本期【期刊跟蹤】選取了與遙感、土地利用變化有關的11篇文章，分別來自Remote Sensing of Environment、International Journal of Applied Earth Observation and Geoinformation、Remote Sensing、Science of The Total Environment、Ecological Indicators、遙感學報、中國科學：地球科學等7個期刊。
1. An efficient approach to capture continuous impervious surface dynamics using spatial-temporal rules and dense Landsat time series stacks
作者：Liu Chong, Zhang Qi, Luo Hui等
期刊：Remote Sensing of Environment
摘要：Impervious surface dynamics have far-reaching consequences on both the environment and human well-being. The expansion of impervious surface is often spontaneous and conscious, particularly in fast developing regions. Thus, monitoring impervious surface dynamics with high temporal frequency in a both accurate and efcient manner is highly needed. Here, we propose an approach to capture continuous impervious surface dynamics using spatial-temporal rules and dense time series stacks of Landsat data. First, a stable area mask based on image classifcation in the start and the end years is generated to remove pixels that are persistent or spatially irrelevant. The Continuous Change Detection (CCD) algorithm is then employed to determine the change points when non-impervious cover converts to impervious surface based on the property of temporal irreversibility. Finally, the CCD time series models are calibrated for pixels with no change or multiple changes. We apply and assess the proposed approach in Nanchang (China), which has been experiencing rapid impervious surface expansion during the past decade. According to the validation results, overall accuracies of image classifcation in the start and the end years are 97.2% and 96.7%, respectively. Our approach generates convincing results for impervious surface change detection, with overall accuracy of 85.5% at the annual scale, which is higher than three commonly used approaches in previous studies. At the continuous scale, the mean biases of the detected time of imperviousness emergence are +0.17 (backward) and −3.42 (forward) Landsat revisit periods (16 days) for pixels with one single change and multiple changes, respectively. The derived impervious surface extent maps exhibit comparable performances with fve widely used products. The present approach offers a new perspective for providing timely and accurate impervious surface dynamics with dense temporal frequency and high classifcation accuracy.
引用格式：Liu, C., Zhang, Q., Luo, H., Qi, S., Tao, S., Xu, H., & Yao, Y. (2019). An efficient approach to capture continuous impervious surface dynamics using spatial-temporal rules and dense Landsat time series stacks. Remote Sensing of Environment, 229, 114–132.
2.Coastline extraction from repeat high resolution satellite imagery
作者：Chunli Dai, Ian M. Howat, Eric Larour, Erik Husby
期刊：Remote Sensing of Environment
摘要：This paper presents a new coastline extraction method that improves water classifcation accuracy by beneftting from an ever-increasing volume of repeated measurements from commercial satellite missions. The widely-used Normalized Difference Water Index (NDWI) method is tested on a sample of around 12,600 satellite images for statistical analysis. The core of the new water classifcation method is the use of a water probability algorithm based on the stacking of repeat measurements, which can mitigate the effects of translational offsets of images and the classifcation errors caused by clouds and cloud shadows. By integrating QuickBird, WorldView-2 and WorldView-3 multispectral images, the fnal data product provides a 2 m resolution coastline, as well as a 2 m water probability map and a repeat-count measurement map. Improvements on the existing coastline (GSHHSthe Global Self-consistent, Hierarchical, High-resolution Shoreline Database, 50 m–5000 m) in terms of resolution (2 m) is substantial, thanks to the combination of multiple data sources.
引用格式：Dai, C., Howat, I. M., Larour, E., & Husby, E. (2019). Coastline extraction from repeat high resolution satellite imagery. Remote Sensing of Environment, 229, 260–270.
