Evaluating and analyzing the comprehensive community 1 disaster reduction capability

: China currently faces the important task of strengthening its regional disaster reduction 6 capabilities. Communities are basic components of urban areas, and ensuring that they participate in 7 disaster reduction is important for urban safety. The "National Demonstration Community for 8 Comprehensive Disaster Reduction (NDCCDR)" is one of the community comprehensive disaster 9 prevention and reduction work organized nationwide by the National Committee for Disaster 10 Reduction and the Ministry of Civil Affairs in order to carry out the disaster relief work efficiently and 11 orderly and strengthen the capacity building of disaster prevention and reduction. In this paper, in view 12 of the imperfect evaluation criteria in the NDCCDR, and according to the connotation of community 13 comprehensive disaster reduction capability (CCDRC), we construct an index system to evaluate the 14 CCDRC; the system is comprised of six primary evaluation indices that measure the ability to evaluate 15 disaster risk, the ability to provide rescue and support, the presence of engineering defenses, the 16 presence of social and economic base support, the ability to manage disasters, and the level of public 17 cognitive ability. These six primary indices include 31 secondary indices. Because the index system is 18 characterized by a small sample size, sparse information, and large spatial extent, for the index system, 19 we evaluate CCDRC using the entropy-weighted grey target model and geographic information system 20 (GIS) overlay analysis. According to the distribution status of NDCCDR, we take the Suzhou New 21 District (SND) as a case study for the empirical measurements and calculations and use ArcMap 10.2 22 software to produce a map of the spatial distribution of CCDRC in this region. The results indicate that 23 the area’s CCDRC is relatively low. The spatial distribution of CCDRC is uneven. However, the 24 CCDRC in the region has a good foundation and it also has large potential for improvement. The 25 results also show that all of the NDCCDR are in the forefront of the case area, but their CCDRC is 26 unbalanced and their primary evaluation indices of CCDRC are also not balanced. Therefore, we 27 propose that the construction of NDCCDR and CCDRC should be combined, from point to face, and 28 ultimately to improve the overall level of CCDRC in the community


