegetation. These selected criteria were obtained after going through the literature and serve as a methodology, as shown in Figure 1.

Multi-Criteria Decision Analysis
Decision-making tools are valuable in tackling issues with numerous actors, criteria and objectives. Generally, MCDA is based on five components, which are: goals, decisionmakers' preferences, alternatives, criteria and results, respectively. In light of many alternatives, differences can be catered between Multi-Attribute Decision Making (MADM) and Multi-Objective Decision Making (MODM), but both offer comparative characteristics. MODM is reasonable for the assessment of consistent options when there is a need to predefine constraints in the form of choice vectors. A set of target functions is optimized considering the limitations while decreasing the performance of at least one goal. In MADM, inherent characteristics are covered by prompting the thought of fewer options, and evaluation becomes difficult as prioritizing turns out to be more difficult. The result is obtained by comparing different alternatives concerning each criterion [10][11][12]. Differ-

Multi-Criteria Decision Analysis
Decision-making tools are valuable in tackling issues with numerous actors, criteria and objectives. Generally, MCDA is based on five components, which are: goals, decision-makers' preferences, alternatives, criteria and results, respectively. In light of many alternatives, differences can be catered between Multi-Attribute Decision Making (MADM) and Multi-Objective Decision Making (MODM), but both offer comparative characteristics. MODM is reasonable for the assessment of consistent options when there is a need to predefine constraints in the form of choice vectors. A set of target functions is optimized considering the limitations while decreasing the performance of at least one goal. In MADM, inherent characteristics are covered by prompting the thought of fewer options, and evaluation becomes difficult as prioritizing turns out to be more difficult. The result is obtained by comparing different alternatives concerning each criterion [10][11][12]. Different multi-criteria techniques are applied in the field of UHI mitigation, thermal comfort improvement, and the selection of the heat stress index. MCDA models are developed according to the researcher's point of view concerning demand and goal. It can be a direct or indirect methodology. In a direct approach, the task of priorities or weights is performed as a result of contributions from a questionnaire. In an indirect approach, all the potential criteria are separated into components and assigned weights as per past comparable issues, and the judgment of decision-makers is based on experience. MCDA is consistently complex because of the involvement of stakeholders and factors which are technical, institutional, legislative, social and financial. The overall strategy of the MCDA technique is presented in Figure 2. A survey has been conducted on the use of different MCDA techniques for UHI and UHS mitigation.
Climate 2021, 9, 102 4 of 23 or indirect methodology. In a direct approach, the task of priorities or weights is performed as a result of contributions from a questionnaire. In an indirect approach, all the potential criteria are separated into components and assigned weights as per past comparable issues, and the judgment of decision-makers is based on experience. MCDA is consistently complex because of the involvement of stakeholders and factors which are technical, institutional, legislative, social and financial. The overall strategy of the MCDA technique is presented in Figure 2. A survey has been conducted on the use of different MCDA techniques for UHI and UHS mitigation. The following methods were applied for UHS [3,[13][14][15][16][17][18][19][20] and are briefly discussed with their limitations in Table 1.  The following methods were applied for UHS [3,[13][14][15][16][17][18][19][20] and are briefly discussed with their limitations in Table 1.

Aim of the Study Method
Step Limitation Reference Green roof adaptation in Thailand to mitigate UHI. The relevant factors were identified in qualitative content analysis, structured alongside two dimensions (internal/external and positive/negative factors), and quantitatively assessed.
Analytical hierarchy process based on expert judgments, strength, weakness, opportunities, threats (SWOT) analysis.
A limited number of factors considered and lengthy pairwise comparisons.
Tachaya Sangkakool [13] Assess the heat stress relative vulnerability of 15 local government areas in metropolitan Sydney.
Multi-criteria outranking approach (build analogy between multi-criteria decision analysis and indicator-based vulnerability assessment (IBVA)). Electric III ranking process.
Stage 3: Distillation and ranking procedures.
Nonlinearities might not be incorporated in the outranking aggregation process.

