tomorrowcities/content/articles/data.md

metadata

author: huseyin.kaya
title: Data Formats
description: Description of the data formats used in web application
image: >-
  https://github.com/TomorrowsCities/tomorrowcities/blob/main/tomorrowcities/content/images/data.png?raw=true
thumbnail: >-
  https://github.com/TomorrowsCities/tomorrowcities/blob/main/tomorrowcities/content/images/data.png?raw=true
alt: Data Formats
createdAt: 2023-10-10T00:00:00.000Z
duration: 6 min read
category:
  - general

[TOC]

Data Structures

Tomorrow's Cities Decision Support Environment (TCDSE) is capable of conducting different hazard scenarios on different infrastructures, hence it needs to use several layers with different data input/output formats.

Storage Formats

Tabular If the data does not contain geo-spatial information, then spreadsheet formats such as Comma-Separated Values (.csv) or Microsoft Excel (.xlsx) can be used to store the data.

GeoGSON The primary choice of the webapp to store geo-spatial information is GeoJSON for couple of reasons. First, unlike SHP format, it doesn't require auxillary files so it is easy to transfer or load a GeoJSON file. Secondly, it is supported by QGIS and geopandas Python libraries. Another advantage is that every GeoJSON is a text file and you can always look at the content by any text editor.

GeoTIFF When the intensity measures are given as a mesh of rectangular grid, then the number of points to store can become very large. In those cases, the size of the GeoJSON files will be very large because GeoJSON is a text file and there will be redundant information. It is possible to compress a GeoJSON but a better approach is to use GeoTIFF files. GeoTIFF is useful because it is a raster file and moreover geo-spatial context is already built-in in the format. Additionally GeoTIFF files are lot smaller than GeoJSON files. Please note that GeoTIFF is only advised to represent intensity measures on a mesh grid.

Multi-band GeoTIFF GeoTIFF format also supports multiple bands which becomes handy when we want to store multiple density maps in single TIF file. This is especially useful for using multiple spectral acceleration intensity maps. In order to convey which band corresponds to which spectral acceleration period, the bands in the GeoTIFF are labelled such as "PGA", "SA 0.3", "SA 1", etc.

Layers

The layers supported by tomorrowcities are listed below. In this section, we will cover all of them and provide information so that the users are able to generate data compatible to web application. Unless otherwise mentioned, all geo-spatial layers have WGS 84 coordinate reference system with EPSG:4326.

• land use
• building
• household
• individual
• intensity
• fragility
• vulnerability
• power nodes
• power edges

Landuse

Landuse plan is presented as a GeoJSON format whose attributes are given below:

population	status	zone	densitycap	luf	landuse_lu	area	zoneid	floorarear	setback	avgincome
0	None	Agricultural Zone	0.0	Agricultural Zone	Agricultural Zone	4099.24144134	4	0.0	0.0	None
0	None	Forest Zone	0.0	Forest Zone	Forest Zone	30843.9849839	5	0.0	0.0	None
7081	None	Residential Very Low Density	5.0	Residential-Very Low Density	Agriculture Cum Resedential	1416.16357154	226	0.0	0.0	lowIncome

Buildings

Buildings are the core component of visioning scenarios. The features of the building with some example data are shown below:

zoneid	bldid	nhouse	residents	specialfac	expstr	fptarea	geometry
4	17	41	178	0	RCi+HC+18s+ResCom	111	MultiPolygon

where

• zoneid (integer) refers to the unique identitifer of the zone that building is located in. The features of the corresponding zone is described in a dedicated zone table.
• bldid (integer) is a unique building identifier.
• nhouse (integer) is the number of household in that building.
• residents (integer) stores the number of individual live in the building.
• expstr (string)

Households

Households are defined in a tabular format whose attributes are shown below:

bldid	hhid	income	nind	commfacid
17	12	lowIncomeA	3	3643

where

• bldid (integer) is the building identifier where the household is located in
• hhid (integer) is the unique identifier for the household
• income (string) is the income level of the household
• nind (integer) is the number of individuals living in the household
• commfacid (integer) is the building identifier of the community facility. In Tomorrow's Cities, it is used to define the hospital associated with the household.

Individuals

Individual layer is a tabular data which can be stored in Excel or JSON format. The attributes are:

hhid	individ	gender	age	head	eduattstat	indivfacid
24448	1	2	8	1	3	-1
1	5552	2	2	8	1	4
2	31586	3	2	9	1	5

where

• hhid refers to household identifier where the individual lives in.
• individ is the unique individual identifier
• gender is a categorical variable for gender: 1: male, 2:female
• age is the age category
• head is a binary feature indicating that the individual is the head of the household
• eduattstat is the education level
• indivfacid is the facility (school or workspace) that the individual is associated with

Intensity Measures

Whether it is flood, debris or earthquake, every hazard map should contain at least two properties: a point geometry and intensity measure denotes by 'im'. The data can be provided via GeoTIFF or GeoJSON format. TIFF files should contain CRS information so that the engine could map the coordinates to a common CRS to conduct calculations.

Vulnerability

The engine has a built-in support for Global Vulnerability Model (GVM) used in Global Earthquake Model (GEM). GVM uses an XML format in which one can define several vulnerability functons. There are almost a thousand XML files in GVM repository precomputed for hundred cities from all continents. When a user uploads an XML file conforming to GVM, the engine can parse, display and use it in the calculations.

The GVM's XML format is very simple. For demonstration purposes, we provide a short but structurally-complete one below. Please note that the values do not represent a real vulnerability function:

<?xml version="1.0" encoding="UTF-8"?>
<nrml xmlns="http://openquake.org/xmlns/nrml/0.5">
<vulnerabilityModel id="vulnerability_model" 
                    assetCategory="buildings" 
                    lossCategory="structural">
  <description>Example vulnerability model</description>

  <vulnerabilityFunction id="CR/LDUAL+CDL+DUM/H12/COM" dist="BT">
    <imls imt="SA(1.0)">0.05 0.1 1.2 2.3 5.2 6.6 8.38 15</imls>
    <meanLRs>1e-08 0.0001 0.002 0.06 0.14 0.19 0.6 0.99</meanLRs>
    <covLRs>1e-08 1e-08 1e-08 1e-08 0.62 0.53 1e-08 1e-08 </covLRs>
  </vulnerabilityFunction>

  <vulnerabilityFunction id="MUR+CLBRS/LWAL+DNO/H1/RES" dist="BT">
    <imls imt="SA(1.0)">0.04 0.2 1.2 2.3 5.2 6.4 9.38 15</imls>
    <meanLRs>1e-08 0.001 0.003 0.06 0.15 0.18 0.7 0.99</meanLRs>
    <covLRs>1e-08 1e-08 1e-08 1e-08 0.2 0.6 1e-08 1e-08 </covLRs>
  </vulnerabilityFunction>

  </vulnerabilityModel>
</nrml>