Introduction
Despite the capacity to understand and transform nature, humans remain vulnerable to various natural disasters.Studies on the techniques, methods, and strategies involved in disaster prevention and mitigation remain important.In the past, engineering defenses have constituted the main approach to disaster mitigation.However, studies have suggested that this method is not always a successful means of disaster prevention and reduction (Gall et al., 2011;Deckers et al., 2010).Risk analysis can help decision makers identify high risk areas and thus use their limited capital and resources in the correct places (Hanne Glas et al., 2016;Shi et al., 2012).Therefore, risk analysis can significantly increase the effectiveness of disaster management, and it has become an active focus of research on disaster prevention and reduction in recent years.As the basic component of urban public disaster prevention and reduction strategies, communities play an important role in urban safety systems.More attention should be paid to disaster risk evaluation at the community scale.The national comprehensive disaster prevention and reduction plan (2016( -2020, China) , China) identified the "conduction of community disaster risk identification and evaluation, and compilation of community disaster risk maps" as a major strategy for disaster prevention and reduction over this time period (General Office of the State Council, 2017).Rapid urbanization and the interactions between various natural disasters mean that urban and rural communities are impacted by many kinds of disasters, sometimes simultaneously.Therefore, community-level disaster management has increasingly become the focus of global disaster management (Li, 2012).Currently, the international community follows the concept of "community-based disaster risk management" (CBDRM) (Bajet et al., 2008).Emphasis is placed on a complete understanding of disaster risk (in addition to natural disasters, including public health, traffic security, social safety, and accidents involving water, gas, and electricity) (UNISDR, 2012), a holistic approach to disaster management (Zhan, 2006) and the universal characteristics of disaster prevention and preparation (Zhou, 2013).
In recent years, China has experienced several major disasters, including the Sichuan (Wenchuan) earthquake, the Yushu, Qinghai earthquake, and the Zhouqu debris flows in Gansu Province.Therefore, the concept of comprehensive disaster reduction capability has gained significant traction.It is generally recognized that comprehensive improvements in the disaster reduction capability will reduce or mitigate the causalities and property loss caused by disasters (Hu, 2013).Prior to the International Day for Disaster Reduction in 2006, the State Council of China convened the "Symposium on enhancing the comprehensive disaster reduction capability", which focused on comprehensively strengthening China's comprehensive disaster reduction capability.The concept of comprehensive disaster reduction involves four main aspects: first, preparing for and defending against various kinds of disasters; second, preparing to respond at different stages during the development of a disaster; third, integrating various resources; and fourth, applying various disaster reduction approaches (Lyu, 2011).
Many studies have addressed the disaster reduction capability in China and other countries, and their results have been adopted by the governmental organizations and committee and applied in actual regional disaster management situations.These studies have mostly focused on single types of disasters ( (Francesco D et al., 2017;Zhang, 2004), single aspects of disaster reduction (Boris F et al., 2016;Daniel Green et al., 2017), and the comprehensive regional disaster reduction capability (Ma, 2007).
Relatively few studies have addressed disaster reduction at the community scale.Although some authors have constructed index systems for the ability of communities to prevent and mitigate disasters, they did not propose an in-depth or specific quantitative method (Smith et al., 2017;Yi, 2012).
The concrete measures contain the use of community-based early warning systems (Paul J. Smith et al., 2017), community infrastructure exposure risk analysis (S. Fuchs et al., 2015； R. Figueiredo et al., 2016； Saif Shabou et al., 2017), disaster risk reduction education (Avianto Amri et al., 2017) and community-level resilience to disaster (Adriana Keating et al., 2017；Estefania Aroca-Jimenez et al., 2017).However, there is little research on the quantitative evaluation method of CCDRC.In this paper, we aim to address the above situation by proposing a quantitative method of evaluating CCDRC.We construct an index system for CCDRC.We quantitatively evaluate CCDRC using grey target modeling and overlay analysis of GIS，and use the Suzhou New District (SND) as a case study to demonstrate our calculations referring to the distribution status of NDCCDR.We analyze the spatial distribution of CCDRC in the study area with the goal of providing decision support for efficient disaster response management by local government.The evaluation of CCDRC will help to defend against various kinds of disasters in the community as a whole, respond at different stages during the development of a disaster, integrate various resources, and coordinate various disaster reduction approaches.Through the evaluation of CCDRC, the overall level of CCDRC and the status of the main evaluation indexes can be grasps, so as to make it easy to take specific measures to effectively strengthen the weak links.The evaluation indices in this paper have the completeness, availability and quantifiable characteristics.The model has the advantages of simple construction, space transferability, simple operation, and multiple characteristics of evaluation results.Therefore, the method introduced in this paper is universal.

Defining the CCDRC
Based on CBDRM ideas, as well as the general philosophy of regional disaster reduction capability in China, we define the CCDRC as follows: a community's ability to avoid or reduce natural disasters and accidents involving public health, traffic security, and major utilities by using engineering and non-engineering measures to integrate resources from the government, non-governmental organizations, community residents, and the general public.These measures are taken during the process of disaster prevention and preparation, emergency response, and post-disaster recovery, with the aim of protecting the life and property of residents and supporting their normal activities (especially vulnerable groups), as well as the normal operation of industrial activities in the community.When evaluating the CCDRC, several aspects are critical: (1) while the disaster reduction capability at the community scale is an integral component of disaster reduction capability at the regional scale, they should not be evaluated or measured in the same way; (2) in addition to community organizations, the government, community, residents, and other organizations all contribute to community disaster reduction; (3) community disaster reduction is defined by the ability to cope with various kinds of disasters (both natural and man-made), not any single disaster; and (4) it is important to comprehensively consider various factors, including the evaluation criteria in the NDCCDR, when constructing an objective and comprehensive index system for evaluating the CCDRC.

