Exploring deep learning for landslide mapping:A comprehensive review

A detailed and accurate inventory map of landslides is crucial for quantitative hazard assessment and land planning.Traditional methods relying on change detection and object-oriented approaches have been criticized for their dependence on expert knowledge and subjective factors.Recent advancements in highresolution satellite imagery,coupled with the rapid development of artificial intelligence,particularly datadriven deep learning algorithms(DL)such as convolutional neural networks(CNN),have provided rich feature indicators for landslide mapping,overcoming previous limitations.In this review paper,77representative DL-based landslide detection methods applied in various environments over the past seven years were examined.This study analyzed the structures of different DL networks,discussed five main application scenarios,and assessed both the advancements and limitations of DL in geological hazard analysis.The results indicated that the increasing number of articles per year reflects growing interest in landslide mapping by artificial intelligence,with U-Net-based structures gaining prominence due to their flexibility in feature extraction and generalization.Finally,we explored the hindrances of DL in landslide hazard research based on the above research content.Challenges such as black-box operations and sample dependence persist,warranting further theoretical research and future application of DL in landslide detection.

Some recent advances in remote sensing-based monitoring of changes in the Greenland Ice Sheet

The mass balance of the Greenland Ice Sheet(GrIS)plays a crucial role in global sea level change.Since the 1960s,remote sensing missions have been providing extensive and continuous observation data for change monitoring of the GrIS.In this paper,we present our recent research results from remote sensing-based GrIS change monitoring.First,historical satellite data are processed and used to fill data gaps and are combined with existing partial maps,completing an ice velocity map of the GrIS from the 1960s to 1980s.This map provides valuable data for estimating the historical mass balance of Greenland.Second,the monthly gravimetry-based mass balance of the GrIS from 2002 to 2020 is estimated by combining Gravity Recovery and Climate Experiment(GRACE)and GRACE Follow On(GRACE-FO)data.It is found that the GrIS has lost a total mass of approximately 4443±75 Gt during this period.Third,based on Global Land Ice Measurements from Space(GLIMS),an updated Greenland glacier inventory is achieved utilizing data collected between 2006 and 2020.This inventory provides more detailed and up-to-data glacier boundaries of Greenland.Overall,these advances provide essential data support for estimating the mass balance of the GrIS,contributing to the advancement of research on global sea level change.

Methods for ice velocity mapping in West Antarctica using historical optical satellite images

The Antarctic Ice Sheet harbors more than 90%of the Earth ice mass,with significant losses experienced through dynamic thinning,particularly in West Antarctica.The crucial aspect of investigating ice mass balance in historical periods preceding 1990 hinges on the utilization of ice velocities derived from optical satellite images.We employed declassified satellite images and Landsat images with normalized cross correlation based image matching,adopting an adaptive combination of skills and methods to overcome challenges encountered during the mapping of historical ice velocity in West Antarctica.A basin-wide synthesis velocity map encompassing the coastal regions of most large-scale glaciers and ice shelves in West Antarctica has already been successfully generated.Our results for historical ice velocities cover over 70%of the grounding line in most of the West Antarctic basins.Through adjustments,we uncovered overestimations in ice velocity measurements over an extended period,transforming our ice velocity map into a spatially deterministic,temporally average version.Among all velocity measurements,Thwaites Glacier exhibited a notable spatial variation in the fastest ice flowline and velocity distribution.Overestimation distributions on Thwaites Glacier displayed a clear consistency with the positions of subsequent front calving events,offering insights into the instabilities of ice shelves.

Building Facade Point Clouds Segmentation Based on Optimal Dual-Scale Feature Descriptors

To address the current issues of inaccurate segmentation and the limited applicability of segmentation methods for building facades in point clouds, we propose a facade segmentation algorithm based on optimal dual-scale feature descriptors. First, we select the optimal dual-scale descriptors from a range of feature descriptors. Next, we segment the facade according to the threshold value of the chosen optimal dual-scale descriptors. Finally, we use RANSAC (Random Sample Consensus) to fit the segmented surface and optimize the fitting result. Experimental results show that, compared to commonly used facade segmentation algorithms, the proposed method yields more accurate segmentation results, providing a robust data foundation for subsequent 3D model reconstruction of buildings.

