Linking greenhouse gas emissions to urban landscape structure: the relevance of spatial and thematic resolutions of land use/cover data

Abstract

Context Emissions of greenhouse gases in urban areas play an important role in climate change. Increasing attention has been given to urban landscape structure-emission relationships (SERs). However, it remains unknown if and to what extent SERs are dependent on observational scale. Objective To assess how changing observational scales (in terms of spatial and thematic resolutions) of urban landscape structure affect SERs. Methods We examined correlations between 16 landscape metrics and greenhouse gas emissions across 52 European cities, through (1) systematic manipulation of spatial and thematic resolutions of the urban land use/cover (ULUC) dataset, and (2) comparison between available standard ULUC datasets with different spatial resolutions. Results Our analyses showed that the observed SERs significantly depend on both thematic and spatial resolutions of the ULUC data. For the 16 landscape metrics, we found diverse spatial/thematic scaling relations exhibiting monotonic, hump-shaped or scaleinvariant trends. For different landscape metrics, the SERs were strongest at different spatial scales,

The resolution‐dependent role of landscape attributes in shaping macro‐scale biodiversity patterns

“…The resulting land cover datasets at different spatial and thematic resolutions were used to calculate landscape metrics. These scaling approaches have been adopted in previous studies that systematically investigated scaling behaviour based on scalograms (e.g., Luan et al, ; Wu, Shen, Sun, & Tueller, ). For each of the 21 resulting land cover datasets, we calculated a comprehensive set of 28 class‐ and landscape‐level metrics describing landscape composition and structure (Supporting Information Table S2; for details on class‐ and landscape‐level metrics, see McGarigal & Marks, ).…”

Remote sensing in landscape ecology

“…As expected, the core focus of recent articles in the journal that used remote sensing remains on spatial patterns and issues of heterogeneity with scaling issues often of considerable interest (Frazier 2014;Luan et al 2018;Egerer et al 2020;Rudge et al 2022;Mondal and Jeganathan 2022;Gann and Richards 2023). Many articles have made use of Landsat sensor data (Moris et al 2022;Hopkins et al 2022) notably taking advantage of the relatively long time series of data that has now been formed (Zhao et al 2015;Bost et al 2019;Jung et al 2020;Fisher et al 2021;Yu et al 2021).…”

Assessing the Effect of Training Sampling Design on the Performance of Machine Learning Classifiers for Land Cover Mapping Using Multi-Temporal Remote Sensing Data and Google Earth Engine

“…Land Use and Land Cover (LULC) maps are not only vital for landscape monitoring, planning, and management but also for studying the impact of climate change and human interventions on the ecosystem processes and services [1][2][3]. The term "land cover" refers to the physical cover present on the surface of the earth, whereas "land use" refers to the purpose for which the land is used.…”

“…The research is carried out in a complex and rugged Himalayan landscape where the availability of reference data is often limited. The study addresses two major research questions: (1) what is the effect of various training sampling methods on the classification results obtained by the RF classifier? and (2) how well do the evaluated ML classifiers perform with respect to each other for land cover mapping in a complex environment?…”