Catalogue of residential buildings and classification according to seismic vulnerability in the city of Murcia (Spain).

<p>The residential building catalogue and seismic vulnerability classification is a supplementary dataset associated with the paper entitled “Classifying Buildings According to Seismic Vulnerability Using Cluster-ANN Techniques. Applied to the City of Murcia, Spain</p>
<p></p>
<p>The goal is to identify building patterns in construction typologies (CTs), related to both structural and non-structural elements that make up a building, called Building Typologies (BTs). The analysis identifies urban modifiers associated with different CTs, and using machine learning techniques, different building configurations are obtained, which are defined as Building Cluster Typologies (BCT).</p>
<p></p>
<p>The results of the BCT identification include the definition of BCT attributes, their representativeness in the study area, the spatial distribution and seismic vulnerability. 11 BCTs associated with 5 CTs are identified, as well as the distribution of seismic vulnerability ranges at the census section scale in the city of Murcia.</p>
<p></p>
<p>In conclusion, the identification of BCTs allows for exposure assessments that take into account the particular characteristics of buildings, grouped according to specific configurations, and facilitates future seismic vulnerability evaluations.</p>
<p></p>
<p>4. Description of the dataset</p>
<p></p>
<p>The dataset is the final data from the research titled: “Classifying Buildings According to Seismic Vulnerability Using Cluster-ANN Techniques. Applied to the City of Murcia, Spain.”</p>
<p></p>
<p>The database is the result of the identification of the different CTs and BCTs in the city of Murcia. The data is organized into 7631 records of residential buildings, which have urban modifier attributes associated with each CT. The combination of CTs and urban modifiers define the different BCTs. In total, for each building record (rows), a CT, the year of construction, ten urban modifiers, and the BCT identification result are associated (in columns).</p>
<p></p>
<p>In addition, the distribution of the Seismic Vulnerability Index (Iv) of the buildings at the census blocks scale is provided. The Iv is expressed in intervals. Minimum, maximum, and average values are provided.</p>
<p></p>
<p></p>
<p>1. Methodology</p>
<p></p>
<p>The methodology for obtaining the present database contains several steps:</p>
<p></p>
<p>1. Extraction of building data from the Spanish Cadastre (https://www.sedecatastro.gob.es/) and configuration of an initial database. </p>
<p></p>
<p>2. Identification of construction typologies (CT) using a typology membership probability matrix for Murcia (RISMUR, 2014). </p>
<p></p>
<p>3. Error detection and correction with remote field work.</p>
<p></p>
<p>4. Assignation of urban modifiers of the buildings in the study area (Martinez-Cuevas and Gaspar-Escribano 2016; Martinez-Cuevas et al., 2017) with remote field work in combination with different GIS processes (Lantada et. al. 2010). The coding of new attributes (urban modifiers) was done according to the GEM taxonomy (https://taxonomy.openquake.org/). </p>
<p></p>
<p>5. Application of a Two-Step Cluster analysis identify clusters. The resulting clusters (CTs with the corresponding urban modifiers) are called Building Cluster Typologies (BCT). The BCT field was assigned to every record of the database.</p>
<p></p>
<p>6. Assignation of vulnerability index (Iv) value to each building (Benedetti and Petrini, 1984; Angeletti et al., 1988; Milutinovic and Tredafiloski, 2003; Lagomarsino and Giovinazzi 2006) and calculation of the average Iv value for the census block level.</p>
<p></p>
<p>FILES</p>
<p>--------------------------</p>
<p>The dataset consists of two files in csv format. </p>
<p></p>
<p> -Buildings_01.csv: database with 7631 building records. Each record consists of an ID, centroid coordinates in ETRS_1989_UTM_Zone_30N and thirteen attributes.</p>
<p></p>
<p> -Iv_CB_02.csv: database with 158 records corresponding to the census blocks of the study area. Each record consists of an ID, centroid coordinates in ETRS_1989_UTM_Zone_30N, census block code and three attributes.</p>
<p></p>

本住宅建筑名录与地震易损性分类数据集为论文《基于聚类-人工神经网络技术的建筑地震易损性分类——以西班牙穆尔西亚市为例》的配套补充数据集。

本数据集旨在识别构成建筑的结构与非结构要素相关的建筑类型（Construction Typologies, CTs）模式，即建筑类型（Building Typologies, BTs）。分析过程将识别与不同建筑类型相关联的城市修正因子，并借助机器学习技术得到不同的建筑配置，此类配置被定义为建筑聚类类型（Building Cluster Typologies, BCT）。

建筑聚类类型识别结果涵盖建筑聚类类型属性定义、其在研究区域内的代表性、空间分布特征与地震易损性。本研究共识别出与5种建筑类型相关联的11种建筑聚类类型，同时得到穆尔西亚市普查街区尺度下的地震易损性区间分布情况。

综上，建筑聚类类型的识别可开展兼顾建筑特定特征的暴露度评估（依据具体配置进行分组），并为后续地震易损性评价提供便利。

4. 数据集说明

本数据集为研究《基于聚类-人工神经网络技术的建筑地震易损性分类——以西班牙穆尔西亚市为例》的最终成果数据。

本数据库为穆尔西亚市不同建筑类型与建筑聚类类型识别的成果。数据集共包含7631条住宅建筑记录，每条记录均附带对应建筑类型的城市修正因子属性。建筑类型与城市修正因子的组合即可定义不同的建筑聚类类型。每条建筑记录（行）共包含建筑类型、建造年份、10项城市修正因子以及建筑聚类类型识别结果共13项属性字段。

此外，本数据集还提供了普查街区尺度下的建筑地震易损性指数（Seismic Vulnerability Index, Iv）分布情况。地震易损性指数以区间形式表示，同时提供最小值、最大值与平均值。

1. 研究方法

本数据库的构建流程包含以下步骤：

1. 从西班牙地籍系统（https://www.sedecatastro.gob.es/）提取建筑数据，并构建初始数据库。

2. 借助穆尔西亚市建筑类型归属概率矩阵（RISMUR, 2014）识别建筑类型（Construction Typologies, CT）。

3. 通过远程实地作业开展误差检测与修正。

4. 结合远程实地作业与多种地理信息系统（Geographic Information System, GIS）处理流程（Lantada等, 2010），为研究区域内的建筑分配城市修正因子（Martinez-Cuevas & Gaspar-Escribano, 2016; Martinez-Cuevas等, 2017）。新属性（城市修正因子）的编码依据全球地震模型（GEM）分类法（https://taxonomy.openquake.org/）完成。

5. 采用两步聚类分析识别聚类组，所得聚类组（包含对应城市修正因子的建筑类型）即被定义为建筑聚类类型（Building Cluster Typologies, BCT），并为数据库所有记录分配建筑聚类类型字段。

6. 为每栋建筑分配地震易损性指数（Iv）值（Benedetti & Petrini, 1984; Angeletti等, 1988; Milutinovic & Tredafiloski, 2003; Lagomarsino & Giovinazzi, 2006），并计算普查街区尺度下的地震易损性指数平均值。

数据集文件
--------------------------

本数据集包含2个CSV格式文件：

- Buildings_01.csv：包含7631条建筑记录的数据库。每条记录包含ID、ETRS_1989_UTM_Zone_30N坐标系下的质心坐标以及13项属性。

- Iv_CB_02.csv：对应研究区域内158个普查街区的数据库。每条记录包含ID、ETRS_1989_UTM_Zone_30N坐标系下的质心坐标、普查街区编码以及3项属性。