3. A cloud detection algorithm for satellite imagery based on deep learning
作者：Jacob Høxbroe Jeppesen, Rune Hylsberg Jacobsen, Fadil Inceoglu
期刊：Remote Sensing of Environment
摘要：Reliable detection of clouds is a critical pre-processing step in optical satellite based remote sensing. Currently, most methods are based on classifying invidual pixels from their spectral signatures, therefore they do not incorporate the spatial patterns. This often leads to misclassifcations of highly reflective surfaces, such as human made structures or snow/ice. Multi-temporal methods can be used to alleviate this problem, but these methods introduce new problems, such as the need of a cloud-free image of the scene. In this paper, we introduce the Remote Sensing Network (RS-Net), a deep learning model for detection of clouds in optical satellite imagery, based on the U-net architecture. The model is trained and evaluated using the Landsat 8 Biome and SPARCS datasets, and it shows state-of-the-art performance, especially over biomes with hardly distinguishable scenery, such as clouds over snowy and icy regions. In particular, the performance of the model that uses only the RGB bands is signifcantly improved, showing promising results for cloud detection with smaller satellites with limited multi-spectral capabilities. Furthermore, we show how training the RS-Net models on data from an existing cloud masking method, which are treated as noisy data, leads to increased performance compared to the original method. This is validated by using the Fmask algorithm to annotate the Landsat 8 datasets, and then use these annotations as training data for regularized RS-Net models, which then show improved performance compared to the Fmask algorithm. Finally, the classifcation time of a full Landsat 8 product is 18.0 ± 2.4 s for the largest RS-Net model, thereby making it suitable for production environments.
引用格式：Jeppesen, J. H., Jacobsen, R. H., Inceoglu, F., & Toftegaard, T. S. (2019). A cloud detection algorithm for satellite imagery based on deep learning. Remote Sensing of Environment, 229, 247–259.
4. Multi-sensor cloud and cloud shadow segmentation with a convolutional neural network
作者：Marc Wieland, Yu Li, Sandro Martinis
期刊：Remote Sensing of Environment
摘要：Cloud and cloud shadow segmentation is a crucial pre-processing step for any application that uses multispectral satellite images. In particular, disaster related applications (e.g., flood monitoring or rapid damage mapping), which are highly time- and data-critical, require methods that produce accurate cloud and cloud shadow masks in short time while being able to adapt to large variations in the target domain (induced by atmospheric conditions, different sensors, scene properties, etc.). In this study, we propose a data-driven approach to semantic segmentation of cloud and cloud shadow in single date images based on a modified U-Net convolutional neural network that aims to fulfil these requirements. We train the network on a global database of Landsat OLI images for the segmentation of five classes (「shadow」, 「cloud」, 「water」, 「land」 and 「snow/ice」). We compare the results to state-of-the-art methods, proof the model's generalization ability across multiple satellite sensors (Landsat TM, Landsat ETM+, Landsat OLI and Sentinel-2) and show the influence of different training strategies and spectral band combinations on the performance of the segmentation. Our method consistently outperforms Fmask and a traditional Random Forest classifier on a globally distributed multi-sensor test dataset in terms of accuracy, Cohen's Kappa coefficient, Dice coefficient and inference speed. The results indicate that a reduced feature space composed solely of red, green, blue and near-infrared bands already produces good results for all tested sensors. If available, adding shortwave-infrared bands can increase the accuracy. Contrast and brightness augmentations of the training data further improve the segmentation performance. The best performing U-Net model achieves an accuracy of 0.89, Kappa of 0.82 and Dice coefficient of 0.85, while running the inference over 896 test image tiles with 44.8 s/megapixel (2.8 s/megapixel on GPU). The Random Forest classifier reaches an accuracy of 0.79, Kappa of 0.65 and Dice coefficient of 0.74 with 3.9 s/megapixel inference time (on CPU) on the same training and testing data. The rule-based Fmask method takes significantly longer (277.8 s/megapixel) and produces results with an accuracy of 0.75, Kappa of 0.60 and Dice coefficient of 0.72.
引用格式：Wieland, M., Li, Y., & Martinis, S. (2019). Multi-sensor cloud and cloud shadow segmentation with a convolutional neural network. Remote Sensing of Environment, 230, 111203.
