For many people across the world, risk of flooding is a yearly concern, impacting both home life and work life when roads become blocked off and streets are flooded. Floods are a growing problem, with climate change and a reduction in green space increasing the risk of extreme weather events and water runoff.
Our idea for this hackathon was to use the range of functionality within Fabric to surface river level and rainfall data to allow a holistic view of potential future flood risk within a given area. This could then incorporate other data sources such as news reports or Twitter/X posts.
We aimed to build a flood prediction system using input from the UK Environment Agency (EA) public APIs. The dataset contains:
Flood Warnings / Events
Flood Area mapping data
Stations for measuring river height and flow levels, rainfall and temperature
For the purposes of this project most of the data ingested is static, with flood warnings and station readings updating daily.
We have created:
Pipelines to ingest the data to the Lakehouse and Warehouse
Stored Procedures to transform the data into a dimensional model
Notebooks to build the ML model and predictions
Power BI report to present the results
Project name
MS Fabric Flood Predictions
Description
For many people across the world, risk of flooding is a yearly concern, impacting both home life and work life when roads become blocked off and streets are flooded. Floods are a growing problem, with climate change and a reduction in green space increasing the risk of extreme weather events and water runoff.
Our idea for this hackathon was to use the range of functionality within Fabric to surface river level and rainfall data to allow a holistic view of potential future flood risk within a given area. This could then incorporate other data sources such as news reports or Twitter/X posts.
We aimed to build a flood prediction system using input from the UK Environment Agency (EA) public APIs. The dataset contains:
Flood Warnings / Events Flood Area mapping data Stations for measuring river height and flow levels, rainfall and temperature For the purposes of this project most of the data ingested is static, with flood warnings and station readings updating daily.
We have created:
Pipelines to ingest the data to the Lakehouse and Warehouse Stored Procedures to transform the data into a dimensional model Notebooks to build the ML model and predictions Power BI report to present the results
Project Repository URL
https://github.com/methodsanalytics/ma_fabric_hack_together_2024/
Project video
https://github.com/methodsanalytics/ma_fabric_hack_together_2024/blob/main/VideoSubmission/Fabric%20Hackathon%20Video.zip
Team members
CookePeter,l-barca,thompson124