Positional Error Model of Line Segments with Modeling and Measuring Errors Using Brownian Bridge

Spatial linear features are often represented as a series of line segments joined by measured endpoints in surveying and geographic information science.There are not only the measuring errors of the endpoints but also the modeling errors between the line segments and the actual geographical features.This paper presents a Brownian bridge error model for line segments combining both the modeling and measuring errors.First,the Brownian bridge is used to establish the position distribution of the actual geographic feature represented by the line segment.Second,an error propagation model with the constraints of the measuring error distribution of the endpoints is proposed.Third,a comprehensive error band of the line segment is constructed,wherein both the modeling and measuring errors are contained.The proposed error model can be used to evaluate line segments’overall accuracy and trustability influenced by modeling and measuring errors,and provides a comprehensive quality indicator for the geospatial data.

Maintenance of Millimeter-level Geodetic Reference Framework

The high-precision terrestrial reference frame,as the spatial benchmark for geodesy,is an important national infrastructure.However,due to the influence of nonlinear factors related to geophysical phenomena,the overall maintenance accuracy of the ITRF framework is still at the centimeter level.Therefore,accurately characterizing the true trajectories of linear motion,nonlinear motion,and geocentric motion of the reference station is the key to achieve the construction and maintenance technology of a millimeter level terrestrial reference framework.Based on long-term global and regional GNSS observation data,more Chinese geodesy scientists devoted much efforts to the maintenance of millimeter-level geodetic reference framework.The main contributions of this work included the followings:①Dynamic maintenance of millimeter-level terrestrial reference frame;②Research progress on the method of maintenance of regional reference frame based on GNSS;③The progress of CGCS2000 frame maintenance in millimeter level accuracy;④Reprocessing and reanalysis of two-decade GNSS observation in continental China;⑤Research on current GNSS velocity field model and deformation in Chinese mainland;⑥The preliminary realization and evaluation of CTRF2020.

Land Use/Land Cover Changes Detection in Lagos City of Nigeria Using Remote Sensing and GIS

The rapid urbanization and population growth of Lagos City, Nigeria, have led to a significant change in land use and cover over the past two decades. The primary objective of this research was to assess the changes in land use and cover and forecast future trends in Lagos for the sustainable development of urbanization. The study utilized remote sensing and GIS technologies to monitor and identify the land use and cover of Lagos from 2000 to 2020. The CA Markov artificial neural network technique for cellular automata was employed to predict changes in land use and cover from 2020 to 2030. In addition, the post-classification comparison method was used to detect changes in classified classes in land use and cover. The study classified satellite images for 2000, 2010, and 2020 to develop land use and cover maps using ERDAS Imagine. The classification was based on six categories, namely 1) water bodies, 2) built-up, 3) bare land, 4) forest, 5) vegetation, and 6) wetlands. The results showed that: 1) the vegetation cover, wetlands, built-up areas, forests, and bare land have undergone significant changes over the past two decades. Built-up areas, wetlands, and forests have increased by 33.57%, 1.01%, and 21.37%, respectively, while vegetation, bare land, and water have decreased by 21.77%, 5.14%, and 17.13%, respectively. 2) Moreover, during 2020-2030, it is projected that 19.18% of forests and 16% of vegetation will decline, while 5.27% of barren land, 0.82% of wetlands, and 15.83% of water will increase. The urban area will be expanded by 42.44%. 3) The simulated results showed that the correction percentage was 82.43%, and the global kappa value was 0.85. The study found that the expansion of urban built-up areas due to population growth was the primary driver of the changes in land use and cover in Lagos. This research provides crucial insights that contribute to sustainable planning and management and helps us better understand the changes in land use and cover in Lagos.

