Mapping Toronto Flood Events by using Esri Operations Dashboard

Dashboard Web application: Toronto Flood Events 2013-2017

By: Mohamad Fawaz Al-Hajjar

Geovisualization Project, @RyersonGeo, SA8905, Fall 2019

Introduction:

Toronto has been affected by many flood events, but the biggest modern event happened in July, 8th, 2013, when a thunderstorm passed over the city and broke the record when Toronto received huge amount of rain reached to 126mm, that caused major transit delays, power outages, flight cancellations and many areas flooded throughout the city; in order to visualize such phenomena and monitor the number of events per Toronto ward, web application dashboard has been implemented to inactively visualize the historical data, which also could be used to map the real time data as an optimal way to utilize the web dashboards.

Geovisualization Methodology

The technology that has been used to interactively visualize flood events data in Toronto is Esri Operations Dashboard, which was released in December, 2017 and has become an effective tool for the Esri users, which allow them to publish their Web Maps via dashboard by applying simple configuration without writing a single line of code. The project has followed the below methodology.

  1. Data Review and Manipulation

After obtaining the open data from two main sources, TRCA Open Data Portal and Toronto Open Data Portal, with other different data sources which have been reviewed and visualized in ArcMap application 10.7.1 release. Some of the data had to be cleansed, such as Flood Plain Mapping Index and property boundary shapefiles, other data were derived from polygon shapefile “flood-reporting-wgs84” for Toronto wards, where the total number of flood events stored by year from 2013-2017. A derived data-set produced as a point shapefile events points by using generating random point tool from polygon in ArcGIS ArcToolbox.

In addition, another data set have been created, the Property boundaries which have been intersected and clipped with the flood plain feature to generate the flooded properties per ward, which is also spatially joined with the wards to inherit its attributes. that could be configured in the dashboard to show the number of flooded properties per ward.

List of Data-Set Used:

Stormevents (derived from Flood reporting polygon) (Toronto open data)

Property per ward (derived from Property boundary and Flood reporting polygons) (Toronto open data)

Flood Events renamed to (Flood reporting polygons) (Toronto open data)

Toronto Shelters (Toronto open data)

GTA Watercourses (TRCA open data)

GTA Flood Plain (TRCA open data)

GTA Waterbodies (TRCA open data)

2. Data Publishing:

After getting the data ready, map produced in ArcMap where data symbolized then published to web map In ArcGIS Online, which will be the core map for the operation dashboard.

3. Creating the Dashboard:

In order to generate an Esri operation dashboard you need to be a member of ArcGIS Online organization, then have a published Web Map or hosted Feature Layer as an input to the dashboard.

Creating the dashboard went through many steps as described below:

  • Login to your ArcGIS Online organization using your username and password.
  • From the main interface click the App Launcher button as below snapshot
Application Launcher button

or you could also click on your Web Map application under Content in ArcGIS Online then click on Create Web App dropdown list to choose Using Operations Dashboard

Create Web App
  • Create Web App box will be opened to fill Title, Tag and Summary
  • The map will be opened into the dashboard, where you will start to add the widgets you need to your application from the drop-down menu as below snapshot.
  • Widgets will be added and configured as needed.

Toronto Flood Events Dashboard has included the most important widgets (Map, Header, Serial Chart, Pie Chart, Indicator, and List)

Once widget selected, the configuration box will be opened which is easy to be configured then will be dragged to be docked as needed

After adding multiple widgets, an important setting needs to be configured in the Map widget to set what is called an Action Framework, that happens when we change the map extent of the geographic area, then the other dashboard elements such as (Serial Chart, Pie Chart, Indicator, and List) will interactively be changed.

  • From the Map Widget go to Configure button, then select Map Actions tab, hit Add Action drop-down list then filter to choose other dashboard elements from the configuration box. the option When Map Extent Changes appears to let you filter and make action to other elements as well. Indeed, this is the most powerful tool in the dashboard.
  • Another configuration could be made in the Header element where you can insert a drop-down menu to map a certain feature by date, type, area or time, which is easily be configured in the dashboard web application.
  • After configuring all required elements, hit save then you can share or publish your dashboard web application with other users out of your organization.
To access the Dashboard click on the link below
Toronto Flood Events 2013-2017

Geovisualization Project Limitations:

The project was encountered two main limitations:

The data limitation:

Data limitations were taken most of the time to be defined, then after defining the available open data, many data cleansing and manipulation has been taking in terms of changing spatial reference to fit with online maps or changing the data format, which are still limited with the variables used, the derived events point generated randomly from the polygon shapefile “flood-reporting-wgs84” for Toronto wards to show the number of events per Toronto ward, which are not available as points from the main source; even though, the points still not accurate in location, but it give an idea about the number of event per ward boundary in different years.

