Riyadh Expansion

This project presents an integrated study of urbanization trends in Riyadh City, Riyadh Province of the Kingdom of Saudi Arabia, by using Geographical Information Systems (GIS) and remote sensing.

The study explores the growth and the directions of urban expansion for five decades from 1972 to 2021.

Study Area

Classifying Method

A medium-resolution (MR) remotely sensed imagery from 30 to 80 m of the Landsat system data, were obtained from the USGS website, was used as the primary source for exhibiting Riyadh city urban expansion. Classifying the Landsat images, using ArcMap 10.8.1, were conducted using the Iso Cluster tool (unsupervised classification).

Bands Combination

(A) Landsat image acquired in 1984 (B) classification using the Iso Cluster tool for the band combination: B3/B4 (C) Classification using the Iso Cluster tool for the band combination: B3/B5. (D) Classification using the Iso Cluster tool for the band combination: composite all bands.

Conducting a test using three different band combinations to decide which one should be used in the Iso Cluster tool.

Band combination composite of all bands was chosen

Landsat 7 Stripes

From July/2003 up to Mars/2013, the only satellite covering the study area was Landsat 7. Even though Landsat 5 was working during that period, the study area, among a few other areas around the world, was not covered

I used a toolset downloaded online to de-stripe the images. Even though I tried many ways to classify the de-striped images, the images could not be classified


An increase of 10.3% in the urban areas from 1972 to 2021. In 1972 the total urban areas were about 614.85 km2, representing 8.1% of the study area. In 2021 the entire urban areas were about 1400.01 km2 (18.4% of the study area).

Thematic maps of Riyadh city 1972-2021

The left maps (a) are reclassified maps into two classes (urban areas and non-urbanized areas) of the output of classifying the Landsat images using the Iso Cluster tool. The urban areas are dark red, and the non-urbanized regions are represented in light blue.

The right images (b) are taken by different Landsat satellites and shown in natural colour.










Accuracy Assessment

Conducting an accuracy assessment by evaluating the overall classification accuracy and Kappa statistics.