Aim of the Study Method
Step Limitation Reference Investigate the inner-dependencies between the benefits, opportunities, cost, risks for proper adoption of green roof installation.
The enhanced fuzzy Delphi method (EFDM) and fuzzy decision-making trial and evaluation laboratory (FDEMATEL) approaches.
Step 1: Select the panel of experts.
Step 2: Design and distribute the questionnaire.
Membership function is: Step 3: Develop initial direct relation fuzzy matrix.
Step 4: Normalize the initial direct relation fuzzy matrix.
Step 5: Develop the total direct and indirect relation fuzzy matrix.
Step 6: Defuzzify the entries in the fuzzy total relation matrix.
Step 7: Produce causal diagrams, values for D + R and D − R were calculated by the following equations: Absence of significant relationships among environmental and economic opportunities.
Design the questionnaire and send it to the experts, 2.
Organize experts' opinions collected from the questionnaire into an estimate, and create the Triangular Fuzzy Numbers (TFNs), 3.
Select the criteria affecting decision making.
Sets of pairwise comparisons according to the direction of influence of the relationship between the criteria/sub-criteria were generated. The comparison scale for pairwise comparison is 0, 1, 2, 3, and 4, which denote no influence, low influence, medium influence, and high influence, respectively.

2.
The direct-relation matrix was generated, which is the average of pairwise comparison matrixes that have been generated in step 1 by 28 experts. An n × n matrix A, in which Aij is the degree to which criterion i affects criterion j.
If the expert decides to change an answer or decides to add any new information, the first round should be repeated, and the process will be time-consuming.
Amir Mahdiyar [16] The study aims to map the UHI of a mid-size city (Rennes, France) and define the relevant land-use factors. The UHI was measured by 22 weather stations in different contexts: urban, suburban, and peri-urban.
Multi-criteria linear regression method used to build a model of the UHI.
1. The first step of the process was to build a regression model by selecting explanatory variables; 2.
The second step of the process was to execute the selected regression during the first step; 3.
The regression coefficients were applied to the associated raster.
Limited variables considered, do not provide reasoning and spatial method.

Aim of the Study Method
Step Limitation Reference Examines major local climate zones (LCZs), with greater coverage area, in the city of Nagpur, India by selecting critical LCZ and mitigation strategies such as greening, cool roof, and cool pavement using ENVI met tool. The study is conducted in three phases. The first stage deals with air temperature and UHI investigation. The second stage covers the issue of identifying criticality using multi-criteria decision making (MCDM) technique. The third stage examines the selection of mitigation strategies, simulation environment, and mitigation priorities.
The technique for order of preference by similarity to the ideal solution (TOPSIS).

1.
Construct decision matrix (X) and assign weightage to the criteria.
Determine the positive ideal (A + ) and negative ideal (A − ) solutions.
Calculate the separation measures from the positive ideal solution (di + ) and the negative ideal solution (di − ) Calculate the relative closeness to the positive ideal solution (Performance Score) and rank the preference order or select the alternative closest to 1.

Aim of the Study Method
Step Limitation Reference An exhaustive study proposing a new index aimed at quantifying the hazard of the absolute maximum UHI intensity in urban districts during the summer season by taking all the parameters influencing the phenomenon into account. In addition, for the first time, the influence of each parameter has been quantified.
Results are achieved by exploiting three synergistically related techniques: analytic hierarchy processes to analyze the parameters involved in the UHI phenomenon; a state-of-the-art technique to acquire a large set of data; and an optimization procedure involving a Jackknife resampling approach to calibrate the index by exploiting the effective UHI intensity measured in a total of 41 urban districts and 35 European cities.

1.
The AHP step 1 consists of the Structure of the Problem to determine an index useful to quantify potential UHII in the urban district.

2.
The AHP step 2 is used to individually analyze each aspect of the defined UHII problem in order to weigh the parameters involved.

