On Google Satellite Imagery, I've created a layer with objects marking houses in a 1 km proximity to large streets in a rural setting using QGIS with the Openlayers Plugin. I would like to identify the top 10 most densly populated hotspots (most houses within a 1km circle) along app. 100 km accumulated street.

I am sure there is already much Information about such task on the web, it would already be very helpful to give me the right vocabulary for what I am trying to do to be able to do my own research.

Backround Information: My Task is to identify high demand spots for water in an underserved area in Western Africa. There is x number of water kiosks installed, connected by roads with an average distance of about 20km from one another. We would like to set up small water selling Points along these roads and therefore need to find suitable Locations which as many People as possible can reach by foot (i.e. 1 km max. distance).

  • 2
    Generate a point density layer from your house locations. Every pixel in the point density layer will show the total number of houses within 1km. Here is a link to doing a point density. digital-geography.com/point-density-raster-qgis. Further processing might include converting your lines to polygons ( or generating points along your line) and doing doing zonal statistics on the resulting vectors to more clearly define locations. – GBG Aug 2 '18 at 16:26
  • You might also want to look at the Heatmap option in symbology. It gives a visual representation of point density. This is useful if you want to play around with different parameters and get a quick comparison. – csk Aug 2 '18 at 18:18
  • The point cluster renderer (also a layer style option) could also be useful, as it gives a point count for point clusters. – csk Aug 2 '18 at 18:26

My suggestion is to simply count houses which falls within 1km-radius buffer zone from points on the road.

(1) Create a point layer, at regular interval along the road, each represents candidate for new water selling point.

Tool: Create points along lines (Processing Toolbox | QGIS geoalgorithms | Vector geometry tools), or Convert lines to points (Processing Toolbox | SAGA | Vector point tools).

The interval can be 1km, 0.5km, or even shorter ... depend on how detail you need to analyze. Add an id field to newly created point layer so that you can identify the location later.

(2) Create buffer around the point

Tool: Fixed distance buffer (from menu Vector | Geoprocessing Tools), with distance = 1000 (m).

(3) Join buffer layer and house points layer to get Counts.

Tool: Join attributes by location (from menu Vector | Data Management Tools), Target layer is your buffer while Join layer is your house points. Select Take summary of intersecting features as the Attribute summary.

(4) Open attribute table of the result Joined layer. Sort (descending) the table by count field and select top 10 rows. See id you have highlighted. On map selected features turn yellow which will help visual check.

enter image description here

(example: selected top three rows)

(5) If needed, save selected features as a new file.

The above workflow is based on QGIS2, some tool name may differ if you use QGIS3.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.