Which water point should we repair to impact the most people?
How many potential customers can I reach with a new water kiosk?
How many people are are within 10km of a health facility?
You can now query population data in maps, dashboards and tables
The population density layers in mWater have been combined into a single, powerful map layer that can also be queried in your tables and dashboards.
So if you want to know how many people live within 1km of an existing or planned water point, you can now use the Population Within expression in mWater to find that out. The same option is available for all geolocated data such as your surveys.
In this guide we'll cover how to build these queries as well as look at some existing use cases in detail.
This feature has been funded by USAID’s WATSAN project, implemented by DAI
Creating a table of water points with estimated population within 1km
Building a population data query
To build a population data query, you only need some geolocated data such as water point site data, schools, health facilities, or surveys with GPS location data.
3) Configure the formula to pick out the location and the desired distance.
You can use the Population Within formula (also called an expression) in many places across the platform, for example in Table widgets if you want to make a list of sites and the population within a range of the site. You can use the same expression in pivot tables or bar charts, as well as on maps as we'll see in examples below.
Note: Population data queries are computationally intensive so please filter your queries to only analyze a few thousand data points at a time to avoid server timeout. Each site or survey row takes 4 milliseconds to calculate, so 1000 sites can be analyzed in 4 seconds. If you try to set a calculation for an entire country, it will time out.
Pick a data source with GPS
Select the formula Population Within
Configure the calculation
Note: To ensure performance, only query a couple of thousand sites or surveys at a time.
Example - Uganda
The Ugandan Water Project triangulates population estimates made by key informants by using queries in mWater to approximate how many people live within 1km / a 30-minute roundtrip of a borehole.
To recreate this approach you can follow these steps:
3) Filter the table down to at most a few thousand sites. You can do this for example by selecting one district or water system they belong to.
4) Add a column for the water point Name.
The system estimates 1,462 people live within 1km of this water point
5) Add a column and select the Population Within formula.
6) Configure the location to point to the GPS Location of the water point site.
7) Set the Within a distance of (meters) field to 1000.
We do not need to select Aggregate results since we are listing each water point separately and are interested in how many people live within 1km of each.
Table: Population estimate within 1km of UWP water points in Buikwe, Uganda
Configuring the Population Within formula to calculate the population within 1000m of the water point GPS
Note: Result aggregation is currently disabled due to a known issue with the Postgres database software.
Note from the Ugandan Water Project:
The population served by water points is an important metric for implementing organizations and donors alike. Several methods exist to estimate the population served by a water point, each with strengths and limitations. As many in the WASH sector are well aware, is not uncommon for population estimates from two separate credible sources to differ significantly.
Acknowledging that no method is perfect, the Ugandan Water Project has committed to corroborating its population estimates using multiple data sources, an approach called triangulation. mWater’s population estimate feature gives the Ugandan Water Project an additional, objective, tool that they use to cross check population estimates provided locally by key informants and existing records.
One way the Ugandan Water Project uses this feature is to estimate how many people live within a set distance of any borehole. While any distance can be analyzed, the Ugandan Water Project typically employs a 1 km cut-off, which approximates the maximum distance for a 30-minute round trip to the borehole, with fetching time included. The 30-minute benchmark is significant because it is a requirement for basic household water access, according to the industry standard JMP Service Ladder.
With improved methods of population estimation supported by this mWater feature, the Ugandan Water Project is better equipped to plan programs and accurately report impact to donors.
Example - Haiti
In Haiti utilities are using this feature to find out how much potential there is for expanding private access or public access. They count the population within 100m of a water point or kiosk and then subtract the known subscribers from this amount to get the number of potential new subscribers.
The potential for new subscribers for public and private access can be estimated from the available data
The table uses Spatial Joins to see how many water points (subscribers) are within 100m of the kiosk already.
Subscribers are tracked as taps in yards or dwellings
The calculation in the table is drawing from the population density data in the system
Example - Population within 10km of a health facility
Population queries are not limited to water points or 1km radiuses. You can choose any data source with GPS location information to run the query on, and you can set the distance as you like. Below is an example of a health facility map with 10km radiuses. By clicking on the area circle, a popup opens providing basic information on the health center, including the calculated estimated population living within 10km.