The impact of spatial scale on local Moran's I clustering of annual fishing effort for Dosidicus gigas offshore Peru

The spatial scale(?shing grid) of ?sheries research af fects the observed spatial patterns of?sheries resources such as catch-per-unit-ef fort(CPUE) and ?shing ef fort. We examined the scale impact of high value(HH) clusters of the annual ?shing ef fort for Dosidicus gigas of fshore Peru from 2009 to 2012.For a multi-scale analysis, the original commercial ?shery data were tessellated to twelve spatial scales from 6′ to 72′ with an interval of 6′. Under these spatial scales, D. gigas clusters were identi?ed using the Anselin Local Moran's I. Statistics including the number of points, mean CPUE, standard deviation(SD),skewness, kurtosis, area and centroid were calculated for these HH clusters. We found that the z-score of global Moran's I and the number of points for HH clusters follow a power law scaling relationship from2009 to 2012. The mean ef fort and its SD also follow a power law scaling relationship from 2009 to 2012.The skewness follows a linear scaling relationship in 2010 and 2011 but ?uctuates with spatial scale in2009 and 2012; kurtosis follows a logarithmic scale relationship in 2009, 2011 and 2012 but a linear scale relationship in 2010. Cluster area follows a power law scaling relationship in 2010 and 2012, a linear scaling relationship in 2009, and a quadratic scaling relationship in 2011. Based on the peaks of Moran's I indices and the multi-scale analysis, we conclude that the optimum scales are 12′ in 2009 ? 2011 and 6′ in 2012, while the coarsest allowable scales are 48′ in 2009, 2010 and 2012, and 60′ in 2011. Our research provides the best spatial scales for conducting spatial analysis of this pelagic species, and provides a better understanding of scaling behavior for the ?shing ef fort of D. gigas in the of fshore Peruvian waters.

Landslide Susceptibility Assessment of the Youfang Catchment using Logistic Regression

A detailed landslide susceptibility map was produced in the Youfang catchment using logistic regression method with datasets developed for a geographic information system(GIS).Known as one of the most landslide-prone areas in China, the Youfang catchment of Longnan mountain region,which lies in the transitional area among QinghaiTibet Plateau, loess Plateau and Sichuan Basin, was selected as a representative case to evaluate the frequency and distribution of landslides.Statistical relationships for landslide susceptibility assessment were developed using landslide and landslide causative factor databases.Logistic regression(LR)was used to create the landslide susceptibility maps based on a series of available data sources: landslide inventory; distance to drainage systems, faults and roads; slope angle and aspect; topographic elevation and topographical wetness index, and land use.The quality of the landslide susceptibility map produced in this paper was validated and the result can be used fordesigning protective and mitigation measures against landslide hazards.The landslide susceptibility map is expected to provide a fundamental tool for landslide hazards assessment and risk management in the Youfang catchment.

Analysis of the wide area differential correction for BeiDou global satellite navigation system

The regional BeiDou Satellite System, or BDS2, broadcasts a differential correction as Equivalent Satellite Clock Correction to correct both orbit and satellite clock errors. For the global BDS, or BDS3, satellite orbit and clock corrections conforming with RTCA standards will be broadcast to authorized users. The hybrid constellation and regional monitoring network pose challenges for the high precision separation of orbit and satellite clock corrections. Three correction models of kinematic,dynamic and Two-way Satellite Time Frequency Transfer(TWSTFT)-based dynamic were studied to estimate the satellite orbit and clock corrections. The correction accuracy of the three models is compared and analyzed based on the BDS observation data. Results show that the accuracies(root mean square, RMS) of dual-frequency real-time positioning for the three models are about 1.76 m, 1.78 m and 2.08 m respectively, which are comparable with the performance of WAAS and EGNOS. With dynamic corrections, the precision of Precise Point Positioning(PPP) experiments may reach about 23 cm after convergence.