Creating an index system for evaluating CCDRC
The NDCCDR uses demonstration as a means to enhance a community's ability to reduce disaster risk.The document defines ten aspects of organizational management mechanism, disaster risk evaluation, infrastructure of disaster prevention and mitigation and so on as the basic elements for compliance with the disaster reduction demonstration community (Office of National Disaster Reduction Committee, 2010).The wide-abroad implemented CBDRM attach importance to software construction but despise hardware environment construction.Compared with these disadvantages, the NDCCDR not only strengthen the planning and construction of software such as community residents' awareness and skills for disaster reduction, disaster reduction publicity and training, but also has taken into account the construction of community disaster reduction hardware such as shelters and material reserves.However, the above indicators are not enough to fully reflect the CCDRC.Therefore, based on the meaning of CCDRC, we consider quantitative factors including a community's economic status, rescue and safeguarding resources, and engineering defenses, as well as qualitative factors including disaster risk evaluation, organizational management, and public awareness of disaster prevention.
Taking account of the universality, availability and quantifiable characteristics of indicators, we construct the evaluation index system of CCDRC.Our index system is made up of six primary indices and a total of 31 secondary indices, which mainly includes two types of attribute (spatial attribute and non spatial attribute), involve binary, numerical, and categorical data.The indices are listed in Table A1 of the Appendix A.

Evaluating the CCDRC
Our index system has the following characteristics: (1) it is hierarchical, but includes numerical, binary, and categorical data; thus, these data cannot be processed in a standard way, and it is difficult to determine their weights using traditional methods; (2) the index data do not have empirical values, and the quantity of data is small, so quantitative evaluation is difficult; and (3) the same index differs spatially, can evolve between communities and can be transformed to an index with consistent polarization.Based on the aforementioned characteristics, we use grey target modeling to evaluate the comprehensive disaster reduction capability of a single community.Next, we use GIS overlay analysis to create a map showing the spatial distribution of the CCDRC throughout the region.

Entropy weighted grey target model
For evaluating data that involves a small sample and sparse and uncertain information, we first set a grey target and take the bull's eye of the grey target as the standard model.The model is divided into different grades based on the degree to which the model to be evaluated is close to the target center (that is, the target center degree).This method is the traditional grey target model (Deng, 2002).We can then consider the degree to which the various evaluation indices influence the target center degree and use the entropy weight method to determine the weight of the evaluation indices, which will yield more objective and fair evaluation results.This methodology constitutes the improved grey target model, or the entropy-weighted grey target model (Li et al., 2012;Li et al., 2013;Li et al., 2016).The target center degree of various spatial units can be calculated by the following Eq.(1): , The fraction on the right side of the equation is the target center coefficient of index j (j=1, 2, …, n) with a spatial unit of i; ∆ 0i (j) is the corresponding grey correlation difference; and w j is the weight of index j.Equation (2) shows the formula for calculating the entropy weight, as follows: where is the conditional entropy of index j (Jin, 1994;Lian, 2004).

GIS spatial overlay analysis
Spatial overlaying is an important spatial analysis method in GIS.The method overlays two or more layers on the same scale in the same region to generate a new layer with multiple attributes.The new layer synthesizes the attributes of the original layers; this new layer represents a new spatial relationship as well as indicating the relationship between the attributes of the original layers based on logical operations (Gong et al., 2001;Yang et al., 2016;He et al., 2015).Overlaying a polygon layer includes both intersection and identity (Fig. 1).Both the range and attribute of a spatial unit will change after the intersection operation.This kind of overlay requires logical operations and also includes complicated topological operations on spatial objects.After the identity operation, the range of the spatial units will not change, but the attributes will, mainly because of the logical operations.When there many layers are included in the operation, their weights must be considered.In this paper, to determine a community's CCDRC, we take the community as the evaluation unit and the layers corresponding to various evaluation indices as the objects of operation; we use the entropy-weighted grey target model for the identity operation.

