摘要：
Accurate geographic data of slums are important for handling urban poverty issues. Previous slum mapping studies using high-resolution or very-high-resolution (HR/VHR) remotely sensed (RS) images are commonly not suitable for city-wide scale tasks. This study aims to efficiently generate a slum map on a city-wide scale using freely accessed multispectral medium-resolution (MR) Sentinel-2 images. Composite slum spectral indices (CSSIs) were initially proposed based on the shapes of spectral profiles of slums and nonslums and directly represent slum characteristics. Specifically, CSSI-1 denotes the normalized difference between the shortwave infrared bands and the red edge band, while CSSI-2 denotes the normalized difference between the blue band and the green band. Furthermore, two methods were developed to test the effectiveness of CSSIs on slum mapping, i.e., the threshold-based method and the machine learning (ML)-based method. Experimental results show that the threshold-based method and the ML-based method achieve intersection over unions (IoU) of 43.89% and 54.45% in Mumbai, respectively. The accuracies of our methods are comparable to or even higher than the accuracies reported by existing methods using HR/VHR images and transfer learning. The threshold-based method exhibits a promising performance in mapping slums larger than 5 ha, while the ML-based method refines mapping accuracies for slum pockets smaller than 5 ha. The threshold-based method and the ML-based method produced the slum map in Mumbai in 2 and 28 min, respectively. Our methods are suitable for rapid large-area slum mapping owing to the high data availability of Sentinel-2 images and high computational efficiency.
期刊：
ISPRS Journal of Photogrammetry and Remote Sensing,2023年205:34-49 ISSN：0924-2716
通讯作者：
Hu, Q
作者机构：
[Xu, Baodong; Shi, Zhihua; Cai, Zhiwen] Huazhong Agr Univ, Coll Resources & Environm, Wuhan 430070, Peoples R China.;[Hu, Q; Hu, Qiong; Yang, Jingya] Cent China Normal Univ, Coll Urban & Environm Sci, Hubei Prov Key Lab Geog Proc Anal & Simulat, Wuhan 430079, Hubei, Peoples R China.;[Zhang, Xinyu; You, Liangzhi; Wei, Haodong] Huazhong Agr Univ, Macro Agr Res Inst, Coll Plant Sci & Technol, Wuhan 430070, Peoples R China.;[Li, Wenjuan; Yang, Jingya] Chinese Acad Agr Sci, Inst Agr Resources & Reg Planning, State Key Lab Efficient Utilizat Arid & Semi arid, Beijing 100081, Peoples R China.;[Zeng, Yelu] China Agr Univ, Coll Land Sci & Technol, Beijing 100083, Peoples R China.
通讯机构：
[Hu, Q ] C;Cent China Normal Univ, Coll Urban & Environm Sci, Hubei Prov Key Lab Geog Proc Anal & Simulat, Wuhan 430079, Hubei, Peoples R China.
关键词：
Agricultural field parcel delineation Deep learning Multimodal satellite data Spatiotemporal fusion Spatial transferability
摘要：
Accurate spatial information for agricultural field parcels is important for agricultural production management and understanding agro-industrialization and intensification. However, traditional remote sensing methods that rely on single-modal or single-date data struggle to identify heterogeneous field parcels, particularly in regions dominated by smallholder farming systems. To address this challenge, we proposed a Dual branch Spatiotemporal Fusion Network (DSTFNet) that integrated very high-resolution (VHR) images and medium-resolution satellite image time series (MRSITS) to extract agricultural field parcels over various landscapes. The DSTFNet consisted of two branches: a spatial branch that extracted spatial features from VHR images and a temporal branch that explored seasonal spectral dynamics from MRSITS data by using ConvLSTM units and an attention module. We evaluated the DSTFNet in four regions across China by using GF-2 and Sentinel-2 data. The results showed that DSTFNet performed well in delineating agricultural field parcels, achieving the highest Matthew's correlation coefficient (MCC) = 0.823 for the field extent, the highest F1-score of edge (Fedge) = 0.865 for field boundary, and the lowest errors of segmentation evaluation index (SEI) = 0.191 for the vectorized field parcels in Hubei province. In addition, DSTFNet significantly outperformed two single-branch models that used VHR or MRSITS alone, as well as existing BsiNet, ResUNet_a, UNet and RAUNet models. DSTFNet also showed good spatial transferability in distinct regions without training data (on average, MCC = 0.728, Fedge = 0.729, and SEI = 0.281 for three target regions). Using limited training data to fine-tune the DSTFNet can further improve its ability to delineate field parcels over complex regions. The visualization analysis of temporal attention weights demonstrated that DSTFNet can well capture cropland spectral dynamics, making it advantageous in extracting diverse cropland parcels. By exploiting important spectral, spatial and temporal information from multimodal satellite data, DSTFNet provided an effective, robust, and transferable solution for accurately delineating agricultural field parcels across heterogeneous farming systems.
