Mamadou talked about his idea for an app or website that can provide information about urban flooding to city dwellers around the world. In many fast-growing cities in the developing world, people are living and building in flood-prone areas. Mamadou has developed a GIS model that can predict the places in a city that will flood. This kind of urban flash flooding is destructive and deadly, and can occur in areas that don’t look risky to the eye.
The vision is for a smartphone app or site that anyone in the world can use to assess their area’s risk for urban flooding. This could save lives!
Our initial plan is to host Mamadou’s urban flood model on the Google Earth Engine platform. Google Earth Engine is a cloud platform that combines computing power with many high-quality geography data sets. There is more information about Mamadou’s urban flooding project in our May meeting notes.
Next, I gave a small demo of Pandas, a data processing library for use in Python. The Pandas DataFrame has become a standard way of manipulating and transferring data, used by analytic, statistical and visualization packages.
We downloaded a dataset of Baltimore City employee salaries, from and the Data.gov repository. I showed some very basic features of Pandas - loading a csv file in to a DataFrame, exploring the data. We converted a dollar amount that was stored as a string with a leading dollar sign in to a float. Here’s an example:
df[['AnnualSalary', 'GrossPay']].replace('\$','',regex=True).astype('float')
Watch for more news about the urban flood watch project here on the mailing list, and please get in touch if you want to help. We will need people with many different skills to turn this vision in to a reality. We welcome your involvement in creating a tool that will help people around the world understand the areas that are at risk of flood in their cities and neighborhoods.