2 Determining who is exposed

!UNDER CONSTRUCTION!

The exposed population is estimated by combining global gridded population and hazard datasets. Each grid cell in the global population dataset is classified by urban/rural typology and exposure to any combination of the four hazard events using the thresholds specified in Table 2.

Thresholds defining exposure

Resampling spatial data

The method to analyze spatial data proceeds as follows. First, the 1km GHSL Degree of Urbanisation data is projected to the same coordinate system and resampled to align with GHS-POP population grid. Second, the hazard data are categorized using the intensity thresholds and return periods in Table 2 to define exposed cells and resampled to align with the 3 arcsecond population grid. Note that the hazard data are available in differ-ent spatial resolutions, which largely reflects how localized each type of event is. Resampling all spatial data to the high-resolution grid defined by the population dataset prevents double counting the population exposed to multiple hazards, since each 3-arcsecond population grid cell is exclusively classified as rural or urban and ex-posed or not exposed, as shown in Figure 2.

Combining spatial data

ADD FIGURE