flood-housing-sindh README

Identifying flood-resilient building plots, with an example in Hyderabad, Sindh, Pakistan.

This project is intended to be a tool in finding flood-resilient building plots in areas without readily accessible flood maps. This tool could be used by building developers or financial institutions to indentify and preliminarily assess potential building locations prior to a site visit, especially when they may be unfamillar with the area.

While this tool is not intended to replace local expertise or on-the- ground surveying, this tool can be used as a preliminary step to streamline the identification of potential building plots.
It is meant to

• Reduce the time and labor required, compard to traditional worklflows
• Compensate for lack of flood maps
• Provide low-bias assessment of potential properties
• Improve base-level donor understaning of the area

Potential building sites are filtered based on ground cover data, then scored simply using elevation, proximity to waterwas, and proximity to the city center. The final product is an interactive map of potential building sites in the specified area. This projects uses Python and QGIS-LTR (both free to the public).

I have created the sample of this project using a case study from Hyderabad, Sindh, Pakistan. In recent years, the Sindh region has experienced flooding which has damaged or destroyed millions of houses, and done nearly 40 billion dollars in damages. Thoughtful building location, in tandem with flood-resilient building design, wil be key in ensuring fewer people are displaced in future flood events.

---

Sources contains information on the data used and where it was obtained and the sources of key documentation used in the writing of code. This also contains download links to files too large to commit to GitHub.

Instructions contains written instructions for the entirety of the project. This includes input and output files for each step, as well as short description for python files and full instructions for steps carried out in QGIS.

Data contains all the data you need, including many of the files you would create in this process. The exception is large source files, which are linked in in the sources document.

---

Step 0: Finding City Center (optional, Python)

This will find the center of the city based on any desired attributes- in this case, banks have been used. This is optional as any arbitrary point could be used.

• HDB_Banks.csv banks and location sin Hyderabad
• geo_center.py finds the center point

Step 1: Create rings from city center (Python)

This creates rings of specified distance from the city center, and creates tools to trim the other files down to size.

• center_rings.py makes rings on the center point

Step 2: Create waterways & waterway rings (QGIS and Python)

This isolates waterways from ground cover files, then creates buffers of specified distance from them.

• PAK-FCO.LCSINDH.tif ground cover data
• water_rings.py makes rings around waterways
• water.tif (created in this step)
• ringmask.tif (created in step 1)

Step 3: Find Buildable Land (QGIS)

This step isolates buildable land from ground cover files.

• PAK-FCO.LCSINDH.tif ground cover data
• ringmask.tif (created in step 1)

Step 4: Create elevation bands (QGIS)

This step creates an elevation file.

• 10n060e_20101117_gmted_mea075.tif elevation data
• ringmask.tif (created in step 1)

Step 5: Put it all together (QGIS and Python)

This step merges the tif files and suses them to score potential building areas. It also prepares this score file for visualization.

• build.tif (made in sstep 3)
• elev.tif (made in step 4)
• water_rings.tif (made in step 2)
• rings.tif (made in step 1)

Step 6: Visualize (Python)

This step makes a final, interactive map.

• islands.gpkg (made in step 5)

To view the final map, either pull and run islands.gpkg or download and open map.html. The final map is scored 1-10, with 10 being hte most favorable.