Search

This study aims to create a terrain classification of Japan that allows both geomorphological and geoengineering classifications coexist without large contradictions and to distinguish landform elements even in urban plains which include noise associated with digital elevation models (DEMs). Because Japan is susceptible to natural disasters, we designed the classification to reflect the ground vulnerability of both alluvial plains and mountains through the application of terrain classification data to landslide susceptibility and seismic zoning. We updated an existing DEM-based terrain classification method for application in the high-resolution 30 m DEM. We used topographic measurements that do not amplify manmade unevenness or noise, which are usually the main problems associated with the use of high-resolution DEMs with high vertical accuracies. We selected the height above the nearest drainage (HAND), slope gradient, surface texture, and local convexity as geometric signatures, which were devised so as not to detect noise. Segment polygon data of terrain units were derived from the raster data of slope and HAND. The polygon data were classified into 40 clusters using the attributes of slope, HAND, and surface texture; then they were grouped into 16 legends following comparisons with the existing geological and geomorphological maps and supplementary reclassification by HAND and local convexity. The derived terrain classification, except for manmade cuts and fills, showed similarities with the existing expert-driven maps and some association with areas where shallow landslides or floods frequently occur. Based on a trial in California using a 30 m DEM, we concluded that the proposed method can be adopted in other regions outside of Japan.

During the occurrence of floods, the government generally tries to determine the extent of the continuously changing inundation (water) area as a disaster response for planning evacuations and designing drainage systems. One of the most common methods to identify this area during the daytime involves the interpretation of aerial optical images and videos that are obtained using an airplane, helicopter, or satellite. However, these aerial approach methods have not been established convincingly for their use during the night time. Therefore, to identify the water area at night, two hand-held optical sensors were used in this study, namely: an ultrasensitive camera (Sony SL camera α7S), and a thermal infrared camera (Optris thermography PI640). The night-time aerial performance was evaluated using a helicopter, along with a supplementary night-time examination on the ground. Herein, we targeted river water areas during normal conditions. As a result, it was determined that the water area could be identified clearly with an ultrasensitive camera that has an ISO sensitivity between 51200 and 102400. The thermal infrared camera was able to identify the water area to have a higher temperature than that of the surrounding area. However, in some cases, the water area boundaries were not clear. Thus, as the thermal infrared camera was susceptible to varying conditions, supplemental use of an ultrasensitive camera was determined to be effective for the identification of inundation (water) areas at night. The significance of this study is established by the fact that the studied sensors can be easily transported in a helicopter or mounted on a UAV for night-time applications.

Global terrain classification data have been used for various issues related to topography such as the estimation of soil types and of ground vulnerability to earthquakes and the creation of seismic hazard maps. However, due to the resolution of digital elevation models (DEMs), the terrain classification data from previous studies could not discriminate small landforms such as plains at the bottom of narrow valleys and small rises in plains. Owing to the greater regional variation of small landforms, there is trade-off between DEMs of higher resolution and the creation of global geomorphological legends. To address this problem, we first merged regions with similar topographic characteristics using slope gradients and HAND (height above the nearest drainage) calculated by the 90-m-spatial-resolution DEMs interpolated from the multi-error-removed improved-terrain DEM (MERIT DEM), and united the polygons with the unit catchments of the MERIT-Basins dataset, so that the polygons contain calculated terrain measurements (slope gradient, HAND, surface texture, local convexity, sinks) and noise types as attributes, as well as the ID number of the unit catchment. In addition, we performed k-means clustering on the dataset using slope gradient, HAND, and surface texture, which can be combined with the dataset as a simple terrain classification. The clustering results were prepared in 15 and 40 global uniform clusters and 15 and 40 clusters for each basin to understand the global appearance of the terrain and provide zoning data for regional problem-solving. The 15 clusters were prepared to observe the outline of the terrain without any processing, whereas the 40 clusters were prepared to group and reclassify the polygons to create zoning data for each region. This dataset showed improvements in terms of capturing the small rises in plains compared to the authors' previous global terrain classification data. This dataset can be used as a proxy and is expected to contribute to modeling and estimation in various applications that are known to be related to topography. The datasets of this article are available at [https://gisstar.gsi.go.jp/terrain2021/].