Evaluation method and process
Setting the community as the unit, we sequentially constructed the impact space of the index sequence {U j } and the standard model {x 0 (U j )} based on grey target modeling from the six primary evaluation indices.The standard value of various indices in the impacting space is related to the index polarity.The maximum standard value of index polarity is the maximum of the corresponding index in the impacting space, and the minimum polarity index is the corresponding minimum.The specific procedures of measurement and calculation are described below.
(1) Grey target transformation and determination of the grey correlation difference information: to avoid the effect of large values (caused by excessively large differences between data values in the standard model sequence) on the relatively small values, we conducted grey target transformation on the various index sequences.After grey target transformation, the value of the evaluation index U j in the i th research unit, Tx i (U j ), was calculated using the following Eq.(3): , Next, we obtained the grey correlation difference for index U j for unit i in equation ( 1), ∆ 0i (j), as follows: , (2) Calculation of conditional entropy j H for index U j : the relative distance between the index U j in unit i and the standard value x 0 (U j ) can be expressed with the closeness degree d ij , as follows: where is the difference between the maximum and minimum of index j for all the spatial units of the study area.The normalized value of the uncertainty measurement (that is, the conditional entropy) for the relative significance of this index can be expressed as follows: , where .If d ij =0, we prescribed .
(3) Calculation of the target center degree and grades for the various spatial units: first, we calculated the entropy weight w j of the various indices j using Eq.( 2).Next, using Eq.( 1), we calculated the target center degree of the corresponding primary index of the different spatial units and the target center degree of the corresponding CCDRC.
(4) Production of the CCDRC spatial distribution map: we import the above calculation results into ArcGIS10.2,and the target layer of each level index of the study area will be set up.Then we carry out the GIS overlay calculation according to the calculation method of the total target center degree, and the CCDRC spatial distribution map of the study area can be obtained.
Using the CCDRC primary indices spatial distribution map and CCDRC spatial distribution map generated in the case study area based on the above model, we can seek the following target: (1) the overall level of CCDRC in case area, (2) the spatial distribution of CCDRC in the region, (3) the potential analysis and improvement measures of CCDRC, (4) the CCDRC level of NDCCDR.
We can obtain the spatial attribute data of the index system from a regional geographic database.
The non spatial attribute data can be obtained by means of community disaster monitoring logs, reports and other historical archives, resident visits and field surveys with the help of local civil affairs department.Therefore, the process of CCDRC calculation and analysis based on entropy weight -grey target model and GIS overlay method in this paper, is generally applicable to most of the provinces in mainland China.If the model is used to evaluate the foreign CCDRC, the benefits will offer to communities that are vulnerable to various kinds of disasters, as well as some challenges, such as different national conditions, unbalanced economic development among countries, great differences in the system of disaster prevention and reduction, and great differences in the organization level of the project.To address these challenges, features that make this approach worth considering in the context -The model used in this paper has comparability between spatial units, and has transferability between regions; using entropy method to determine index weights can avoid the arbitrariness and unilateralism of subjective weight determination.
-This method is not focused on the index itself, and does not need to establish a function relationship between the indexes, but rather to model the order relation represented by the index value, so it is very easy to operate, and the results of evaluation are diversified.
On the whole, it is world widely feasible to apply the methods introduced in this paper to evaluate the CCDRC.

The distribution status of NDCCDR
Since the selection of NDCCDR was organized in 2008, nearly three thousand community have been selected (or once been selected) in succession.this paper addresses the measurement and analysis of these aspects so that they can be strengthened.

