: A system is described which: (a) replaces existing sets of diverse terrain indices with a group of statistics for point-characteristics; (b) calculates all of these statistics in a single computer run from a single data set; and (c) utilizes available altitude matrix data The procedures are applicable to Altitude matrix data at any grid mesh From altitudes in each 3 x 3 submatrix, a quadratic surface is fitted and solved for its first and second horizontal and vertical derivatives at the central point This yields the slope gradient, slope aspect, profile convexity, and plan convexity at every point in the original matrix, except for the peripheral rows and columns These 'point' descriptors are presented as: (1) line-printer shaded maps; (2) histograms; (3) scatter plots of each pair; (4) matrix of pair-wise correlations, plus circular regressions on aspect, and several multiple regressions, and (5) summary (moment-based) statistics In general, the five basic descriptors have little relation to each other, except that gradient is usually a quadratic function of altitude A comparison is made with other approaches, such as spectral analysis and fractal modelling The long-distance persistence properties of terrain mean that considerable extra variance at long wavelengths is usually incorporated when the study area is extended Hence the auto correlation function varies with the length of series or size of area Variance of derivatives are also affected, but means, skews, and kurtoses are not

Statistical Characterization of Altitude Matrices by Computer. Report 6. An Integrated System of Terrain Analysis and Slope Mapping.

Himalayan glacier tongues are commonly debris covered and they are an important source of melt water. However, they remain relatively unstudied because of the inaccessibility of the terrain and the difficulties in field work caused by the thick debris mantles. Observations of debris-covered glaciers are therefore scarce and airborne remote sensing may bridge the gap between scarce field observations and coarse resolution space-borne remote sensing. In this study we deploy an Unmanned Aerial Vehicle (UAV) before and after the melt and monsoon season (May and October 2013) over the debris-covered tongue of the Lirung Glacier in Nepal. Based on stereo-imaging and the structure for motion algorithm we derive highly detailed ortho-mosaics and digital elevation models (DEMs), which we geometrically correct using differential GPS observations collected in the field. Based on DEM differencing and manual feature tracking we derive the mass loss and the surface velocity of the glacier at a high spatial accuracy. On average, mass loss is limited and the surface velocity is very small. However, the spatial variability of melt rates is very high, and ice cliffs and supra-glacial ponds show mass losses that can be an order of magnitude higher than the average. We suggest that future research should focus on the interaction between supra-glacial ponds, ice cliffs and englacial hydrology to further understand the dynamics of debris-covered glaciers. Finally, we conclude that UAV deployment has large potential in glaciology and it may revolutionize methods currently applied in studying glacier surface features.

High-Resolution Monitoring of Himalayan Glacier Dynamics Using Unmanned Aerial Vehicles

Most fi eld erosion studies in agricultural areas provide little information on the probable errors involved. Here, for the fi rst time, we compare the accuracy, time and cost of conventional and new methodologies for gully surveying, and provide a model to estimate the effort required to achieve a specifi ed accuracy. Using a terrestrial LiDAR survey of a 7.1-m-long gully reach as a benchmark data set, the accuracies of different measurement methods (a new 3D photo-reconstruction technique, total station, laser profi lemeter, and pole) are assessed for estimating gully erosion at a reach scale. Based on further fi eld measurements performed over nine gullies (>100 m long), a simulation approach is derived to model the expected volume errors when 2D methods are used at the gully scale. All gullies considered were located near Cordoba, Spain. At the reach scale, the fi eld measurements using 3D photo-reconstruction and total station techniques produced cross-sectional area error values smaller than 4%, with other 2D methods exceeding 10%. For volume estimation, photo-reconstruction proved similar to LiDAR data, but 2D methods generated large negative volume error (EV) values (<–13% for laser profi lemeter and pole). We show that the proposed error expressions derived from the model are in line with the reach-scale fi eld results. A measurement distance factor (MDF) is defi ned that represents the ratio between cross-section distance and the gully length, and thus refl ects relative survey effort. We calculate the required MDF for specifi ed values of EV, illustrating how MDF decreases with increasing gully length and sinuosity.

