Analyzing the extent and use of impervious land in rural landscapes


Loading...

Date

2024-12

Publication Type

Journal Article

ETH Bibliography

yes

Citations

Altmetric

Data

Abstract

The amount of impervious surface is increasing rapidly worldwide. Although urban expansion has been studied extensively, the alteration of impervious land cover in rural regions remains under-examined. In particular, insights into the utilization of these sealed surfaces are crucially needed to unravel the underlying dynamics of land use changes beyond urban areas. This study focuses on rural regions from a Swiss case study and presents an analysis of the use of sealed surfaces in such regions, rather than solely quantifying the extent of sealed surfaces. Utilizing a synergistic approach that merges detailed cadastral plans with very-high-resolution remote sensing imagery and sophisticated deep learning algorithms, we characterized the uses of sealed surfaces, including buildings and their surroundings. Our findings reveal that 2.1 % of the study area’s rural regions comprises sealed surfaces - an area comparable to the sealed surfaces in the urban regions. Within these rural regions, transport infrastructure represents 68 % of this impervious surface. Buildings account for 12 %, and their surroundings, constituting 13 %, are utilized primarily for agricultural purposes, including farming and livestock activities. The deep learning approach achieved a classification accuracy of 72 % for a shallow model and 79 % for a deeper model, indicating that mapping building types is possible with reasonable accuracy. The outcomes of this study underscore the critical need to factor in the presence and utilization of impervious land cover within rural regions for the sustainable management of land resources.

Publication status

published

Editor

Book title

Volume

5 (4)

Pages / Article No.

625 - 636

Publisher

Elsevier

Event

Edition / version

Methods

Software

Geographic location

Date collected

Date created

Subject

Soil sealing; Rural areas; Very-high-resolution aerial imagery; Building types; Convolutional neural networks; Cadastral data

Organisational unit

03823 - Grêt-Regamey, Adrienne / Grêt-Regamey, Adrienne check_circle

Notes

Funding

Related publications and datasets