Calculation of CCDRC in the region
We surveyed and conducted statistical analyses on all 83 communities in the study area; we obtained effective sample data for 72 communities, which we used as the basic data.According to the grey target theory, we built the community-based influence space of the index sequence from six aspects, namely, disaster risk evaluation capability, rescue and support capability.In the light of formula ( 1) -( 6), we calculated the entropy weight, constructed the grey correlation difference information, and then obtained the target center degree.During the calculation, we need to pay attention to the index polarity.Except for number of group event disasters, number of fire disasters and the disaster risk intensity, all other secondary indices are maximum polarity indices.
We calculate the rescue and support capability for eight communities on Shishan Street as an example.Following the procedures above, we sequentially conducted grey target transformation, spatially determined the grey correlation difference information and conditional entropy, and calculated the entropy weight of each evaluation index (the calculated entropy weights for all the primary and secondary indices are shown in Table A1 of the Appendix A).Finally, we obtained the target center degree of the various primary evaluation indices.The results are shown in Table 1.Note: the entropy weight corresponding to the various evaluation indices is the normalized weight determined with the 72 spatial units (communities) in the research area used as the reference.
The target center degree indicates the strength of the rescue and support capability.Table 1 shows that, of the eight Shishan Street communities, the rescue and support capability is strongest in the four NDCCDR, Wanfeng, Xinsheng, Xintai, and Shishan (Fig. 3 (b)); it is relatively weak in the other communities.

Grading the CCDRC
After inputting the data from Table 1 into ArcGIS 10.2, we conducted the GIS overlay operation using the entropy-weighted gray target model described above, resulting in a map of the distribution of the rescue and support capability in the Shishan Street communities.We repeated this operation for all the communities in the study area and then graded their capabilities based on the target center degree.
We used anomalies to create four capability grades, excellent, good, moderate, and poor.Next, we created a map of the distribution of the rescue and support capability grades for the communities in the study area.We repeated the same operation for the disaster risk evaluation, engineering defense, social and economic base support, disaster management and public cognitive capabilities of all the communities, yielding the individual grade distribution maps for each capability.Finally, based on the entropy weight of the primary evaluation indices (Table A1 of the Appendix), we used GIS overlay analysis to determine the total target center degree of the CCDRC for all the communities in the study area.Once divided into grades, based on the minimum information principle, the target center degree should not be smaller than 1/(1+ζ) (where ζ is the resolution coefficient, and its value in Eq. ( 1) is 0.5).
Therefore, the minimum of the total target center degree should be 0.3333.The results show that the maximum total target center degree was 0.6941 for the Wanfeng community of Shishan Street.
Therefore, we created four grades in the interval [0.3333, 0.6941] and used these grades to create a map of the spatial distribution of grades of CCDRC (Fig. 5).The blank regions in the figure show communities where we could not obtain qualified sample data.

