Search

Regional landslide identification is important for the risk management of landslide hazards. The traditional methods of regional landslide identification were mainly conducted by a human being. In previous studies, automatic landslide recognition mainly focused on new landslides distinct from the environment induced by rainfall or earthquake, using the image classification method and semantic segmentation method of deep learning. However, there is a lack of research on the automatic recognition of old loess landslides, which are difficult to distinguish from the environment. Therefore, this study uses the object detection method of deep learning to identify old loess landslides with Google Earth images. At first, a database of loess historical landslide samples was established for deep learning based on Google Earth images. A total of 6111 landslides were interpreted in three landslide areas in Gansu Province, China. Second, three object detection algorithms including the one-stage algorithm RetinaNet and YOLO v3 and the two-stage algorithm Mask R-CNN, were chosen for automatic landslide identification. Mask R-CNN achieved the greatest accuracy, with an AP of 18.9% and F1-score of 55.31%. Among the three landslide areas, the order of identification accuracy from high to low was Site 1, Site 2, and Site 3, with the F1-scores of 62.05%, 61.04% and 50.88%, respectively, which were positively related to their recognition difficulty. The research results proved that the object detection method can be employed for the automatic identification of loess landslides based on Google Earth images.

Accurate and repeated forest inventory data are critical to understand forest ecosystem processes and manage forest resources. In recent years, unmanned aerial vehicle (UAV)-borne light detection and ranging (lidar) systems have demonstrated effectiveness at deriving forest inventory attributes. However, their high cost has largely prevented them from being used in large-scale forest applications. Here, we developed a very low-cost UAV lidar system that integrates a recently emerged DJI Livox MID40 laser scanner (~$600 USD) and evaluated its capability in estimating both individual tree-level (i.e., tree height) and plot-level forest inventory attributes (i.e., canopy cover, gap fraction, and leaf area index (LAI)). Moreover, a comprehensive comparison was conducted between the developed DJI Livox system and four other UAV lidar systems equipped with high-end laser scanners (i.e., RIEGL VUX-1 UAV, RIEGL miniVUX-1 UAV, HESAI Pandar40, and Velodyne Puck LITE). Using these instruments, we surveyed a coniferous forest site and a broadleaved forest site, with tree densities ranging from 500 trees/ha to 3000 trees/ha, with 52 UAV flights at different flying height and speed combinations. The developed DJI Livox MID40 system effectively captured the upper canopy structure and terrain surface information at both forest sites. The estimated individual tree height was highly correlated with field measurements (coniferous site: R2 = 0.96, root mean squared error/RMSE = 0.59 m; broadleaved site: R2 = 0.70, RMSE = 1.63 m). The plot-level estimates of canopy cover, gap fraction, and LAI corresponded well with those derived from the high-end RIEGL VUX-1 UAV system but tended to have systematic biases in areas with medium to high canopy densities. Overall, the DJI Livox MID40 system performed comparably to the RIEGL miniVUX-1 UAV, HESAI Pandar40, and Velodyne Puck LITE systems in the coniferous site and to the Velodyne Puck LITE system in the broadleaved forest. Despite its apparent weaknesses of limited sensitivity to low-intensity returns and narrow field of view, we believe that the very low-cost system developed by this study can largely broaden the potential use of UAV lidar in forest inventory applications. This study also provides guidance for the selection of the appropriate UAV lidar system and flight specifications for forest research and management.

Mangrove forest ecosystems are distributed at the land–sea interface in tropical and subtropical regions and play an important role in carbon cycles and biodiversity. Accurately mapping global mangrove aboveground biomass (AGB) will help us understand how mangrove ecosystems are affected by the impacts of climatic change and human activities. Light detection and ranging (LiDAR) techniques have been proven to accurately capture the three-dimensional structure of mangroves and LiDAR can estimate forest AGB with high accuracy. In this study, we produced a global mangrove forest AGB map for 2004 at a 250-m resolution by combining ground inventory data, spaceborne LiDAR, optical imagery, climate surfaces, and topographic data with random forest, a machine learning method. From the published literature and free-access datasets of mangrove biomass, we selected 342 surface observations to train and validate the mangrove AGB estimation model. Our global mangrove AGB map showed that average global mangrove AGB density was 115.23 Mg/ha, with a standard deviation of 48.89 Mg/ha. Total global AGB storage within mangrove forests was 1.52 Pg. Cross-validation with observed data demonstrated that our mangrove AGB estimates were reliable. The adjusted coefficient of determination (R2) and root-mean-square error (RMSE) were 0.48 and 75.85 Mg/ha, respectively. Our estimated global mangrove AGB storage was similar to that predicted by previous remote sensing methods, and remote sensing approaches can overcome overestimates from climate-based models. This new biomass map provides information that can help us understand the global mangrove distribution, while also serving as a baseline to monitor trends in global mangrove biomass.