Comparing the accuracy of several field methods for measuring gully erosion

The quality of digital elevation models (DEMs), as well as their spatial resolution, are important issues in geomorphic studies. However, their influence on landslide susceptibility mapping (LSM) remains poorly constrained. This work determined the scale dependency of DEM-derived geomorphometric factors in LSM using a 5 m LiDAR DEM, LiDAR resampled 30 m DEM, and a 30 m ASTER DEM. To verify the validity of our approach, we first compiled an inventory map comprising of 267 landslides for Sihjhong watershed, Taiwan, from 2004 to 2014. Twelve landslide causative factors were then generated from the DEMs and ancillary data. Afterward, popular statistical and machine learning techniques, namely, logistic regression (LR), random forest (RF), and support vector machine (SVM) were implemented to produce the LSM. The accuracies of models were evaluated by overall accuracy, kappa index and the receiver operating characteristic curve indicators. The highest accuracy was attained from the resampled 30 m LiDAR DEM derivatives, indicating a fine-resolution topographic data does not necessarily achieve the best performance. Additionally, RF attained superior performance between the three presented models. Our findings could contribute to opt for an appropriate DEM resolution for mapping landslide hazard in vulnerable areas.

/pdf/evaluating-scale-effects-of-topographic-variables-in-4llf4gtngf.pdf

Evaluating scale effects of topographic variables in landslide susceptibility models using GIS-based machine learning techniques.

Three-dimensional digital technology is important in the maintenance and monitoring of cultural heritage sites. This study focuses on using a combination of terrestrial laser scanning and unmanned aerial vehicle (UAV) photogrammetry to establish a three-dimensional model and the associated digital documentation of the Magoksa Temple, Republic of Korea. Herein, terrestrial laser scanning and UAV photogrammetry was used to acquire the perpendicular geometry of the buildings and sites, where UAV photogrammetry yielded higher planar data acquisition rate in upper zones, such as the roof of a building, than terrestrial laser scanning. On comparing the two technologies’ accuracy based on their ground control points, laser scanning was observed to provide higher positional accuracy than photogrammetry. The overall discrepancy between the two technologies was found to be sufficient for the generation of convergent data. Thus, the terrestrial laser scanning and UAV photogrammetry data were aligned and merged post conversion into compatible extensions. A three-dimensional (3D) model, with planar and perpendicular geometries, based on the hybrid data-point cloud was developed. This study demonstrates the potential for using the integration of terrestrial laser scanning and UAV photogrammetry in 3D digital documentation and spatial analysis of cultural heritage sites.

Three-Dimensional Digital Documentation of Cultural Heritage Site Based on the Convergence of Terrestrial Laser Scanning and Unmanned Aerial Vehicle Photogrammetry

The Chinese Loess Plateau suffers from serious gully erosion induced by natural and human causes. Gully-affected areas detection is the basic work in this region for gully erosion assessment and monitoring. For the first time, an unmanned aerial vehicle (UAV) was applied to extract gully features in this region. Two typical catchments in Changwu and Ansai were selected to represent loess tableland and loess hilly regions, respectively. A high-powered quadrocopter (md4-1000) equipped with a non-metric camera was used for image acquisition. InPho and MapMatrix were applied for semi-automatic workflow including aerial triangulation and model generation. Based on the stereo-imaging and the ground control points, the highly detailed digital elevation models (DEMs) and ortho-mosaics were generated. Subsequently, an object-based approach combined with the random forest classifier was designed to detect gully-affected areas. Two experiments were conducted to investigate the influences of segmentation strategy and feature selection. Results showed that vertical and horizontal root-mean-square errors were below 0.5 and 0.2 m, respectively, which were ideal for the Loess Plateau region. The overall extraction accuracy in Changwu and Ansai achieved was 84.62% and 86.46%, respectively, which indicated the potential of the proposed workflow for extracting gully features. This study demonstrated that UAV can bridge the gap between field measurement and satellite-based remote sensing, obtaining a balance in resolution and efficiency for catchment-scale gully erosion research.