GRACE RL05-based ice mass changes in the typical regions of antarctica from 2004 to 2012

The Antarctic ice sheet is the largest block of ice on Earth, a tiny change of its ice sheet will have a significant impact on sea level change, so it plays an important role in global climate change. The Gravity Recovery and Climate Experiment （GRACE） mission, launched in 2002, provides an alternative method to monitor the Antarctic ice mass variation. The latest Release Level 05 （ RL05 ） version of GRACE time-variable gravity （TVG） data, derived from GRACE observations with improved quality and time-span over 10 years, were released by three GRACE data centers （CSR, JPL and GFZ） in April 2012, which gives us a chance to re-estimate the ice mass change over Antarctic more accurately. In this paper, we examine ice mass changes in regional scale, including Antarctic Peninsula （AP, West Antarctica）, Amundsen Sea Embayment （ASE, West Antarctica）, Lambert-Amery System （LAS, East Antarctica） and 27 drainage basins based on three data sets. The AP mass change rates are -12.03±0.74 Gt/a （CSR, 2004-2012）, -13.92±2.33 Gt/a （JPL, 2004 -2012） , -12.28±0.76 Gt/a （GFZ, 2005-2012） , with an acceleration of -1.50±0.25 Gt/a^2, -1.54±0.26 Gt/a^2, -0. 46±0.28 Gt/a^2 respectively, the ASE mass change rates are -89.22±1.93 Gt/a, -86.28± 2.20 Gt/a, -83.67±1.76 Gt/a with an acceleration of -10. 03±0. 65 Gt/a^2, -8.74±0. 74 Gt/a^2 and -5.69 ±0.68 Gt/a^2, and the LAS mass ehange rates are -4.31±1.95 Gt/a, -7.29±2. 84 Gt/a, 1.20±1.35 Gt/a with an acceleration of -0. 18±0.62 Gt/a^2, 3.55±0.95 Gt/a^2 and 0.97±0.49 Gt/a^2. The mass change rates derived from the three RL05 data are very close to each other both in AP and ASE with the uncertainties much smaller than the change rates, and mass losses are significantly accelerated since 2007 in AP and 2006 in ASE, respectively. However, the mass change rates are significantly different in LAS, negative rate from CSR and JPL data, but positive rate from GFZ data, the uncertainties are even larger than the correspondent change rates. With regard to the 27 drainage basins, seven basins （basin 3-9） located in the east Antarctica show positive mass change rates, and the rest twenty basins are characterized by negative mass change rates during the time span of the three RL05 data.

Multipath Error Analysis of COMPASS Triple Frequency Observations

The BeiDou-2 satellite navigation system broadcasts triple frequency data. In this paper, the pseudorange multipath is extracted by using the geometry-free and ionosphere-free combination of one pseudorange and two phase measurements, and the phase multipath is extracted by using triple frequency phase measurements, respectively. By using several days’ static observation data, we exact the noisy pseudorange and phase multipath of three types of satellites, GEO, IGSO and MEO satellites. Because of the low frequency characteristics of the multipath, the low frequency wavelet filter is further used to recover the high-precision low frequency multipath signals that are specified by their amplitudes, periods and phases. The results show that the multipath periods are about 86160s, 86158s and 46391s for GEO, IGSO and MEO satellites, respectively, which coincide with that of the corresponding satellite orbits. Then we use the extracted multipath signals to correct the pseudorange measurements in order to improve the accuracy of point positioning. The positioning accuracy in East-West direction can be significantly improved by using the multipath corrected pseudorange measurements, and in the other two directions the positioning accuracy can also be improved to some extent.

An iterative algorithm of NWTLS-EC for three dimensional-datum transformation with large rotation angle

The Gauss-Markov （GM） model and the Errors-in-Variables （EIV） model are frequently used to perform 3D coordinate transformations in geodesy and engineering surveys. In these applications, because the observation errors in original coordinates system are also taken into account, the latter is more accurate and reasonable than the former. Although the Weighted Total Least Squares （WTLS） technique has been intro- duced into coordinate transformations as the measured points are heteroscedastic and correlated, the Variance- Covariance Matrix （VCM） of observations is restricted by a particular structure, namely, only the correlations of each points are taken into account. Because the 3D datum transformation with large rotation angle is a non- linear problem, the WTLS is no longer suitable in this ease. In this contribution, we suggested the nonlinear WTLS adjustments with equality constraints （NWTLS-EC） for 3D datum transformation with large rotation an- gle, which removed the particular structure restriction on the VCM. The Least Squares adjustment with Equality （LSE） constraints is employed to solve NWTLS-EC as the nonlinear model has been linearized, and an iterative algorithm is proposed with the LSE solution. A simulation study of 3D datum transformation with large rotation angle is given to insight into the feasibility of our algorithm at last.