期刊：
Pest Management Science,2023年79(7):2591-2602 ISSN：1526-498X
通讯作者：
Ran Meng<&wdkj&>Ran Meng Ran Meng Ran Meng
作者机构：
[Meng, Ran; Lv, Zhengang; Zhou, Longfei; Xu, Binyuan; Sun, Rui] Huazhong Agr Univ, Coll Resources & Environm, Wuhan, Peoples R China.;[Meng, Ran] HIT Inst Artificial Intelligence Co Ltd, Harbin, Peoples R China.;[Yang, Wanneng; Chen, Gengshen] Huazhong Agr Univ, Natl Ctr Plant Gene Res Wuhan, Natl Key Lab Crop Genet Improvement, Hubei Hongshan Lab, Wuhan, Peoples R China.;[Liang, Linlin] Chinese Acad Sci, Aerosp Informat Res Inst, Beijing, Peoples R China.;[Zhao, Feng] Cent China Normal Univ, Coll Urban & Environm Sci, Key Lab Geog Proc Anal & Simulat Hubei Prov, Wuhan, Peoples R China.
通讯机构：
[Ran Meng; Ran Meng Ran Meng Ran Meng] C;College of Resources and Environment, Huazhong Agricultural University, Wuhan, China<&wdkj&>HIT Institute for Artificial Intelligence Co. Ltd, Harbin, China
作者机构：
[Tao, Jianbin; Zhou, Yang; Zhang, Xinyue; Jiang, Qiyue; Zhou, Y] Cent China Normal Univ, Key Lab Geog Proc Anal & Simulat Hubei Prov, Sch Urban & Environm Sci, Wuhan 430079, Peoples R China.;[Liu, Yiqing] Beijing Normal Univ, Inst Disaster Risk Sci, Fac Geog Sci, Beijing 100875, Peoples R China.
通讯机构：
[Zhou, Y ] C;Cent China Normal Univ, Key Lab Geog Proc Anal & Simulat Hubei Prov, Sch Urban & Environm Sci, Wuhan 430079, Peoples R China.
关键词：
cropping intensity;temporal mixture analysis;endmember;unmixing;time series images
摘要：
Agricultural cropping intensity plays an important role in evaluating the food security and the sustainable development of agriculture. The existing indicators measuring cropping intensity include cropping frequency and multiple cropping index. As a nominal measurement, cropping frequency classifies crop patterns into single-cropping and/or double-cropping and leads to information loss. Multiple cropping index is calculated on the basis of statistical data, ignoring the spatial heterogeneity within the administrative region. Neither of these indicators can meet the requirements of precision agriculture, and new methods for fine cropping intensity mapping are still lacking. Time series remote sensing data provide vegetation phenology information and reveal temporal development of vegetation, which can be used to facilitate the fine cropping intensity mapping. In this study, a new temporal mixture analysis method is introduced to estimate the abundance level cropping intensity from time series remote sensing data. By analyzing phenological characteristics of major land-cover types in time series vegetatiosacan indices, a novel feature space was constructed by using the selected PCA components, and three unique endmembers (double-cropping, natural vegetations and water bodies) were found. Then, a linear spectral mixture analysis model was applied to decompose mixed pixels by replacing spectral data with multi-temporal data. The spatio-temporal continuous, fine resolution, abundance level cropping intensity maps were produced for the North China Plain and the middle and lower reaches of the Yangtze River Valley. The experiments indicate a good result at both county and pixel level validation. The method of manually delineating endmembers can well balance the accuracy and efficiency. We also found the size of the study area has little effect on the unmixing accuracy. The results demonstrated that the proposed method can model cropping intensity finely at large scale and long temporal span, at the same time with high efficiency and ease of implementation.
