Google Earth Engine can easily compute statistics on gridded raster datasets. While computing statistics on image datasets is easy, there are some issues that must be taken into account when working with population datasets.
Data used this time:
GHSL: Global Human Settlement Layers, Population Grid 1975-1990-2000-2015 (P2016)
GHSL relies on the design and implementation of new spatial data mining techniques that allow the automated processing and extraction of analysis and knowledge from large volumes of heterogeneous data including: global, granular streams of satellite imagery data, census data, and populations source or voluntary geographic information source.
This dataset describes the distribution and density of the population, expressed as the number of people per cell, with reference years: 1975, 1990, 2000, 2015.
Residential population estimates are provided by CIESIN GPW v4. These estimates are decomposed from census or administrative units into grid cells, informed by the distribution and density of buildings for each corresponding era in the GHSL global layer. (See "Development of a new 250-meter-resolution open and free multi-temporal global population grid").
This dataset is produced in the world Molvid projection method (EPSG:54009).
Dataset Availability
1975-01-01T00:00:00 - 2015-12-31T00:00:00