Spatial pattern analysis of forest trees based on the vectorial mark

Analysis of spatial patterns to describe the spatial correlation between a tree location and marks(i.e.,structural variables),can reveal stand history,population dynamics,competition and symbiosis.However,most studies of spatial patterns have concentrated on tree location and tree sizes rather than on crown asymmetry especially with direct analysis among marks characterizing facilitation and competition among of trees,and thus cannot reveal the cause of the distributions of tree locations and quantitative marks.To explore the spatial correlation among quantitative and vectorial marks and their implication on population dynamics,we extracted vertical and horizontal marks(tree height and crown projection area)characterizing tree size,and a vectorial mark(crown displacement vector characterizing the crown asymmetry)using an airborne laser scanning point cloud obtained from two forest stands in Oxfordshire,UK.Quantitatively and vectorially marked spatial patterns were developed,with corresponding null models established for a significance test.We analyzed eight types of univariate and bivariate spatial patterns,after first proposing four types.The accuracy of the pattern analysis based on an algorithm-segmented point cloud was compared with that of a truly segmented point cloud.The algorithm-segmented point cloud managed to detect 70–86%of patterns correctly.The eight types of spatial patterns analyzed the spatial distribution of trees,the spatial correlation between tree size and facilitated or competitive interactions of sycamore and other species.These four types of univariate patterns jointly showed that,at smaller scales,the trees tend to be clustered,and taller,with larger crowns due to the detected facilitations among trees in the study area.The four types of bivariate patterns found that at smaller scales there are taller trees and more facilitation among sycamore and other species,while crown size is mostly homogeneous across scales.These results indicate that interspecific facilitation and competition mainly affect tree height in the study area.This work further confirms the connection of tree size with individual facilitation and competition,revealing the potential spatial structure that previously was hard to detect.

Variations in China's terrestrial water storage over the past decade using GRACE data

Variations in China’s terrestrial water storage from March 2003 to February 2013 were determined using data from the Gravity Recovery and Climate Experiment(GRACE)monthly gravity field model provided by the Center for Space Research. The results were compared with the variations in surface water estimated using the Global Land Data Assimilation System(GLDAS) hydrological model. The results indicated a decline in terrestrial water storage in the Shanxi and Xinjiang Tianshan regions over the past decade with a downward trend that reached 7.76 ± 0.71 mm/a and 5.8 ± 0.67 mm/a, respectively. Anthropogenic activities were considered to be the major cause of this terrestrial water loss(especially groundwater) in these regions. In contrast, the intersection of the Xinjiang and Tibet Autonomous Regions and the Qinghai Province showed an upward trend in the terrestrial water storage at a rate of 9.06 ± 0.37 mm/a, which is closely related to the high-quality local ecological environment and lack of human activities. At the intersections between the Chongqing, Guizhou, and Hunan Provinces and between the Jiangxi, Zhejiang, and Fujian Provinces, the terrestrial water storage increased at rates of7.86 ± 0.9 and 8.68 ± 0.8 mm/a, respectively. These two regions received abundant annual precipitation; moreover, there was no considerable variation in the amount of groundwater storage over the past decade. In addition, the empirical orthogonal function(EOF) method used in this study could eliminate correlated errors in the GRACE monthly gravity field model more effectively than the traditional polynomial fitting method, and it did not generate false signals.

Application of a simultaneous iterations reconstruction technique for a 3-D water vapor tomography system

The simultaneous iterations rithms of the ART family. It is used reconstruction technique （SIRT） widely in tomography because of is one of several reconstruction algoits convenience in dealing with large sparse matrices. Its theoretical background and iteration model are discussed at the beginning of this paper. Then, the implementation of the SIRT to reconstruct the three-dimensional distribution of water vapor by simulation is discussed. The results show that the SIRT can function effectively in water vapor tomography, obtain rapid convergence, and be implemented more easily than inversion.