5. Optimal dates for assessing long-term changes in tree-cover in the semi-arid biomes of South Africa using MODIS NDVI time series (2001–2018)
作者：Moses Azong Choa, Abel Ramoelo
期刊：International Journal of Applied Earth Observation and Geoinformation
摘要：The varying proportions of tree and herbaceous cover in the grassland and savanna biomes of Southern Africa determine their capacity to provide ecosystem services.First, a 16-day NDVI time series was generated from MODIS NDVI data, i.e. MOD13A2 16-day NDVI composite data. Secondly, percentage tree-cover data for 100 sample polygons (4 × 4) pixels for areas that have not undergone change in tree cover between 2001 and 2018 were derived using high resolution Google Earth imagery. Next, a time series consisting of the coefficients of determination (R2) for the NDVI/tree-cover linear regression were computed for the 100 polygons. Lastly, a threshold R2 > 0.5 was used to determine the optimal period of the year for mapping tree-cover.
It emerged that the narrow period from Julian day 161–177 (June 10–26) was the most consistent period with R2 > 0.5 in the region. 18 tree-cover maps (2001–2018) were generated using linear regression model coefficients derived from Julian day 161 for each year. Kendall correlation coefficient (tau) was used to determine areas of significant (p < 0.05 and p < 0.01) increasing or decreasing trend in tree-cover. Areas (polygons) that showed increasing tree-cover appeared to be more widespread in the trend map as compared to areas of decreasing tree-cover. An accuracy assessment of the map of increasing tree-cover was conducted using Google Earth high resolution images. Out of 330 and 200 mapped polygons verified using p < 0.05 and 0.01 thresholds, respectively, 180 (54% accuracy) and 132 (65% accuracy) showed evidence of tree recruitment.
引用格式：Moses Azong Choa, Abel Ramoelo. Optimal dates for assessing long-term changes in tree-cover in the semi-arid biomes of South Africa using MODIS NDVI time series (2001–2018) . Int. J. Appl. Earth Obs. Geoinf. 81,27–36
6. Wetland Classification with Multi-Angle/Temporal SAR Using Random Forests
作者：Sarah Banks, Lori White, Amir Behnamian, Zhaohua Chen, Benoit Montpetit , Brian Brisco, Jon Pasher and Jason Duff
期刊：Remote Sensing
摘要：To better understand and mitigate threats to the long-term health and functioning of wetlands, there is need to establish comprehensive inventorying and monitoring programs. Here, remote sensing data and machine learning techniques that could support or substitute traditional field-based data collection are evaluated. For the Bay of Quinte on Lake Ontario, Canada, different combinations of multi-angle/temporal quad pol RADARSAT-2, simulated compact pol RADARSAT Constellation Mission (RCM), and high and low spatial resolution Digital Elevation and Surface Models (DEM and DSM, respectively) were used to classify six land cover classes with Random Forests: shallow water, marsh, swamp, water, forest, and agriculture/non-forested. Results demonstrate that high accuracies can be achieved with multi-temporal SAR data alone (e.g., user’s and producer’s accuracies≥90% for a model based on a spring image and a summer image), or via fusion of SAR and DEM and DSM data for single dates/incidence angles (e.g., user’s and producer’s accuracies≥90% for a model based on a spring image, DEM, and DSM data). For all models based on single SAR images, simulated compact pol data generally achieved lower accuracies than quad pol RADARSAT-2 data. However, it was possible to compensate for observed differences through either multi-temporal/angle data fusion or the inclusion of DEM and DSM data (i.e., as a result, there was not a statistically significant difference between multiple models). With a higher repeat-pass cycle than RADARSAT-2, RCM is expected to be a reliable source of C-band SAR data that will contribute positively to ongoing efforts to inventory wetlands and monitor change in areas containing the same land cover classes evaluated here.
引用格式：Sarah Banks, Lori White, Amir Behnamian, Zhaohua Chen, Benoit Montpetit , Brian Brisco, Jon Pasher and Jason Duff. Wetland Classification with Multi-Angle/Temporal SAR Using Random Forests. Remote Sensing, 2019, 11(6): 670 – 697.