Technology Accessibility:

It is clearly represented when we use Esri operations dashboard, which is only available to the member of ArcGIS Online organization and whom how have that access, still be able to get the benefits out of it by hitting the published location.

Visualizing the Story of Forest Fires in BC with Operational Dashboards

By: Anderson Webber

Background

2017 and 2018 were the two worst fire years in the BC’s known history, diminishing provincial air quality and destroying healthy ecosystems beyond natural levels. These consecutive record-breaking years have led to many discussions regarding the causes of such fires, with the hope to better understand why these events occurred and hopefully prevent such events from reoccurring. The purpose of this project is to help aid in the understanding of BC wildfires through an interactive summary of the 179,000 wildfires occurring within the province over the last 68 years via an Operations Dashboard.

This image has an empty alt attribute; its file name is UEvIDLI36LdfJ3DBJcZZI1UVB6dJopmn-_2ZbtI7bAB67OThU9KmieQf5J5_ZHrI_8ar62-rI2dNWdogy94w7yWQ99vqxKKxX8BLaQ1mw2VxCVySZelO-5UylqDFJOpKSWpxlK77

Data and Technology

Dashboards have become a very trendy tool for geovisualizations, designed to display location aware visualizations and analytics packaged in an easy to use web or mobile app. ESRI largely markets Operational Dashboards for real-time analytics which aid in such tasks as emergency response, but in this example I will be looking at the dashboard’s utility in understanding a large historical data set; past BC wildfires. BC wildfire data was sourced from the BC open data catalogue containing point locations and attribute data of any fire incident updated on April 1st 2019 with all previous years data since 1950. The following tutorial will allow anybody with an ArcGIS license to replicate this project.

Step 1. Getting the data

Data was downloaded from here:

https://catalogue.data.gov.bc.ca/dataset/fire-incident-locations-historical

Step 2. Data Cleaning

Beyond the date of report and location, the wildfire CSV contained information for each fire including its size, cause and which zone it occurred in… In order to create the widgets I wanted, the data had to be cleaned a bit. To determine which months the fires were worst for example, the date field in the format YYYYMMDD had to be split into three separate fields; Year, Month and Day. This could be done in any data manipulation software such as SQL server, alterex or excel. The simple query needed to select included:

=LEFT(D2,4) to select year

=MID(D2,5,2) to select month 

=RIGHT(D2,2) to select days

Step 3. Hosting layer file online

Once data was split it had to be uploaded into ArcGIS online as a hosted layer file in order to be brought into the dashboard webapp. Hosted layers allow for the uploading of large files which can be used in web, desktop or mobile apps. In order to post hosted features you must be part of an ArcGIS organization, and have privileges to post hosted layers. When you add a layer through ArcGIS Online, this is what your options should look like to host a layer:

Step 4. Making a web map

Before creating the dashboard itself, the user must create a web map first. This is done by clicking the “Create” tab then clicking the “Map” button. Once the map is open simply click the “Add +” button and bring in the hosted feature classes we just created. Now you have made a web map with the fire data. Edit symbology and add any other layers you would like in this step. I chose a dark basemap and red ‘firefly’ symbology. 

Step 5. Adding fields

Depending on what you want your dashboard to display, more data cleaning and manipulating could be done at this step. As I wanted to see what the different sizes of fires were within BC, I created an ESRI Arcade expression which would calculate a field with ranges. To do this I created a new field with data type ‘Text’ in the Fires table called ‘Sizes’ and calculated the field with: 

iif($feature[“SIZE_HA”] <=0.25 , “<0.25”,

iif($feature[“SIZE_HA”] <=10, “0.25-10”,

iif($feature[“SIZE_HA”] <=100, “10-100”,

iif($feature[“SIZE_HA”] <=1000, “100-1,000”,

iif($feature[“SIZE_HA”] <=10000, “1,000-10,000”,

iif($feature[“SIZE_HA”] <=50000, “10,000-50,000”,

iif($feature[“SIZE_HA”] >50000, “>50,000”, 0)))))))

Step 6. Creating the dashboard

Now we are ready to make the dashboard! On the same webmap which was just made, click the Share button -> Create new Webapp -> Operations Dashboard. You can call it whatever you like. Now you have a dashboard shell.

