Visualizing access to health in Haiti

In this presentation, we will look at visualizations of access to health facilities in Haiti.

The datasets we are working with are a population estimate, 2003, from the US Census Bureau. The resolution of this dataset is 100m² per cell/pixel, but the analysis in this presentation will be done at an aggregation of 1km² per pixel.

The health facility dataset is from GeoCommons.

The total population represented is 8.4M; Haiti's population has grown to 10M+ in 2013. We will be working at a resolution of 1km² per cell; shown above.

Here, health facilities (in red) have been overlaid on top of the population layer. The dataset, from GeoCommons, was primarily collected in 2005 and revised in 2009.

Below, the redder the pixel, the farther the people living there have to travel to a health facility.

Methodology: straight line distance in m (on UTM projection).

A different view: The longer the line, the farther people have to travel to the nearest health facility--lines connection population points to health facilities. (Inspired by [1])

On this map, population is represented by color and lines are the same as before. Deeper purple lines indicate bigger populations.

• These maps are the first steps of analysis, more sophisticated algorithsm can dive deeper, including into recommendations.

• The pre-requisite is good data; for garbage in, garbage out.

• The analysis should be made interactive for local-level users.

• And finally, the analysis, however sophisticated, cannot exist in a vacuum. It needs to be interpreted and understood by those who actually know on-the-ground realities.

We are working on all of these problems at the Sustainable Engineering Lab.