Mapping the Land Cover of Africa at 10 m Resolution from Multi-Source Remote Sensing Data with Google Earth Engine

Abstract

The remote sensing based mapping of land cover at extensive scales, e.g., of whole continents, is still a challenging task because of the need for sophisticated pipelines that combine every step from data acquisition to land cover classification. Utilizing the Google Earth Engine (GEE), which provides a catalog of multi-source data and a cloud-based environment, this research generates a land cover map of the whole African continent at 10 m resolution. This land cover map could provide a large-scale base layer for a more detailed local climate zone mapping of urban areas, which lie in the focus of interest of many studies. In this regard, we provide a free download link for our land cover maps of African cities at the end of this paper. It is shown that our product has achieved an overall accuracy of 81% for five classes, which is superior to the existing 10 m land cover product FROM-GLC10 in detecting urban class in city areas and identifying the boundaries between trees and low plants in rural areas. The best data input configurations are carefully selected based on a comparison of results from different input sources, which include Sentinel-2, Landsat-8, Global Human Settlement Layer (GHSL), Night Time Light (NTL) Data, Shuttle Radar Topography Mission (SRTM), and MODIS Land Surface Temperature (LST). We provide a further investigation of the importance of individual features derived from a Random Forest (RF) classifier. In order to study the influence of sampling strategies on the land cover mapping performance, we have designed a transferability analysis experiment, which has not been adequately addressed in the current literature. In this experiment, we test whether trained models from several cities contain valuable information to classify a different city. It was found that samples of the urban class have better reusability than those of other natural land cover classes, i.e., trees, low plants, bare soil or sand, and water. After experimental evaluation of different land cover classes across different cities, we conclude that continental land cover mapping results can be considerably improved when training samples of natural land cover classes are collected and combined from areas covering each Köppen climate zone.

Automated Production of a Land Cover/Use Map of Europe Based on Sentinel-2 Imagery

“…Therefore, their appropriate preparation is a crucial step in the workflow. However, the collection of the qualitatively homogenous training dataset on a continental or global scale is a very demanding task [34,45]. Field surveys are infeasible due to the high cost, the required time and workload.…”

“…It is common practice to use, besides the spectral bands of satellite imagery, other explanatory variables (classification features), such as spectral indices derived from the analysed images [34,[58][59][60][61]. Therefore, to enhance recognition of land cover classes, we fed the classifier with similar features.…”

“…There is no need to organise field campaigns, or prepare training samples through the visual interpretation of EO. Indeed, such processes are prerequisites in numerous projects [33,34]. The joint application of the complementary CLC and HRL datasets, along with spectral indices derived from satellite imagery, help to improve the quality of the training set by filtering out erroneous samples.…”

“…The collection of a sample suitable for large-scale mapping is often seen as a bottleneck, and may require a tremendous amount of manual work [31,32]. For example, Gong et al [20,33] and Li et al [34] produced large-scale, high resolution land cover databases, however they required the enormous and usually cost-intensive manual visual image interpretation that seriously hampers repetitive and regular (e.g., annual) map production or their update [1].…”

Remote Sensing Applications in Monitoring of Protected Areas

“…The improved capacity of data science and infrastructure, e.g., cloud computing, Google Earth Engine (GEE) and big Earth data approaches, facilitates data sharing and the integration and modeling processes [87][88][89]. For example, the capacity and service from GEE open opportunities for explorations that benefit from decades of data acquisition from remote sensing [90][91][92][93][94][95][96].…”