7. Prospects for Imaging Terrestrial Water Storage in South America Using Daily GPS Observations
作者：Vagner G. Ferreira, Christopher E. Ndehedehe, Henry C. Montecino, Bin Yong, Peng Yuan, Ahmed Abdalla and Abubakar S. Mohammed
期刊：Remote Sensing
摘要：Few studies have used crustal displacements sensed by the Global Positioning System (GPS) to assess the terrestrial water storage (TWS), which causes loadings. Furthermore, no study has investigated the feasibility of using GPS to image TWS over South America (SA), which contains the world’s driest (Atacama Desert) and wettest (Amazon Basin) regions. This work presents a resolution analysis of an inversion of GPS data over SA. Firstly, synthetic experiments were used to verify the spatial resolutions of GPS-imaged TWS and examine the resolving accuracies of the inversion based on checkerboard tests and closed-loop simulations using 「TWS」 from the Noah-driven Global Land Data Assimilation System (GLDAS-Noah). Secondly, observed radial displacements were used to image daily TWS. The inverted results of TWS at a resolution of 300 km present negligible errors, as shown by synthetic experiments involving 397 GPS stations across SA. However, as a result of missing daily observations, the actual daily number of available stations varied from 60–353, and only 6% of the daily GPS-imaged TWS agree with GLDAS-Noah TWS, which indicates a root-mean-squared error (RMSE) of less than 100 kg/m . Nevertheless, the inversion shows agreement that is better than 0.50 and 61.58 kg/m in terms of the correlation coefficient (Pearson) and RMSE, respectively, albeit at each GPS site.
Figure 6. Top panels show (a) synthetic checkerboard patterns with positive and negative mass loadings, (b) inverted checkerboard patterns, and (c) residuals of the differences between 「observed」 and inverted mass loadings at a spatial resolution of 1.0∘×1.0∘. Middle Panels (d,e,f) show the same results as Panels (a–c), respectively, but for a spatial resolution of 2.0∘×2.0∘. Bottom Panels (g–i) show the same results as Panels (a–c), respectively, but for a spatial resolution of 3.0∘×3.0∘. The dashed lines indicate a buffer zone of 5.0∘ w.r.t. the boundary of South America.
