The model for the Water Scarcity Clock is based on water data from grids of the size of 0.5x0.5 degrees. Each grid contains population and overall water availability for that area. Based on this data we calculate the average water availability per capita per year for each grid. The grids will be summed up in order to get the country number. The sum of all countries is the headline number.
Based on these grids, we are able to express what percentage of the population lives in areas with an average water availability of less than 500m³, 500m³ to 1,000m³, 1,000m³ to 1,700m³ and more than 1,700m³. The headline number shown will be the sum of people living in areas with less than 1000m³ per capita per year over all countries.
In order to account for small natural climate variations between years, we use 20-year rolling averages of water availability. Using these averages allows us to provide a more accurate estimate of water scarcity. For example, a region may experience an unseasonably mild winter or perhaps endure particularly heavy snowfall. These years may be outliers and thus misleading when it comes to analyzing and forecasting longer-term trends.
For the population numbers we use SSP2. The water data is based on 3 Global hydrological and 5 General Circulation models. Combining those we get 15 different model combinations and we use the average of all of these. As a result, the figures shown will be independent from the choice of a certain model combination (implicitly assuming a uniform distribution across model combinations).