Result analysis and suggestion
Comparing the spatial distribution of the grades of CCDRC in Fig. 5 to the administrative divisions shown in Fig. 3 (b), we observe the following characteristics: (1) The CCDRC in the study area is generally weak.Communities with a poor CCDRC account for 63.1% of the study area, and communities with a moderate grade account for 28.6%; communities with an excellent grade only account for 8.3%.The distribution of these grades is not accidental but is derived from the spatial distribution of the primary evaluation indices.For convenience of comparison, we set the interval of the target center degree for all the primary evaluation indices to [0.3333, 1].We   (3) There is great potential to improve the CCDRC.The area ratio for the different grades of the primary evaluation indices and CCDRC in the study area can be obtained from Fig. 5 and Fig. 6 (see Table 2).Note: the primary evaluation indices in the table are shown in descending order from left to right based on their entropy weights.
The results show that communities with moderate or higher CCDRC account for 36.9% of the study area; with the exception of the public cognitive capability and the rescue and support capability (with the largest weights), communities with a grade of moderate or higher for other indices account for over 76.1% of the study area, and communities with an excellent disaster management capability (with the third highest entropy weight) account for 63%.To further analyze the potential for improvements in the CCDRC, we use Tongan, which has a moderate CCDRC, as an example.Most secondary evaluation indices of its engineering defense and community social and economic base support capabilities are close to or better than the average level of the study area (as shown in Fig. 7, where the vertical axis is the ratio between an evaluation index and the index's average value in the study area).The analysis above indicates that there is a relatively large potential for improving the CCDRC in the research area.
Above analyses suggest that we should focus on the following several aspects to enhance the CCDRC in the study area: (1) Several measures should be taken to improve the cognitive level in the communities.Because its weight is largest, the public cognitive capability significantly affects the CCDRC.Figure 6(a) and Table 2 both demonstrate that the public cognitive capability in the research area is insufficient, communities with a poor grade account for 82.4% of the study area.In modern communities with highly developed means of communication, many measures can be used to improve the public cognitive capability.For example, some measures include creating official ways to disseminate disaster reduction information, developing publicity material, setting up disaster early-warning display screens, and increasing the amount of publicity material to increase residents' knowledge of disaster prevention and reduction.Better publicizing disaster reduction activities will help residents understand the dangers of disasters and instill a common sense of proper emergency behavior.Enhancing residents' consciousness regarding disaster prevention and reduction will also help attract volunteers to join disaster prevention and reduction teams and eventually strengthen the overall cognitive capability of the public in the study area.(2) It is also important to collectively manage and reinforce the effectiveness of disaster relief measures and safeguards.Because the disaster management capability of communities in the study area is relatively good (Fig. 6(c)), we compare it to the target center degree of the community rescue and support capability to yield a plot of the target center degree for the community rescue and support capability (Fig. 8).This index is close to the minimum of the target center degree (0.333), in contrast to the disaster management capability.We conclude that it is important to reinforce the disaster rescue and support capability, including strengthening coordination between the relevant governmental departments, investing in multiple aspects of disaster reduction and allocation of per capita medical resources, appointing disaster information personnel, and setting aside more rescue and emergency material.These measures will fundamentally strengthen the disaster rescue and support capability in the study area.(3) Similarly, the engineering disaster defense capability should also be strengthened.As shown in Figure 6(d) and Table 2, the engineering defense capability in the study area is a bimodal distribution.
Communities with an excellent grade (including Dongzhu, Shishan Street, and some of Xushuguan) account for 33.8% of the study area, while communities with a moderate or poor grade account for 35.5% and 30%, respectively.We conclude that internal resources in the study area should be allocated in a better manner; the engineering defense capability of the communities graded excellent can be leveraged in planning, construction, maintenance and management.These strategies will reinforce the engineering defense capabilities of communities throughout the study area.