3.
The summary of priority is obtained by multiplying each criteria weight by the intensity range weight and adding the results.
Based on literature quantitative analysis.
Sangiorgio [3] Weighting Criteria and Prioritizing of Heat Stress Indices in Surface Mining.
The viewpoints of occupational health experts and the qualitative Delphi methods were used to extract the most important criteria. Then, the weights of 11 selected criteria were determined by the Fuzzy Analytic Hierarchy Process.
Finally, the fuzzy TOPSIS technique was applied for choosing the most suitable heat stress index.

1.
The formation of implementing team and monitoring the Delphi process; 2.
Selecting the experts and participants; 3.
Adjusting the questionnaire for the first round; 4.
Sending the questionnaire to experts; 6.
Analyzing the obtained responses in the first round; 7.
Preparing the second-round questionnaire considering the required revisions; 8.
Sending the second questionnaire to the same experts; 9.
Analyzing the results of the second questionnaire; 10. Determining the relative weights of each criterion using the fuzzy AHP; 11. Choosing a heat stress index among the existing ones in the study using the fuzzy TOPSIS method.

Decision Support Tools
From an environmental perspective, decision-making involves multiple complex steps for various stakeholders with different objectives and priorities. Most concerned people tend to attempt heuristic or intuitive approaches in order to simplify the problem to make it manageable. By following this approach, stakeholders lose important information and may discard the contradictory facts and factors of uncertainty and risks. In other words, it is not suitable for making thoughtful choices that can focus on all the important points of the process [21]. Therefore, a proper strategic decision-making tool is helpful to assess the decision-makers to bring about the process strategically and manage the multitude of ideas properly [22,23]. Additionally, during the process of decision-making, practitioners are supposed to take the elements of biodiversity, social innovation, governance, and urban management into consideration within a socio-ecological framework [24,25].
These tools are defined as an approach involving any techniques, models, frameworks (one project's framework can be seen in Figure 3), or methodologies that strategically manage and support the decision-making [26]. Moreover, decision-making tools help to evaluate and monitor the co-benefits systematically [27] and processes for connecting, reflecting and investigating, exploring, and modeling while suggesting proper solutions [28].

Decision Support Tools
From an environmental perspective, decision-making involves multiple complex steps for various stakeholders with different objectives and priorities. Most concerned people tend to attempt heuristic or intuitive approaches in order to simplify the problem to make it manageable. By following this approach, stakeholders lose important information and may discard the contradictory facts and factors of uncertainty and risks. In other words, it is not suitable for making thoughtful choices that can focus on all the important points of the process [21]. Therefore, a proper strategic decision-making tool is helpful to assess the decision-makers to bring about the process strategically and manage the multitude of ideas properly [22,23]. Additionally, during the process of decision-making, practitioners are supposed to take the elements of biodiversity, social innovation, governance, and urban management into consideration within a socio-ecological framework [24,25].
These tools are defined as an approach involving any techniques, models, frameworks (one project's framework can be seen in Figure 3), or methodologies that strategically manage and support the decision-making [26]. Moreover, decision-making tools help to evaluate and monitor the co-benefits systematically [27] and processes for connecting, reflecting and investigating, exploring, and modeling while suggesting proper solutions [28]. One such example of these tools is the Adaptation Planning Support Tool (APST), which is specifically designed to focus on the impacts due to climate change. This toolbox has been proven to be useful for policy-makers and has been applied practically in many cities [30].
The Mitigation Impact Screening Tool (MIST) is another decision software-based tool developed by the US Environmental Protection Agency (EPA) for an assessment of the impacts of UHI mitigation strategies' (mainly albedo and vegetation) increase on the reduction in urban air temperatures, ozone, and energy consumption for over 200 US cities [31]. The tool is currently unavailable as it was disabled by the EPA due to the update of One such example of these tools is the Adaptation Planning Support Tool (APST), which is specifically designed to focus on the impacts due to climate change. This toolbox has been proven to be useful for policy-makers and has been applied practically in many cities [30].
The Mitigation Impact Screening Tool (MIST) is another decision software-based tool developed by the US Environmental Protection Agency (EPA) for an assessment of the impacts of UHI mitigation strategies' (mainly albedo and vegetation) increase on the reduction in urban air temperatures, ozone, and energy consumption for over 200 US cities [31]. The tool is currently unavailable as it was disabled by the EPA due to the update of the methodology and data inputs. Nevertheless, some authors have analyzed how it functioned, as it attempted to provide a practical and customized assessment for UHI reduction.
Furthermore, there are various nature-based solutions and their implementation can offer multiple benefits, for example, Stadtklimalotse, Wiki, REGKLAM, SUPER (Sustainable Urban Planning for Ecosystem Services and Resilience), and many more [32,33]. Table 2 represents the review of tools designed for policy-makers and urban planners to use during the process of decision-making for urban heat, climate change, heat vulnerability, health heat events, etc.

