We present here a web-based decision support tool for rural water
supply, intended to be used by planners or community members. This
tool can be used to understand current water source choice and demand.
This tool can also be used to simulate different scenarios such as
changes in the price, quality, or placement of available sources.
Get started by following the prompts below to: (1) upload water source data, (2) input village characteristics, and (3) execute the model to make source choice and water demand predictions.
This tool requires users to upload data on all available water sources
in their study site. If using a GeoJSON file, users should include
information on the name, price, and quality of each source using the
'properties' object; longitude and latitude information should be
included using the 'geometry' object. (Click
to download an example JSON file for rural Kenya that you can use to
run the model; you will want to save it locally to your computer). The
model can also import KML files; the name of each source should be in
the 'name' object and the longitude and latitude information should be
included using the 'coordinates' object.
Select the data file for your study site below, or select the JSON example file you saved to your computer:
Once you have selected your data file, click the 'Import' button to display your data on the map and in the table below.
Notice the table below is editable. Use this feature to correct
mistakes in the data, or add missing source characteristics. (Note
that because KML file formats do not allow users to include source
characteristics, KML file users will need to manually change the
source price and quality characteristics from their defaults
(default settings: price=0, and quality='fair')).
If you need to delete a particular source, use the 'delete' button in the rightmost column. If you need to add a new source to your data, click the 'Add' button. The new source will appear on the map, and in the last row of the table. Edit the default source characteristics as necessary.
To calculate demand, we need to know the population at each location
in the study site. The map below is overlayed with grid cells that
will be used to input population density data. The default population
in each cell is
households per grid cell.
To indicate a higher population density in a particular grid cell, left-click on the cell. To indicate a lower population density in a grid cell, right-click on the cell. When you click or hover over a particular cell, the population density of that cell will be printed below.
grid cell. Note: you can zoom in and/or use the 'Satellite' view to
help approximate the population density.
To calculate demand, we also need to know the prevalence of private sources (like a private well or piped connection) at each location in the study site. Just as before, the map below is overlayed with grid cells, but now you will input the percent of households that primarily use a private source for water collection. The default percent of households that primarily use a private source in each cell is percent.
To indicate a higher prevalence of private source usage in a particular grid cell, left-click on the cell. To indicate a lower prevalence of private source usage in a grid cell, right-click on the cell. When you click or hover over a particular cell, the prevalence of private source usage in that cell will be printed below.
Percent of households using a private source:
Lastly, we need know how rain water consumption changes with the seasons. Use the graph below to input the percent of household water demand that is supplied by rainwater in each month of the year.
We now have all the information we need to predict: (1) the source
that each household will primarily use, (2) household demand, and (3)
aggregate demand at each source. For more information regarding the
underlying modeling, we refer you to Wagner et al. 2019 (link).
Default parameters for the simulation can be edited below:
Quality 'good' coefficient:
Quality 'poor' coefficient:
Click the 'Simulate' button to simulate household source choice and demand.
This may take several minutes.