Conclusions
Community disaster prevention and reduction is a basic component of urban disaster prevention and reduction and plays an important role in the urban security system.In this paper, we constructed an index system to evaluate the CCDRC; the system includes six primary and 31 secondary indices.We used entropy-weighted gray target modeling to determine the CCDRC, and GIS spatial overlay analysis to map the spatial distribution of disaster reduction capability grades.We focused on the SND as a case study and obtained the following conclusions: (1) The area's CCDRC is relatively low; communities with a poor grade account for 63.1% of the study area, and those with an excellent grade account for only 8.3%.Additionally, the spatial distribution of CCDRC is uneven.The eastern communities have relatively good capability, while the western communities adjacent to Lake Taihu have poorer capability.The Shishan Street community, located in the southeastern part of the study region, has the strongest CCDRC.However, the CCDRC in the region has a good foundation and it also has large potential for improvement.After analyzing these results, we propose that CCDRC can be most improved by enhancing their level of public cognitive ability, rescue and support capacity, and engineering defenses.
(2) In general, the CCDRC of the NDCCDR is at the forefront in the study area, but the CCDRC among the NDCCDR is not balanced.As shown in Fig. 5, among the NDCCDR, the CCDRC of Ylian is rated "excellent", but Hxiang or Yxin is "good".The CCDRC primary indices of the NDCCDR are also quite different from each other.As we can see in Fig. 6, the public cognitive capability and the rescue and support capability of the NDCCDR show a general poor feature, but their engineering defense capabilities and disaster management capability are in the overall advantage.Above discussions fully indicate that the CCDRC of NDCCDR is not necessarily good.On the contrary, the evaluation and construction of CCDRC is not only the focus of community disaster prevention and reduction work, but also the standard for the continuous improvement, construction and promotion of the NDCCDR.Based on above analysis, we can combine the NDCCDR construction with the CCDRC construction, from point to face, and ultimately improve the overall level of CCDRC in the region.
This paper takes the municipal area as the case study area, analyzes and compares the CCDRC and its primary indices between communities in the jurisdiction area.All the pertinent suggestions are beneficial to the regional functional departments to carry out disaster prevention and mitigation planning, resource allocation, resident mobilization and administrative decisions within the jurisdiction, so as not to complicate the implementation of the CCDRC construction due to coordination across jurisdictions.Suzhou is regarded as a strong economic city in China.Great efforts have been made to prevent and reduce disaster (It is evident from the number of CCDRC.).However, from the analysis results of this paper, the CCDRC of Suzhou is still not satisfactory.It also shows that our CCDRC building has a long way to go.The overall strengthening of China's CCDRC will continue for a long time.
Due to the large number of indices in this paper, there must be redundancy between data, so information reduction should be carried out before evaluation.From the perspective of overall plan of national comprehensive disaster reduction work, we should prepare to respond at different stages during the development of a disaster.Therefore, the resilience capability to disaster should also be included in the index system.These two aspects are the deficiencies of this paper and need to be solved in future research.
Appendix A: Ratio of open space (%) 0.006 The higher the open space (such as green areas) ratio in the community, the stronger the buffering capability of disasters and the settlement capability of post-disaster personnel.
Internal road density (km/km 2 ) 0.013 The road length inside the unit area of the community; the higher the density, the more efficient the disaster prevention and emergency response.

Standard proportion of
building fortifications (%)

0.001
The ratio between the planned residential area and the total area of community buildings; the higher the standard fortification ratio, the stronger the residential defense Note: The indices without a data type annotation are numerical.
of other countries or regions include, -Besides from the open network GIS platform such as Google earth, many open databases for disaster prevention and mitigation have been put into use within many countries, and the spatial information needed for the evaluation of CCDRC can be online obtained.It is convenient to obtain community attribute information through residents' visit, field investigation and non-governmental organization's disaster prevention and reduction report under the model of CBDRM.
Figure2(a) is the nationwide distribution map of NDCCDR in 2017.The map shows that the selected communities are mainly distributed in the relatively developed capital region(Beijing, Tianjin and Hebei, 117), the Yangtze River Delta (Jiangsu, Zhejiang, Shanghai, 241), the Pearl River Delta (Guangdong, 126) and Shandong Province (92), accounting for 38.9% of the total number.Figure2(b)  shows that there are respectively 15 communities in Nanjing and Suzhou cities, which are cities with the largest number of NDCCDR in Jiangsu province (115).The distribution status shows that the NDCCDR construction of Suzhou is in the forefront of China；nevertheless， the construction of NDCCDR is a demonstration project of comprehensive disaster reduction work in China, and it is also a component part of strengthening the comprehensive disaster reduction ability across our country.Taking SND as an example, with 8 communities currently selected as the NDCCDR, the NDCCDR construction work is obviously among the highest in China.However, the region has jurisdiction over 82 communities, and the number of NDCCDR accounts for only 10% of the total community.In addition, the evaluating index system of NDCCDR is imperfect compared with that of CCDRC, and the NDCCDR is not necessarily consistent with the CCDRC.According to above two aspects, we select SND as the research area to calculate and analyze the CCDRC.The characteristics and distribution of the regional CCDRC can partially reflect the current situation and construction direction of CCDRC in China.