Results and Discussion
Multi-criteria mathematical models [3,[12][13][14][15][16][17][18][19] are a valuable, theoretical, qualitative, and quantitative way of decision making and also a first step towards developing a DST. These models are supported by expert assistance which considers the socio-cultural factors and local environment. They cover the criteria which can be assessed statistically, e.g., cost analysis and political and administrative support.
The AHP is a qualitative approach and depends on the judgments of the people who are involved in the task, but lengthy pairwise comparisons might lead to inconsistency. Multi-criteria outranking is also controversial, and questions were raised about outranking procedures, nonlinearities' incorporation, and aggregation processes. Similarly, in FDE-MATEL, no significant relationships could be found for some criteria. Another issue is that the questionnaire can have a low response rate, be time demanding, and have a low probability of filtering out specific opinions.
Most of the time, this is a trial and error process. Decision-making for urban heat mitigation involves multiple and complex steps that vary on different stakeholders with various adaptation measures and needs. Plus, during the process of decision-making, practitioners should take into consideration the criteria of biodiversity, social innovation, governance, and metropolitan management within a socio-ecological framework.
Some North American, European and Australian DSTs are critically analyzed concerning all the criteria which were considered in this review paper. The results are summarized and classified in Table 3. For future development, recommendations of approaches learnt from the surveyed tools are highlighted by a color-coding scale shown in Table 4.
The DSS [49] was developed in the framework of the European project "Development and application of mitigation and adaptation strategies for counteracting the global phenomenon UHI". This tool is user-friendly and covers many aspects which are needed to support urban planners.
It is known that every testing (pilot) site is different depending on several factors such as climate, population, group of persons, building infrastructure, availability of existing interventions and number of heat events. The development of a DST depends on the scale of the project. Objectives and limited spatial coverage are always a drawback because all decision results are based on different pilot sites' data and tools are based on those characteristics.

Conclusions
Decision-making is a difficult task that has to go through different phases such as identifying reliable and efficient measures, assessing the challenges to investigate the case studies, and building a systematic framework for decision support. The MCDA approach is a valuable and very important initial step to develop a DST to deal with UHS. Toolkits in the form of handbooks are neither spatial nor interactive. Web-based tools are mostly interactive and can provide an assessment of green, blue and grey interventions on heat impact in real-time and help decision-makers to take actions on the heat vulnerability of the suspected area. In these tools, economic and environmental assessment can be performed quite easily through a graphical interface; however, the results always depend on input data which are often difficult to obtain.
In this review and comparative study, we conclude that despite many existing publications and reported tools, there is still room for improvement, which can be achieved by a holistic approach dealing with subjective and objective aspects of heat stress, combining various inputs from sensors as well as from experts and residents' feedback, and using different techniques such as MCDA, GIS, urban planning and, in the end, artificial intelligence tools to correlate these aspects with each other to develop a reliable DSS for the mitigation of heat stress.
Author Contributions: The paper was a collaborative effort between the authors. A.M.Q. and A.R. contributed collectively to developing the methodology of this survey, tools comparison and the manuscript preparation. All authors have read and agreed to the published version of the manuscript.

Funding:
The COOL-TOWNS (Spatial Adaptation for Heat Resilience in Small and Medium-Sized Cities in the 2 Seas Region) project receives funding from the Interreg 2 Seas program 2014-2020 co-funded by the European Regional Development Fund under subsidy contract N • 2S05-040.
Institutional Review Board Statement: Not applicable.