The time-averaged shear-wave velocity in the upper 30 m depth from the ground surface, or VS30, is often used as a predictor to describe local site effects in ground-motion models. Although VS30 is typically determined from in situ measurements, it is not always feasible to obtain such measurements due to project restrictions or site accessibility. This motivates the development and use of proxy-based VS30 predictions that leverage more readily available secondary information such as surface geology, topographic slope, or geomorphic terrain classes to estimate the mean VS30 and associated uncertainty. Traditionally, empirical distributions of VS30 have been observed to have long right tails, leading to high levels of associated uncertainty. In this study, we present a physical framework that is grounded in fundamental principles of geostatistics and probability to explain the uncertainty and skewness associated with VS30 measurements. Specifically, by invoking Lyapunov’s central limit theorem, we hypothesize that the distribution of VS30 can be theoretically approximated by a reciprocal–normal distribution. We show that a non-normal and skewed distribution of VS30 is to be expected and is not a sign of measurement error or sampling bias, although sampling bias can exaggerate such skewness. Our framework also enables us to propose the mode as a characteristic value of VS30 measurements, as opposed to the mean or median, which can overestimate the most probable value.

An amendment to this paper has been published and can be accessed via the original article.

In this presentation, we report science classes in which the purpose was to learn the history of local geology, and a workshop to learn the relationship between landforms and natural hazards using Minecraft. Minecraft (Mojang/Microsoft) is a sandbox computer game for exploration and crafting in 3-D virtual worlds. It is very popular among the young generation (100 million users in the world), and by using the game it is easy to construct virtual worlds and exploration mechanisms. The science classes were conducted twice for students aged 12 to 13 in a junior high school in Miyagi Prefecture, Japan. Using Minecraft, we have constructed a virtual world tailored to their school, including the school buildings and paleoenvironments. In the game, students travel around the school buildings to learn and to solve basic knowledge questions based on references from their school science textbooks, then they go to the underground strata and into past worlds to learn and to solve advanced questions which refer to papers on regional geology. A questionnaire which was given to over 150 students after the first class showed that the students enjoyed the class and obtained a general understanding of geological knowledge. The second class was based on a reviced game after referring to the results of the questionnaire. In the workshop, we used a 3-D topographic model of Japanese flood plains and surrounding terraces and mountains. This example was conducted for 15 to 18-year-old students as a workshop with a small number of students, less than 10. At first, we explained to the students how landforms are associated with natural hazards such as flooding and earthquake shaking, and explained how to find and view thematic maps like hazard maps that could be observed as interactive web maps published by Japanese public agencies and institutes. Next, the students were asked where they wanted to build a house on the virtual terrain. Through their constructions, we considered the balance between playing and learning. This study was supported by JSPS KAKENHI Grant Number JP18K18548.

Geographic information on soil organic carbon (SOC) is important for climate change research and understanding the global carbon cycle. At present, this information is limited for forest soils in mountainous regions that have complex topographies and are affected by tephra deposition. This study predicted the SOC concentration (OC), bulk density (BD), apparent rock coarse fraction (CF), and SOC stock (OCS) at 0–5, 5–15, and 15–30 cm in the soil horizon for the forested areas (ca. 230 000 km2) of Japan. This was carried out with a 10 m grid resolution using a digital soil mapping (DSM) approach. To determine the appropriate spatial prediction model, we evaluated the accuracy of a two dimensional approach for ordinary kriging (OK) and regression kriging (RK), and a three dimensional approach for random forest model (RF). The RF was based on topography, climate and vegetation factors (RFenv), distance and direction from 152 volcanos (RFenv+vol), distance from survey points (RFenv+sp); these factors were also combined (RFall) as explanatory attributes. The results demonstrated an improvement in the average root mean square errors (RMSEs) of RFenv+vol, RFenv+sp, and RFall by approximately 12%, 14%, 5%, and 21% for OC, BD, CF, and OCS using 10-fold cross-validation on a site-by-site basis, respectively, compared with OK. Based on the relatively small difference in improvement among RFenv+vol, RFenv+sp, and RFall (