Figure 2 .
Figure 2. Distribution status of the NDCCDR in China.

Figure 3 .
Figure 3. Overview of research area and administrative district.In this research, the employed data used for empirical calculation and analysis are provided by Ministry of Civil Affairs of Suzhou and were collected in 2015.The spatial size of SND is 22.65 km2 and the average size of a community is 0.33 km2.There are totally five vector layers: administrative boundaries for towns and communities, and roads, water system, and building footprints (Fig.4).The data were updated and examined before use.TableA1of the Appendix A shows the indices of the open space area ratio, the fortified area ratio of buildings, and the community road density, which were obtained using the statistical computation function of ArcGIS 10.2 software.

Figure 4 .
Figure 4. Image layer of spatial data in the study area.
then divided the grades based on anomaly values to obtain the spatial grade distribution maps for the primary indices (Fig.6).The figure is descending ordered from the upper left to the lower right based on the entropy weight of the primary evaluation indices.It can be seen that the public cognitive capability (Fig.6 (a)) and the rescue and support capability (Fig.6 (b)), despite their maximum weight values(0.381and 0.288,respectively, visible from Table A1 of the Appendix ), show a general poor feature, which is the main reason for the generally weak CCDRC in the study area.

Figure 6 .
Figure 6.Spatial grade distribution map of various primary evaluation indexes in the community and the town of Tongan (in the central part of the study area) have an excellent disaster management capability, but their other primary evaluation indices are weaker.The uneven spatial distribution of the disaster reduction capability is also related to the location.Shishan Street is comprised of new urban villages constructed at the end of the twentieth century; it is bordered to the east by the urban area of Suzhou (in the Gusu District), which has well-developed community facilities and a high population.The Hxiang and Ylian communities are located in the central part of the town of Xushuguan, which has a similar setting as Shishan Street.The towns of Tongan, Dongzhu, and Zhenhu, in the western study area, are located on the edge of the Suzhou urban area.Most structures are houses built by individual farmers or as part of settlement communities; living expenses are relatively low, and there is a large transient population.The CCDRC in these communities is thus relatively weak.

Figure 7 .
Figure 7. Relative level of engineering defense and social economic base support capability in Tongan Town.

Figure 8 .
Figure 8. Curve for the target center degree of rescue and support capability provide human, financial, resource and environmental support for disaster prevention and mitigation is mainly affected by the level of socio-economic development, the amount of disposable financial revenue and the level of infrastructure development in a region.Ratio of population from 18-60 years old (%) 0.009 Ratio of population aged 18-60 to the total registered household registration population in the community (village).the ability to organize and coordinate various forces reasonably and effectively in order to effectively achieve disaster prevention and mitigation, to formulate reasonable policies, systems and mechanisms, and to flexibly use various methods.It is mainly influenced by such factors as the perfection of the legal system, the ability of social mobilization, the ability of scientific and technological support.comprehensive disaster reduction plans are regularly reviewed, and specific improvement measures are formulated to address the insufficiencies.Social mobilization mechanism 0.007 Whether a social mobilization mechanism is established.the means of raising the public's awareness of disaster prevention and mitigation through traditional propaganda methods such as holding propaganda activities for disaster prevention and mitigation, issuing propaganda materials and first aid drills, installing disaster early warning display screens, setting up publicity signals for disaster reduction and other modern communication technologies, and the proportion of volunteers participating in disaster prevention and mitigation can also be reflected.Public ability in this area.materials that have been developed (such as popular science books for disaster prevention and reduction).Disaster early-warning display screen 0.035 Whether the community (village) has a display terminal for disaster early warning (display screen)

Table 1 .
Target center coefficients, entropy weights, and primary target center degrees for the rescue and support capability of communities on Shishan Street.

Table 2 .
Area ratio for the different grades of the primary evaluation indices and CCDRC in the study area (unit: %).

Table A1 .
Evaluation Index system and the entropy weights of the CCDRC.