The Shuttle Radar Topography Mission (SRTM) digital elevation model (DEM) is one of the most complete and frequently used global-scale DEM products in various applications. However, previous studies have shown that the SRTM DEM is systematically higher than the actual land surface in vegetated mountain areas. The objective of this study is to propose a procedure to calibrate the SRTM DEM over large vegetated mountain areas. Firstly, we developed methods to estimate canopy cover from aerial imagery and tree height from multi-source datasets (i.e., field observations, airborne light detection and ranging (LiDAR) data, Geoscience Laser Altimeter System (GLAS) data, Landsat TM imagery, climate surfaces, and topographic data). Then, the airborne LiDAR derived DEM, covering ~5% of the study area, was used to evaluate the accuracy of the SRTM DEM. Finally, a regression model of the SRTM DEM error depending on tree height, canopy cover, and terrain slope was developed to calibrate the SRTM DEM. Our results show that the proposed procedure can significantly improve the accuracy of the SRTM DEM over vegetated mountain areas. The mean difference between the SRTM DEM and the LiDAR DEM decreased from 12.15 m to −0.82 m, and the standard deviation dropped by 2 m.

Urban areas are one of the most important components of human society. Their extents have been continuously growing during the last few decades. Accurate and timely measurements of the extents of urban areas can help in analyzing population densities and urban sprawls and in studying environmental issues related to urbanization. Urban extents detected from remotely sensed data are usually a by-product of land use classification results, and their interpretation requires a full understanding of land cover types. In this study, for the first time, we mapped urban extents in the continental United States using a novel one-class classification method, i.e., positive and unlabeled learning (PUL), with multi-temporal Moderate Resolution Imaging Spectroradiometer (MODIS) data for the year 2010. The Defense Meteorological Satellite Program Operational Linescan System (DMSP-OLS) night stable light data were used to calibrate the urban extents obtained from the one-class classification scheme. Our results demonstrated the effectiveness of the use of the PUL algorithm in mapping large-scale urban areas from coarse remote-sensing images, for the first time. The total accuracy of mapped urban areas was 92.9% and the kappa coefficient was 0.85. The use of DMSP-OLS night stable light data can significantly reduce false detection rates from bare land and cropland far from cities. Compared with traditional supervised classification methods, the one-class classification scheme can greatly reduce the effort involved in collecting training datasets, without losing predictive accuracy.

Mapping mangrove extent and species is important for understanding their response to environmental changes and for observing their integrity for providing goods and services. However, accurately mapping mangrove extent and species are ongoing challenges in remote sensing. The newly-launched and freely-available Sentinel-2 (S2) sensor offers a new opportunity for these challenges. This study presents the first study dedicated to the examination of the potential of original bands, spectral indices, and texture information of S2 in mapping mangrove extent and species in the first National Nature Reserve for mangroves in Dongzhaigang, China. To map mangrove extent and species, a three-level hierarchical structure based on the spatial structure of a mangrove ecosystem and geographic object-based image analysis is utilized and modified. During the experiments, to conquer the challenge of optimizing high-dimension and correlated feature space, the recursive feature elimination (RFE) algorithm is introduced. Finally, the selected features from RFE are employed in mangrove species discriminations, based on a random forest algorithm. The results are compared with those of Landsat 8 (L8) and Pléiades-1 (P1) data and show that S2 and L8 could accurately extract mangrove extent, but P1 obviously overestimated it. Regarding mangrove species community levels, the overall classification accuracy of S2 is 70.95%, which is lower than P1 imagery (78.57%) and slightly higher than L8 data (68.57%). Meanwhile, the former difference is statistically significant, and the latter is not. The dominant species is extracted basically in S2 and P1 imagery, but for the occasionally distributed K. candel and the pioneer and fringe mangrove A. marina, S2 performs poorly. Concerning L8, S2, and P1, there are eight (8/126), nine (9/218), and eight (8/73) features, respectively, that are the most important for mangrove species discriminations. The most important feature overall is the red-edge bands, followed by shortwave infrared, near infrared, blue, and other visible bands in turn. This study demonstrates that the S2 sensor can accurately map mangrove extent and basically discriminate mangrove species communities, but for the latter, one should be cautious due to the complexity of mangrove species.