Step 7. Add widgets

Now comes the fun part. Click the “+” dropdown on the top left of the dashboard and add whatever widgets you want. Widgets can be dragged, resized and stacked allowing for a high level of customization.

Step 8. Making charts interactive

To make charts interactive, within the widget configured in the ‘actions’ toolbar and add whatever action you like. This means that selecting a bar on the chart below will change the points, indicators and all accompanying data visualizations to the months chosen. The same methods can be applied to any other aspect, including the map itself.

Keep playing around with the widgets. You can also add images. The final product http://arcg.is/1WDSyy. A screenshot can be seen below:

Limitations

Limitations for this project regarded both the data and software itself. For starters, in order to create an Operations Dashboard, you need an authorized ArcGIS account which is not free and therefore accessible to everyone. Another major limitation has to do with the size of the data set. With almost 180,000 fire points, the rending of these points online can encounter problems such as lag and may even may crash if you have limited ram. The third limitation regards the projection. ArcGIS Online defaults to a global WGS 1983 projection, which is not optimal for looking presenting data at the provincial level. Finally, the user’s screen size also has a major impact on the experience of the user. Although it can be opened on mobile devices and tablets, the dashboard becomes more limited, as graphics and titles are cut off or crushed together, taking away from the visual appeal and usability of the dashboards.

Visualizing Tornado Occurrences in the United States of America from 1950 to 2018.

Kellie Manifold

Geovis Project Assignment @RyersonGeo, SA8905, Fall 2019

With a large growing population in the United States of America more people are being affected by tornadoes each year.  Tornadoes are a mobile, destructive vortex of violently rotating winds with a funnel-shaped cloud and immense wind speeds. Over the past several decades tracking tornadoes has become more common, and there has been an increase in the amount of tornadoes recorded each year. With this, tornado tracking has become more frequent, as more technology becomes available.  This increase in tornado occurrences and the technology to track and record them has resulted in a large dataset of tornado occurrences throughout the United States of America. All recorded tornado occurrences from 1950 – 2018 are kept in a database on NOAA National Weather Service’s website. This data set is very large and difficult to visualize.

To help visualize the distribution of tornado occurrences in the USA, the dataset was used to make an ESRI Dashboard. ESRI Dashboard was chosen because it is a good data visualization tool. Dashboards are very user friendly and allows for the creation of many charts, tables, and indicators to display large amounts of data, on a single web page. The dashboard also allows for user interaction, so that they get a better understanding of the data at hand.  

The following steps were used to create the ESRI Dashboard on tornadoes in the United States of America.

First it important to make sure that you have an ArcGIS online account.

The next step is to gather your data. The data for this dashboard was downloaded from the NOAA National Weather Service and the National Centers for Environmental Protection. The file that was downloaded contains the location of all recorded tornadoes in the USA from 1950 – 2018.

Next; Import this data into ArcGIS Pro. The points will be shown through their Latitude and Longitude location. For the purpose of this project only the contiguous USA was looked at, so Puerto Rico, Alaska and Hawaii were removed. This was done through a select by attribute query which can be seen below.

These states were then removed from the map, and attribute table.

The next step to creating the dashboard is to upload the layer to your ArcGIS online account. This is done through publishing your map as a web map. This will then add the layers used onto your ArcGIS online account.

The following steps are through ArcGIS online. Once you are signed into your account, and the published web layer is in your contents, you can now create the dashboard. First, it prompts you to choose a theme colour, light or dark. Dark was chosen for the tornadoes dashboard as it looks more visually appealing.

From here you can add the web layer you published, in this case the tornado layer, and add any other widgets you would like. Such as serial chart, pie chart, indicator, or a list.

Once you decide which widget you would like, it is important that in the Actions tab in the configuration window that all other widgets and filters are selected.

Having these selected will allow for all other map elements and widgets to correspond to changes selected by the user. Make sure this is done for all the widgets that are added. In the case of tornados the widgets used include category selectors for each state, year, or month, an indicator to show the amount of tornados given the selection criteria, a pie chart to show the months tornados occurred, and a serial chart to display the amount of tornados per year.

Once all the widgets are added you can drag them to rearrange the appearance on the dashboard.

Once you have configured all your widgets to get the layout you want, you have successfully created a dashboard.

Thank you for following along!

The final product for this process can be seen here: http://arcg.is/1bDib10

Please take a look and interact with the dashboard.