引用格式：Vagner G. Ferreira, Christopher E. Ndehedehe, Henry C. Montecino, Bin Yong, Peng Yuan, Ahmed Abdalla and Abubakar S. Mohammed. Prospects for Imaging Terrestrial Water Storage in South America Using Daily GPS Observations. Remote Sensing, 2019, 11(6): 679 – 723.
8.高分三號影像水體信息提取
作者：谷鑫志，曾慶偉，諶華，陳爾學，趙磊，於飛，塗寬
期刊：遙感學報
摘要：國內外針對陸地水體信息提取、洪澇災害快速響應方面具有較深入的研究，但是多採用發展較早、圖像質量可靠的可見光影像及國外星載SAR影像。中國合成孔徑雷達(SAR)衛星高分三號(GF-3)已獲取了大量多極化、全極化SAR數據，為了將GF-3影像快速應用到環境保護、水資源管理等行業中，本研究分析了水體與其他目標具有的不同後向散射特性，將閾值分割法與馬爾可夫隨機場(MRF)相結合，發展了一種檢測精度較高、自動化程度強的水體信息提取方法。該方法首先通過直方圖統計的方法對不同成像模式、不同極化的GF-3影像進行後向散射強度分析，在閾值分割的研究基礎上，比較了最大類間方差法(Otsu)和Kittler and Illingworth(KI)二值化法在水體-非水體分類中的效果。然後結合DEM和GF-3軌道參數排除因陰影現象產生的輻射失真對圖像概率分布的影響，得到初始的水體信息分布圖，再經過Fisher變換和馬爾可夫隨機場(MRF)的迭代運算，綜合利用GF-3影像的多極化信息和空間上下文信息，以最大後驗概率準則輸出最終的水體分布圖。利用了湖南省東北部不同成像模式的兩景GF-3影像進行試驗，在成像時間接近的光學影像中隨機選擇檢驗樣點進行精度評價。實驗結果表明，KI方法在GF-3水體提取應用中比Otsu方法具有更強的優勢，剔除圖像陰影區域後，自動化確定的閾值與目視解譯閾值更加接近，通過MRF模型優化以後，實現了對水體信息的連貫提取，對圖像噪聲具有較強的抑制作用。本研究對水體目標的提取精度均達到了85%以上，實驗結果精度優於基於光學影像的水體指數法，整個流程需要很少的人工經驗參與，具有自動化程度強、檢測精度高的優勢。
引用格式：谷鑫志, 曾慶偉, 諶華, 陳爾學, 趙磊, 於飛, 塗寬. 2019. 高分三號影像水體信息提取. 遙感學報, 23(3): 555–565Gu X Z, Zeng Q W, Shen H, Chen E X, Zhao L, Yu F and Tu K. 2019. Study on water information extraction using domestic GF-3 image. Journal of Remote Sensing, 23(3): 555–565
9.全球城市人居環境不透水面與綠地空間特徵製圖
作者：匡文慧
期刊：中國科學：地球科學
摘要：本研究基於城市等級尺度地表結構和景觀分類原理, 融合了GlobeLand30數據等, 建立了MODIS NDVI、 夜間燈光數據(DMSP/OLS)和全球各分區30個典型城市Landsat TM獲取的不透水面和綠地空間像元組分比例之間回歸關係, 發展了全球城市建成區內不透水面和綠地空間組分比例數據集. 基於Google Earth高解析度影像驗證表明, 城市不透水面製圖的平均相對誤差(MRE)為0.19. 本文評估的全球城市用地面積為76.29×104km2, 主要分布於北半球的歐洲中部地區、 亞洲東部地區和北美洲的中東部地區, 其中北美洲、 歐洲和亞洲城市用地面積為66.3×104km2, 佔全球城市用地總面積的86.91%, 全球前50位國家城市用地佔全球城市用地面積總量的59.32%. 全球城市不透水面面積為45.26×104km2, 集中分布於北美洲中南部、 亞洲東部、 歐洲大部分地區, 以及全球海岸沿線地區. 其中, 北美洲、 歐洲、 亞洲城市不透水面面積佔全球城市不透水面積的84.25%. 研究也發現, 全球不同發達程度國家之間城市不透水面分布和綠地布局呈現高度的空間差異特徵, 各大洲建成區內城市不透水面比例由大到小依次為: 非洲>70%>南美洲>大洋洲>亞洲>60%>北美洲>歐洲>50%. 全球各大洲綠地面積集中分布在北美洲東南部、 歐洲西南部、 亞洲東部和西部. 其中, 北美洲、 歐洲和亞洲綠地面積佔全球綠地總面積的89.44%. 歐洲和北美洲的發達國家更注重城市景觀設計及城市不透水面和綠地的有效鑲嵌, 城市建成區內綠地面積比例相對較高, 而部分發展中國家和欠發達國家建成區內綠地面積比例相對較低, 城市宜居環境有待提升.