Detecting spatio-temporal urban surface changes using identified temporary coherent scatterers

Synthetic aperture radar(SAR) is able to detect surface changes in urban areas with a short revisit time, showing its capability in disaster assessment and urbanization monitoring.Most presented change detection methods are conducted using couples of SAR amplitude images. However, a prior date of surface change is required to select a feasible image pair. We propose an automatic spatio-temporal change detection method by identifying the temporary coherent scatterers. Based on amplitude time series, χ^(2)-test and iterative single pixel change detection are proposed to identify all step-times: the moments of the surface change. Then the parameters, e.g., deformation velocity and relative height, are estimated and corresponding coherent periods are identified by using interferometric phase time series. With identified temporary coherent scatterers, different types of temporal surface changes can be classified using the location of the coherent periods and spatial significant changes are identified combining point density and F values. The main advantage of our method is automatically detecting spatio-temporal surface changes without prior information. Experimental results by the proposed method show that both appearing and disappearing buildings with their step-times are successfully identified and results by ascending and descending SAR images show a good agreement.

A public Cloud-based China's Landslide Inventory Database(Cs LID): development, zone, and spatiotemporal analysis for significant historical events, 1949-2011

Landslide inventory plays an important role in recording landslide events and showing their temporal-spatial distribution. This paper describes the development, visualization, and analysis of a China's Landslide Inventory Database(Cs LID) by utilizing Google's public cloud computing platform. Firstly, Cs LID(Landslide Inventory Database) compiles a total of 1221 historical landslide events spanning the years 1949-2011 from relevant data sources. Secondly, the Cs LID is further broken down into six zones for characterizing landslide cause-effect, spatiotemporal distribution, fatalities, and socioeconomic impacts based on the geological environment and terrain. The results show that among all the six zones, zone V, located in Qinba and Southwest Mountainous Area is the most active landslide hotspot with the highest landslide hazard in China. Additionally, the Google public cloud computing platform enables the Cs LID to be easily accessible, visually interactive, and with the capability of allowing new data input to dynamically augment the database. This work developed a cyber-landslide inventory and used it to analyze the landslide temporal-spatial distribution in China.

Monthly gravity field solution from GRACE range measurements using modified short arc approach

In this paper we present a series of monthly gravity field solutions from Gravity Recovery and Climate Experiment（GRACE） range measurements using modified short arc approach,in which the ambiguity of range measurements is eliminated via differentiating two adjacent range measurements.The data used for developing our monthly gravity field model are same as Tongji-GRACEOl model except that the range measurements are used to replace the range rate measurements,and our model is truncated to degree and order 60,spanning Jan.2004 to Dec.2010 also same as Tongji-GRACE01 model.Based on the comparison results of the C_（2,0）,C_（2,1）,S_（2,1）,and C_（15,15）,S_（15,15）,time series and the global mass change signals as well as the mass change time series in Amazon area of our model with those of Tongji-GRACE01 model,we can conclude that our monthly gravity field model is comparable with Tongji-GRACE01 monthly model.

A Robust Fault Detection Algorithm for the GNSS/INS Integrated Navigation Systems

Integrity is an important index for GNSS-based navigation and positioning, and the receiver autonomous integrity monitoring (RAIM) algorithm has been presented for integrity applications. In the integrated navigation systems of a global navigation satellite system (GNSS) and inertial navigation system (INS),the conventional RAIM algorithm has been developed to extended receiver autonomous integrity monitoring (ERAIM). However, the ERAIM algorithm may fail and a false alarm may generate once the measurements are contaminated by significant outliers, and this problem is rarely discussed in the existing literatures. In this paper, a robust fault detection and the corresponding data processing algorithm are proposed based on the ERAIM algorithm and the robust estimation. In the proposed algorithm, weights of the measurements are adjusted with the equivalent weight function, and the efficiency of the outlier detection and identification is improved, therefore, the estimates become more reliable, and the probability of the false alarm is decreased. Experiments with the data collected under actual environments are implemented, and results indicate that the proposed algorithm is more efficient than the conventional ERAIM algorithm for multiple outliers and a better filtering performance is achieved.