/pdf/detection-of-catchment-scale-gully-affected-areas-using-36lomut5pk.pdf

Detection of Catchment-Scale Gully-Affected Areas Using Unmanned Aerial Vehicle (UAV) on the Chinese Loess Plateau

Gully maps are important prerequisites for the study of gully erosion and land degradation. Many digital elevation model (DEM)-based methods have been proposed to enable automated gully mapping. However, the accuracy of a gully map derived from a DEM is inevitably affected by the DEM resolution. This study investigates the effects of DEM resolution on the accuracy of gully maps. A series of DEMs with resolutions of 0.1–10 m is employed to map gully areas. The effects of DEM resolution on the error in the mapped gully area and on the position error are described by regression models. The results from two catchments in hilly areas of the Loess Plateau in China are as follows. DEMs with resolutions of 0.5–2 m are the most suitable for gully mapping. Very high-resolution DEMs increase local position errors and over-predict the extents of gullies, whereas DEMs with coarser resolutions cause the downward migration of mapped gully boundaries, resulting in the under-prediction of gully areas. However, the effects of DEM resolution on gully maps are not constant but vary in space. The spatial disparities of the resolution effects are related to the gully morphology. The resolution effects on the gully maps in V-shaped gullies are stronger than those in U-shaped gullies. The findings of this study can be used to select a suitable DEM resolution for gully mapping in loess hilly areas and contribute to understanding the characterization of gullies by using DEMs.

Effects of DEM resolution on the accuracy of gully maps in loess hilly areas

Terraces are typical artificial landforms on the Loess Plateau, with ecological functions in water and soil conservation, agricultural production, and biodiversity Recording the spatial distribution of terraces is the basis of monitoring their extent and understanding their ecological effects The current terrace extraction method mainly relies on high-resolution imagery, but its accuracy is limited due to vegetation coverage distorting the features of terraces in imagery High-resolution topographic data reflecting the morphology of true terrace surfaces are needed Terraces extraction on the Loess Plateau is challenging because of the complex terrain and diverse vegetation after the implementation of “vegetation recovery” This study presents an automatic method of extracting terraces based on 1 m resolution digital elevation models (DEMs) and 03 m resolution Worldview-3 imagery as auxiliary information used for object-based image analysis (OBIA) A multi-resolution segmentation method was used where slope, positive and negative terrain index (PN), accumulative curvature slope (AC), and slope of slope (SOS) were determined as input layers for image segmentation by correlation analysis and Sheffield entropy method The main classification features based on DEMs were chosen from the terrain features derived from terrain factors and texture features by gray-level co-occurrence matrix (GLCM) analysis; subsequently, these features were determined by the importance analysis on classification and regression tree (CART) analysis Extraction rules based on DEMs were generated from the classification features with a total classification accuracy of 8996% The red band and near-infrared band of images were used to exclude construction land, which is easily confused with small-size terraces As a result, the total classification accuracy was increased to 94% The proposed method ensures comprehensive consideration of terrain, texture, shape, and spectrum characteristics, demonstrating huge potential in hilly-gully loess region with similarly complex terrain and diverse vegetation covers

Extraction of Terraces on the Loess Plateau from High-Resolution DEMs and Imagery Utilizing Object-Based Image Analysis

Classification of land use and land cover from remote sensing images has been widely used in natural resources and urban information management. The variability and complex background of land use in high-resolution imagery poses greater challenges for remote sensing semantic segmentation. To obtain multi-scale semantic information and improve the classification accuracy of land-use types in remote sensing images, the deep learning models have been wildly focused on. Inspired by the idea of the atrous-spatial pyramid pooling (ASPP) framework, an improved deep learning model named RAANet (Residual ASPP with Attention Net) is constructed in this paper, which constructed a new residual ASPP by embedding the attention module and residual structure into the ASPP. There are 5 dilated attention convolution units and a residual unit in its encoder. The former is used to obtain important semantic information at more scales, and residual units are used to reduce the complexity of the network to prevent the disappearance of gradients. In practical applications, according to the characteristics of the data set, the attention unit can select different attention modules such as the convolutional block attention model (CBAM). The experimental results obtained from the land-cover domain adaptive semantic segmentation (LoveDA) and ISPRS Vaihingen datasets showed that this model can enhance the classification accuracy of semantic segmentation compared to the current deep learning models. 

/pdf/raanet-a-residual-aspp-with-attention-framework-for-semantic-2gc9xspt.pdf

Jiaming Na

Papers

Detection of Catchment-Scale Gully-Affected Areas Using Unmanned Aerial Vehicle (UAV) on the Chinese Loess Plateau

Effects of DEM resolution on the accuracy of gully maps in loess hilly areas

Extraction of Terraces on the Loess Plateau from High-Resolution DEMs and Imagery Utilizing Object-Based Image Analysis

RAANet: A Residual ASPP with Attention Framework for Semantic Segmentation of High-Resolution Remote Sensing Images

Gully boundary extraction based on multidirectional hill‐shading from high‐resolution DEMs