引用格式：匡文慧. 全球城市人居環境不透水面與綠地空間特徵製圖.中國科學: 地球科學 49, 1151 (2019); doi: 10.1360/N072018-00164
1. Land use/land cover changes and its impact on ecosystem services in ecologically fragile zone: A case study of Zhangjiakou City, Hebei Province, China
作者：An Huang, Yueqing Xu, Piling Sun, Guiyao Zhou, Chao Liu, Longhui Lu, Ying Xiang, Hui Wang
期刊：Ecological Indicators
摘要：Land use/land cover (LULC) changes are likely to become more frequent and intense as a result of anthropogenic activities and may significantly affect human welfare by modifying ecosystem services (ESs). Understanding the impact of LULC changes on ESs value and the interactions among ESs could result in improvements in current land use policies and provide a scientific basis for the formulation of new policies in ecologically fragile zones. A case study was conducted in Zhangjiakou City, which is considered a typical ecologically fragile mountainous area in China, to examine the effects of LULC changes on ESs value and the interactions among ESs, including carbon sequestration and oxygen production (CSOP), water yield (WY), soil conservation (SC), sand-fixing (SF), and agricultural production (AP) from 2000 to 2015. Our results showed that ESs in Zhangjiakou City benefited substantially from existing land use policies and their 「win-win expectations.」 There were dramatic changes in the LULC types over the study period, especially in forestland, grassland, and arable land, with a significant impact on ESs value. LULC changes resulted in a significant increase in ESs value (US$ 3147.44 million), with the maximum increase occurring in AP (US$ 2255.19 million). However, LULC significantly decreased the value of the WY by $61.91 million, which mainly resulted from the degradation of arable land, forestland, and grassland. Strong trade-off relationships between WY and SC, CSOP and SC, SC and SF, and SC and AP were observed in 2000. Trade-off relationships were markedly weaken by LULCCs but increased by human activities when related to AP. Finally, a new spatialization approach of AP was designed and quantitative method of trade-off index was improved based on economic value. These results could offer some suggestions for land space optimization and ecological construction in Zhangjiakou City as well as in the similar regions in China, and provide some sci- entifically basis on the research area of coordination development of multi-functions of land use or geographical functions.
引用格式：Huang A, Xu Y, Sun P, et al. Land use/land cover changes and its impact on ecosystem services in ecologically fragile zone: A case study of Zhangjiakou City, Hebei Province, China[J]. Ecological Indicators, 2019, 104: 604-614
2. Multifactor-based environmental risk assessment for sustainable land-use planning in Shenzhen, China
作者：Qian Li, Yang Yu, Xiaoqian Jiang, Yuntao Guan
期刊：Science of The Total Environment
摘要：The health and resilience of urban ecosystems are highly dependent on interactions between the natural and built environments. Rapid urbanization, however, brings potential risks to urban ecosystems. Therefore, it is important to justify land-use planning and to identify opportunities for regulating sustainable land use. We developed a framework of land-use planning based on an Environmental Risk Assessment (ERA) and applied it to an environmentally sensitive urban area in Shenzhen, China. The ERA used the analytic hierarchy process method to determine weights for various indicators, and further performed a multifactor-based spatial superposition analysis in the ModelBuilder of a geographic information system to produce a risk map. The selected indicators were topography, hydrology, ecosystem, land use, and traffic. The risk map was first compared with an existing map of ecological control lines to ensure the reliability of the ERA, and then applied to establish the necessity and priority of land-use measures for four different land-use types: nature reserves, green space, urban areas, and spare land. The map indicated that nature reserves and green space make up 84.7% of the area at high risk of degradation, whereas urban areas make up 13.7%. It also showed that a majority of the high-risk urban areas are distributed around water-source reserves and the Pingshan River, and 87.6% of them are residential, industrial, and commercial lands in which the potential risks of non-point source pollution are high. Corrective actions should be considered on an urgent basis in high-risk urban areas, where low-impact development practices are considered effective in reducing non-point source pollution at the source. Validation results affirmed that the proposed ERA approach can reliably provide insights into the distribution of environmental risks in the study area. The proposed framework of ERA-based land-use planning is applicable to sustainable urban development and revitalization of other urban regions.
引用格式：Li Q, Yu Y, Jiang X, et al. Multifactor-based environmental risk assessment for sustainable land-use planning in Shenzhen, China. Science of The Total Environment, 2019, 657: 1051-1063.
供稿 / 賈凱、彭凱鋒、呂金霞、荔琢、王曉雅、蔣梓傑
製作 / 荔琢 指導 / 蔣衛國
跟蹤和發布生態環境、生態系統、生態水文、溼地生態、水資源、洪水災害、溼地資源、城市生態、城市溼地、海綿城市等方向的國內外學術研究進展、遙感和空間信息技術應用前沿資訊。
生態水文遙感前沿
微信號 : gh_f2514dbfc97d
聯繫方式：jiangweiguo@bnu.edu.cn