Canopy height is an important forest structure parameter for understanding forest ecosystem and improving global carbon stock quantification accuracy. Light detection and ranging (LiDAR) can provide accurate canopy height measurements, but its application at large scales is limited. Using LiDAR-derived canopy height as ground truth to train the Random Forest (RF) algorithm and therefore predict canopy height from other remotely sensed datasets in areas without LiDAR coverage has been one of the most commonly used method in large-scale canopy height mapping. However, how variances in location, vegetation type, and spatial scale of study sites influence the RF modelling results is still a question that needs to be addressed. In this study, we selected 16 study sites (100 km2 each) with full airborne LiDAR coverage across the United States, and used the LiDAR-derived canopy height along with optical imagery, topographic data, and climate surfaces to evaluate the transferability of the RF-based canopy height prediction method. The results show a series of findings from general to complex. The RF model trained at a certain location or vegetation type cannot be transferred to other locations or vegetation types. However, by training the RF algorithm using samples from all sites with various vegetation types, a universal model can be achieved for predicting canopy height at different locations and different vegetation types with self-predicted R2 higher than 0.6 and RMSE lower than 6 m. Moreover, the influence of spatial scales on the RF prediction accuracy is noticeable when spatial extent of the study site is less than 50 km2 or the spatial resolution of the training pixel is finer than 500 m. The canopy height prediction accuracy increases with the spatial extent and the targeted spatial resolution.

In the dwindling natural mangrove today, mangrove reforestation projects are conducted worldwide to prevent further losses. Due to monoculture and the low survival rate of artificial mangroves, it is necessary to pay attention to mapping and monitoring them dynamically. Remote sensing techniques have been widely used to map mangrove forests due to their capacity for large-scale, accurate, efficient, and repetitive monitoring. This study evaluated the capability of a 0.5-m Pléiades-1 in classifying artificial mangrove species using both pixel-based and object-based classification schemes. For comparison, three machine learning algorithms—decision tree (DT), support vector machine (SVM), and random forest (RF)—were used as the classifiers in the pixel-based and object-based classification procedure. The results showed that both the pixel-based and object-based approaches could recognize the major discriminations between the four major artificial mangrove species. However, the object-based method had a better overall accuracy than the pixel-based method on average. For pixel-based image analysis, SVM produced the highest overall accuracy (79.63%); for object-based image analysis, RF could achieve the highest overall accuracy (82.40%), and it was also the best machine learning algorithm for classifying artificial mangroves. The patches produced by object-based image analysis approaches presented a more generalized appearance and could contiguously depict mangrove species communities. When the same machine learning algorithms were compared by McNemar’s test, a statistically significant difference in overall classification accuracy between the pixel-based and object-based classifications only existed in the RF algorithm. Regarding species, monoculture and dominant mangrove species Sonneratia apetala group 1 (SA1) as well as partly mixed and regular shape mangrove species Hibiscus tiliaceus (HT) could well be identified. However, for complex and easily-confused mangrove species Sonneratia apetala group 2 (SA2) and other occasionally presented mangroves species (OT), only major distributions could be extracted, with an accuracy of about two-thirds. This study demonstrated that more than 80% of artificial mangroves species distribution could be mapped.