作者机构：
[Chen, Wei; Li, Siliang; Jiang, Bohan] Tianjin Univ, Inst Surface Earth Syst Sci, Sch Earth Syst Sci, Tianjin 300072, Peoples R China.;[Chen, Wei; Li, Siliang; Jiang, Bohan] Haihe Lab Sustainable Chem Transformat, Tianjin 300192, Peoples R China.;[Chen, Wei; Li, Siliang; Jiang, Bohan] Tianjin Bohai Rim Coastal Earth Crit Zone Natl Obs, Tianjin 300072, Peoples R China.;[Dai, Xiaoai] Chengdu Univ Technol, Coll Earth Sci, Chengdu 610059, Peoples R China.;[Xu, Min] Chinese Acad Sci, Aerosp Informat Res Inst, State Key Lab Remote Sensing Sci, Beijing 100101, Peoples R China.
通讯机构：
[Wei Chen] I;[Xiaoai Dai] C;College of Earth Sciences, Chengdu University of Technology, Chengdu 610059, China<&wdkj&>Institute of Surface-Earth System Science, School of Earth System Science, Tianjin University, Tianjin 300072, China<&wdkj&>Haihe Laboratory of Sustainable Chemical Transformation, Tianjin 300192, China<&wdkj&>Tianjin Bohai Rim Coastal Earth Critical Zone National Observation and Research Station, Tianjin 300072, China
作者：
Wu, Tieniu;Cheng, Antai;Lin, Henry;Zhang, Hailin;Jie, Yi
期刊：
地球科学学刊,2023年34(5):1556-1566 ISSN：1674-487X
通讯作者：
Wu, TN
作者机构：
[Cheng, Antai; Wu, Tieniu; Jie, Yi; Zhang, Hailin] Cent China Normal Univ, Key Lab Geog Proc Anal & Simulat, Wuhan 430079, Hubei Province, Peoples R China.;[Lin, Henry; Wu, Tieniu] Penn State Univ, Dept Ecosyst Sci & Management, University Pk, PA 16802 USA.;[Cheng, Antai; Wu, Tieniu; Jie, Yi; Zhang, Hailin] Cent China Normal Univ, Coll Urban & Environm Sci, Wuhan 430079, Peoples R China.
通讯机构：
[Wu, TN ] C;Cent China Normal Univ, Key Lab Geog Proc Anal & Simulat, Wuhan 430079, Hubei Province, Peoples R China.;Penn State Univ, Dept Ecosyst Sci & Management, University Pk, PA 16802 USA.;Cent China Normal Univ, Coll Urban & Environm Sci, Wuhan 430079, Peoples R China.
关键词：
MIS 9;climatic fluctuation;Paleosol;S3;Chinese Loess Plateau;environmental geology
摘要：
Marine Isotope Stages (MIS) 9 has been proposed as an analog for the present warm period. However, detailed studies of this geological time period are rare in loess-paleosol sequence. In the Chinese Loess Plateau (CLP), the corresponding stratum is the third paleosol layer (S3). Here, we report the terrestrial reconstruction of climatic fluctuations during MIS 9 by analyzing the paleo-climate indexes of S3 with high sampling density. Our results showed that: (1) During the period of MIS 9, the main climatic sub-cycle was 29 ka; (2) MIS 9 could be divided into five sections, MIS 9a, 9b, 9c, 9d, and 9e. Among them, MIS 9a, 9c, and 9e were warm stages, while MIS 9b and 9d were cool intervals; and 3) There were also three swift warm-wet events and one cool-dry event, which occurred around 332-331, 324-323, 311-310, and 331-329 ka BP, respectively. The overall trend of paleo-climate fluctuation correlated approximately with SPECMAP, LR04 stack and Iberian margin deep-sea cores. This study suggested that the paleosol records in the southern margin of the CLP have global significance and contain more detailed climatic signals than marine deposits.
