Skip to content

Using Data for Real-Time Flood Response

Presented by Lisa Carroll
in Environment, Technology

Sponsored by Microsoft Canada

The Okanagan Valley is a picturesque region featuring mountainous landscapes, quaint lakeside communities, and award-winning wineries. But in 2017, the region experienced devastating floods. The floods led to widespread damage of residential, commercial, and natural areas, many of which are still recovering. For the City of Kelowna, these events reinforced the need for enhanced flood management capabilities to allow for a more proactive response to fluctuating weather and water levels.

Collaboration in Water-Level Data Modernization

Flood management in the Okanagan is not a simple task. There are many ever-changing elements to consider that need to be carefully monitored. These include rainfall and the spring snowmelt (freshet season) feeding into the surrounding rivers and lakes. In an emergency, the Central Okanagan Emergency Operation Centre (EOC) responds. The city knew that there had to be a way to use data analytics and advanced technologies to help their regional EOC in advance of an emergency flood situation. This led to a collaboration with Microsoft to embark on their water-level data modernization project using Microsoft Azure IoT Hub and Azure Stream Analytics.

The Mayor by George B. Cuff

“As climate change impacts have started to affect our region more significantly, we recognized the need to be more data-centred in our decision making and not simply rely on what’s out there,” said Kelowna City Manager Doug Gilchrist. “In other words, we had to take control of our own destiny and invest in the tools needed to investigate, manage, and monitor our water systems at a more detailed level.”

To do this, the city collected data from 24 diverse sources that included real-time and historic:

  • streamflow rates;
  • lake levels;
  • snow depths;
  • precipitation; and
  • temperatures.

They then cleaned and standardized the data to make it consistent before storing it in a local database that also had real-time data feeding in (from the many water meters in the area). With this data, they were able to create a Microsoft Power BI dashboard to monitor the near real-time situation for all lakes and creeks in the area.

The team also used predictive analytics to help with the flood protection planning. They created a predictive model to determine if Lake Okanagan was once again at risk of flooding. After some calibration, the model was reliably predicting the daily lake level rise within five millimetres of what was actually occurring. This model showed that the lake would in fact flood again, but this time they had seven to 12 days lead time. This allowed the city to get a head start on flood prevention measures and protect the surrounding areas and their residents.

Planning for Future Data Modernization

The City of Kelowna unfortunately has other challenges to face beyond flood response. For example, record breaking temperatures and wildfires in the area. After the success seen with the water-level data modernization project, they may soon turn their attention to other environmental issues.

“We would be really interested in finding a solution to forest fire behaviour,” says Gilchrist. “Obviously, this is completely different and less predictable, but we’d like to explore how best to invest in technologies to assist with forest fire prevention and mitigation.”

Guide to Good Municipal Governance SE Web banner

To learn more, visit: https://www.microsoft.com/en-ca/sites/municipalities/MW

Municipal World Insider and Executive Members: You might also be interested in Caroline Jackson’s article: Future-proofing for the realities of climate change.


Lisa Carroll is the Public Sector Lead for Microsoft Canada.

Related resource materials:

Next Story
See All Feature Stories

A straight path to the circular economy