Light detection and ranging (Lidar) data can be used to create wall-to-wall forest structure and fuel products that are required for wildfire behavior simulation models. We know that Lidar-derived forest parameters have a non-negligible error associated with them, yet we do not know how this error influences the results of fire behavior modeling that use these layers as inputs. Here, we evaluated the influence of error associated with two Lidar data products—canopy height (CH) and canopy base height (CBH)—on simulated fire behavior in a case study in the Sierra Nevada, California, USA. We used a Monte Carlo simulation approach with expected randomized error added to each model input. Model 1 used the original, unmodified data, Model 2 incorporated error in the CH layer, and Model 3 incorporated error in the CBH layer. This sensitivity analysis showed that error in CH and CBH did not greatly influence the modeled conditional burn probability, fire size, or fire size distribution. We found that the expected error associated with CH and CBH did not greatly influence modeled results: conditional burn probability, fire size, and fire size distributions were very similar between Model 1 (original data), Model 2 (error added to CH), and Model 3 (error added to CBH). However, the impact of introduced error was more pronounced with CBH than with CH, and at lower canopy heights, the addition of error increased modeled canopy burn probability. Our work suggests that the use of Lidar data, even with its inherent error, can contribute to reliable and robust estimates of modeled forest fire behavior, and forest managers should be confident in using Lidar data products in their fire behavior modeling workflow.

Automatic classification of light detection and ranging (LiDAR) data in urban areas is of great importance for many applications such as generating three-dimensional (3D) building models and monitoring power lines. Traditional supervised classification methods require training samples of all classes to construct a reliable classifier. However, complete training samples are normally hard and costly to collect, and a common circumstance is that only training samples for a class of interest are available, in which traditional supervised classification methods may be inappropriate. In this study, we investigated the possibility of using a novel one-class classification algorithm, i.e., the presence and background learning (PBL) algorithm, to classify LiDAR data in an urban scenario. The results demonstrated that the PBL algorithm implemented by back propagation (BP) neural network (PBL-BP) could effectively classify a single class (e.g., building, tree, terrain, power line, and others) from airborne LiDAR point cloud with very high accuracy. The mean F-score for all of the classes from the PBL-BP classification results was 0.94, which was higher than those from one-class support vector machine (SVM), biased SVM, and maximum entropy methods (0.68, 0.82 and 0.93, respectively). Moreover, the PBL-BP algorithm yielded a comparable overall accuracy to the multi-class SVM method. Therefore, this method is very promising in the classification of the LiDAR point cloud.

As a large carbon pool, global forest ecosystems are a critical component of the global carbon cycle. Accurate estimations of global forest aboveground biomass (AGB) can improve the understanding of global carbon dynamics and help to quantify anthropogenic carbon emissions. Light detection and ranging (LiDAR) techniques have been proven that can accurately capture both horizontal and vertical forest structures and increase the accuracy of forest AGB estimation. In this study, we mapped the global forest AGB density at a 1-km resolution through the integration of ground inventory data, optical imagery, Geoscience Laser Altimeter System/Ice, Cloud, and Land Elevation Satellite data, climate surfaces, and topographic data. Over 4000 ground inventory records were collected from published literatures to train the forest AGB estimation model and validate the resulting global forest AGB product. Our wall-to-wall global forest AGB map showed that the global forest AGB density was 210.09 Mg/ha on average, with a standard deviation of 109.31 Mg/ha. At the continental level, Africa (333.34 ± 63.80 Mg/ha) and South America (301.68 ± 67.43 Mg/ha) had higher AGB density. The AGB density in Asia, North America and Europe were 172.28 ± 94.75, 166.48 ± 84.97, and 132.97 ± 50.70 Mg/ha, respectively. The wall-to-wall forest AGB map was evaluated at plot level using independent plot measurements. The adjusted coefficient of determination (R2) and root-mean-square error (RMSE) between our predicted results and the validation plots were 0.56 and 87.53 Mg/ha, respectively. At the ecological zone level, the R2 and RMSE between our map and Intergovernmental Panel on Climate Change suggested values were 0.56 and 101.21 Mg/ha, respectively. Moreover, a comprehensive comparison was also conducted between our forest AGB map and other published regional AGB products. Overall, our forest AGB map showed good agreements with these regional AGB products, but some of the regional AGB products tended to underestimate forest AGB density.