Urban land cover mapping under the Local Climate Zone scheme using Sentinel-2 and PALSAR-2 data
Graphical abstract
Introduction
With the increasing concentration of population in urban areas, the conversion from natural to urban land cover induces changes in the physical characteristics of the urban surface (e.g., albedo, thermal capacity, heat conductivity, and moisture content) (Voog and Oke, 2003). The urban heat island (UHI) effect, which is characterized by higher air and surface temperature in urban areas than in surrounding rural areas, is regarded as the most significant consequences of these changes (Voog and Oke, 2003). As it affects health and well-being of urban residents as well as urban environments, the UHI is a recent “hot topic” for research worldwide (Mohajerani et al., 2017).
Stewart and Oke (2012) were the first to develop a standardized worldwide land cover classification system to describe physical properties of urban morphology and corresponding thermal performance in UHIs, namely the climate-oriented Local Climate Zone (LCZ) scheme. As a standardized framework to characterize urban forms and functions, the LCZ scheme is applied globally as part of the World Urban Database and Portal Tools (WUDAPT) initiative (Danylo et al., 2016). The LCZ scheme has gone beyond its original purpose of studying the UHI phenomena and is now recognized as a valuable framework for characterizing urban land cover with satellite remote sensing images (Ching, 2013; Qiu et al., 2019).
Because of the spatial and spectral heterogeneity of urban areas, separating different land cover types with spectral characteristics using only optical images is a challenging process (Cockx et al., 2014). Recent studies demonstrated that synthetic aperture radar (SAR) data can provide complementary information to optical data (Darmawan et al., 2015), and their integration often leads to increased classification accuracy (Bechtel et al., 2016; Xu et al., 2017a; Xu et al., 2017b; Pavanelli et al., 2018).
Fully polarimetric SAR collects complete backscattering information in the coherency matrix T3. Based on T3, different target decomposition methods were developed to transform the backscattering information into basic backscattering mechanisms for geophysical parameter inversion as well as terrain and target classification (Cloude and Pottier, 1997; Touzi et al., 2004; Yamaguchi et al., 2005). Therefore, fully polarimetric SAR offers a powerful way to detect and characterize targets, based on electromagnetic radiation characteristics and backscattering mechanisms (Lee and Pottier, 2009; Muneyoshi and Junichi, 2013; Angelliaume et al., 2018; Ohki and Shimada, 2018).
Recent studies, taking advantage of the spectral, spatial, and polarimetric information, have combined optical and fully polarimetric SAR data to improve land cover classification accuracy (Bagan et al., 2012; Zhu et al., 2012; Merchant et al., 2019; Muthukumarasamy et al., 2019). However, the integration of backscattering intensity, decomposed components and polarimetric parameters from fully polarimetric SAR and multi-spectral imagery for land cover classification needs further investigation.
A dataset combined by multi-source data is likely to be imbalanced and with data redundancy when the dimensionality is relatively high. Using such a dataset for land cover classification would be a challenge. The subspace method has shown potential for classifying remote sensing data with high-dimensionality, due to the ability of reducing data dimensionality and extracting the features simultaneously (Bagan et al., 2012). To demonstrate the proposed subspace method, we applied it to datasets integrated by Sentinel-2 and fully polarimetric PALSAR-2 data in Shanghai, China, and investigated the robustness of the subspace method in comparison with the Support Vector Machines (SVM) (Mountrakis et al., 2011; Koda et al., 2018) and Maximum Likelihood Classifier (MLC) (Lillesand et al., 2015).
The remainder of this paper is structured as follows. In Section 2, we described the study area, Seninel-2 and PALSAR-2 imagery, ancillary data, LCZ scheme, classification method, classification process and accuracy assessment. Section 3 shows the experimental results. In Section 4, we analyzed the experimental results and discussed the remaining challenges and possible solutions for future work. Finally, Section 5 summarizes and concludes our study.
Section snippets
Study area
We chose Shanghai, a commercial and financial center of mainland China, as our study area (Fig. 1). Shanghai (centered at 34°05′N, 121°43′E) lies on the alluvial plain of the Yangtze River Delta, along the east coast of China. The land area is a nearly flat plain, with an elevation of less than 4 m above sea level. Shanghai has a northern subtropical marine monsoon climate and experiences four distinct seasons (Cui and Shi, 2012), with an annual precipitation of 1388.8 mm, an annual mean
Results
For comparison, we calculated confusion matrices of the classified LCZ maps generated from six datasets by three classification methods. The OAs, PAs, and UAs derived from confusion matrices are presented in the Table S2. Then, we arranged the OAs of three classifiers using datasets of C1-C6 in ascending order as C1, C2, C5, C3, C4, C6 by the subspace method; C1, C2, C5, C3 and C6, C4 by SVM; and C6, C5, C1, C4, C2, C3 by MLC. Using C1, OAs by the subspace method, SVM, and MLC were 65.9%,
Discussion
We generated LCZ maps using different combinations of multi-spectral bands from multi-spectral Sentinel-2A MSI and features of fully polarimetric PALSAR-2 by the subspace method, SVM and MLC. The performance of three classifiers using different datasets was assessed by calculating the confusion matrices of classification results. The OA of classification maps was improved by adding polarimetric features into multi-spectral data. The subspace method performed better than SVM and MLC using
Conclusion
Combining multi-spectral Sentinel-2 MSI imagery and fully polarimetric PALSAR-2 data benefits land cover classification based on the LCZ scheme in a dense urban area. Using dataset accounting for both physical and spectral information, combined by four scattering intensity images and multi-spectral bands, can substantially improve classification accuracy compared with the use of multi-spectral bands alone. In addition, scattering parameters of the coherency matrix T3 (i.e., amplitude and phase)
Declaration of Competing Interest
None.
