High‑Resolution Multi‑Temporal Mapping of Global Urban Land

Type & Source

A groundbreaking multi‑temporal global impervious surface dataset derived from Landsat imagery spanning the years 1990 to 2010, in five‑year intervals. This product is the first global dataset of its kind offering 30‑meter resolution over a multi‑decadal timescale.

Purpose

Developed to address the limitations of existing global urban land products, which typically offer coarse spatial resolution (500–1000 m), inconsistent definitions, and limited temporal coverage. The dataset supports long‑term research into global urban expansion and its environmental impacts.

Methodology

1. Built on Google Earth Engine, leveraging its cloud-based processing power and complete Landsat archive.
2. Uses an automated extraction approach based on the Normalized Urban Areas Composite Index (NUACI).
3. Includes region-specific calibration based on the “urban ecoregions” stratification framework from prior literature.
4. Demonstrated high accuracy across regions, with global Kappa values between 0.4280 and 0.4953, and comparable performance in country-level assessments (e.g., ~0.3306 for China, ~0.4163 for the U.S.).

Citation

Liu, X., Hu, G., Chen, Y., Li, X., Xu, X., Li, S., Pei, F., & Wang, S. (2018). High-resolution multi-temporal mapping of global urban land using Landsat images based on the Google Earth Engine Platform. Remote Sensing of Environment, 209, 227–239. https://doi.org/10.1016/j.rse.2018.02.055.

Access & Documentation

http://www.geosimulation.cn/GlobalUrbanLand.html