Acknowledgement
This work was supported by the National Natural Science Foundation of China (Grant No. 41771372 and 41730642), and the Science and Technology Commission of Shanghai Municipality (Grant no.18511102300).
References (53)
- et al.
Sensitivity of the subspace method for land cover classification
Egypt. J. Remote Sens. Space Sci.
(2018) A perspective on urban canopy layer modeling for weather, climate and air quality applications
Urban Clim.
(2013)- et al.
Quantifying uncertainty in remote sensing-based urban land-use mapping
Int. J. Appl. Earth Obs.
(2014) - et al.
Urbanization and its environmental effects in Shanghai, China
Urban Clim.
(2012) Explaining the unsuitability of the kappa coefficient in the assessment and comparison of the accuracy of thematic maps obtained by image classification
Remote Sens. Environ.
(2020)- et al.
Change detection from remotely sensed images: from pixel-based to object-based approaches
ISPRS J. Photogramm. Remote Sens.
(2013) - et al.
Spatial and temporal patterns of China’s cropland during 1990–2000: an analysis based on Landsat TM data
Remote Sens. Environ.
(2005) - et al.
The urban heat island effect, its causes, and mitigation, with reference to the thermal properties of asphalt concrete
J. Environ. Manag.
(2017) - et al.
Support vector machines in remote sensing: a review
ISPRS J. Photogramm. Remote Sens.
(2011) - et al.
Local climate zone-based urban land cover classification from multi-seasonal Sentinel-2 images with a recurrent residual network
ISPRS J. Photogramm. Remote Sens.
(2019)
Fusion of PolSAR and PolInSAR data for land cover classification
Int. J. Appl. Earth Obs.
Assessment of spectral polarimetric temporal and spatial dimensions for urban and peri-urban land cover classification using Landsat and SAR data
Remote Sens. Environ.
SAR imagery for detecting sea surface slicks: performance assessment of polarization-dependent parameters
IEEE Trans. Geosci. Remote Sens.
Improved subspace classification method for multispectral remote sensing image classification
Photogramm. Eng. Remote. Sens.
Combination of AVNIR-2 PALSAR and polarimetric parameters for land cover classification
IEEE Trans. Geosci. Remote Sens.
Classification of local climate zones using SAR and multispectral data in an arid environment
IEEE J. Sel. Topics Appl. Earth Observ. Remote Sens.
Shanghai Statistical Yearbook
Fusion of sentinel-1A and sentinel-2A data for land cover mapping: a case study in the lower Magdalena region, Colombia
J. Maps.
An entropy based classification scheme for land applications of polarimetric SAR
IEEE Trans. Geosci. Remote Sens.
Assessing the Accuracy of Remotely Sensed Data: Principles and Practices
Contributing to WUDAPT: a local climate zone classification of two cities in Ukraine
IEEE J. Sel. Topics Appl. Earth Observ. Remote Sens.
Impact of topography and tidal height on ALOS PALSAR Polarimetric measurements to estimate aboveground biomass of mangrove Forest in Indonesia
J. Sens.
Study on classification of land cover targets over parts of Burdwan, West Bengal using RISAT-1 hybrid polarimetric SAR images and optical remote sensing data
Difference subspace and its generalization for subspace-based methods
IEEE PAMI
A novel MKL model of integrating LiDAR data and MSI for urban area classification
IEEE Trans. Geosci. Remote Sens.
Supervised spatial classification of multispectral LiDAR data in urban areas
PLoS One
Cited by (22)
Impacts of local climate zone mapping quality on urban near-surface air temperature simulation in WRF-UCM
2024, Sustainable Cities and SocietyMapping local climate zones for cities: A large review
2023, Remote Sensing of EnvironmentInvestigating urban heat-related health risks based on local climate zones: A case study of Changzhou in China
2023, Sustainable Cities and SocietyMapping urban morphology changes in the last two decades based on local climate zone scheme: A case study of three major urban agglomerations in China
2023, Urban ClimateCitation Excerpt :In particular, remote sensing images combined with machine learning have been widely employed for LCZ mapping due to their accessibility, broad coverage, and tolerable accuracy (Bechtel et al., 2015; See et al., 2015; Zheng et al., 2018). Traditional machine learning algorithms include random forest (RF), support vector machines (SVM), maximum likelihood classifier (MLC), etc. (Chung et al., 2021; La et al., 2020; Xu et al., 2017). To facilitate the development of the LCZ community, Bechtel et al. (2015) created the World Urban Database and Access Portal Tools (WUDAPT), which provides a global open platform for consistent LCZ data and has mapped the LCZ of 120 cities worldwide (Chung et al., 2021).
Improving air quality in Guangzhou with urban green infrastructure planning: An i-Tree Eco model study
2022, Journal of Cleaner ProductionSocioeconomic inequalities and the surface heat island distribution in Santiago, Chile
2022, Science of the Total Environment