Polygon-based terrain classification data were created globally using 280 m digital elevation models (DEMs) interpolated from the multi-error-removed improved-terrain DEM (MERIT DEM). First, area segmentation was performed globally with the logarithmic value of slope gradient and the local convexity calculated from the DEM. Next, by adding surface texture, k-means clustering was performed globally and the polygons were grouped into 40 clusters. Then, we tried to reclassify these 40 clusters into geomorphologic terrain groups. In this study, we attempted reclassification and grouping using local information from Japan as a test case. The 40 clusters were compared with Japanese geological and geomorphological data and were then reclassified into 12 groups that had different geomorphological and geological characteristics. In addition, large shape landforms, mountains, and hills were subdivided by using the combined texture. Finally, 15 groups were created as terrain groups. Cross tabulations were performed with geological or lithological maps of California and Australia in order to investigate if the Japanese grouping of the clusters was also meaningful for other regions. The classification is improved from previous studies that used 1-km DEMs, especially for the representation of terrace shapes and landform elements smaller than 1 km. The results were generally suitable for distinguishing bedrock mountains, hills, large highland slopes, intermediate landforms (plateaus, terraces, large lowland slopes), and plains. On the other hand, the cross tabulations indicate that in the case of gentler landforms under different geologic provinces/climates, similar topographies may not always indicate similar formative mechanisms and lithology. This may be solved by locally replacing the legend; however, care is necessary for mixed areas where both depositional and erosional gentle plains exist. Moreover, the limit of the description of geometric signatures still appears in failure to detect narrow valley bottom plains, metropolitan areas, and slight rises in gentle plains. Therefore, both global and local perspectives regarding geologic province and climate are necessary for better grouping of the clusters, and additional parameters or higher resolution DEMs are necessary. Successful classification of terrain types of geomorphology may lead to a better understanding of terrain susceptibility to natural hazards and land development.

The purpose of this research is to give children a geographical viewpoint, and to encourage an interest in, and awareness of, landforms and geology. We created a system based on an exploration type computer game and verified the educational effects. Moreover, we aim to reach not only the virtual aspect but we also have a goal of creating interest in the actual field. As a secondary effect, by using a computer game that attracts children’s interest, we aim to make the experience of solving issues subjective and active even if the player is a passive child, a child with little inquiry, or a child who is not adept at self-assertion. With this new approach, we also hope to interact with young generations who usually do not interact with researchers.Many thematic maps of geography and geology are already published on the Web. They are popular among those who need to collect and view the information for some reason or with those who are interested in observing topographic maps and are interested in geology. However, in particular, the approach to children who do not have such motivation needs one more step: a mechanism to induce an inquiring mind, and a mechanism that leads to finding the information and having interest in the real field.The platform of this research is Minecraft Education Edition (Mojang/Microsoft). Minecraft is very popular game software which has exceeded one hundred million users worldwide in recent years, and in Japan there are many elementary and junior high school student enthusiasts of Minecraft. In the game a user explores a virtual world made of cubic blocks. The blocks imitate vegetation, rock formations, and other items, and can create various puzzles. In recent years, the release of the Education Edition assumes use in classrooms.In this research, we have constructed a virtual world tailored to a specific junior high school which teaches science classes to first grade students. First, we re-created the actual school buildings and also included the underground geologic strata based on data from boring. In addition, we created a mechanism to expand children’s imagination and knowledge about past environments which can be understood from the geological strata. We also provided checkpoints and gave challenges regarding knowledge about the formation of the land. Together with this modern world, we created ancient virtual worlds so users may understand the geological history around the school’s location.Through the experience of this research, we were able to confirm the mechanisms for promoting motivation in children and aiding their understanding of science. It can be applied to systems other than Minecraft, and it can contribute to educational support in a wide variety of fields.

We developed classification schemes for topography in the artificially modified alluvial plains. In order to determine the ground conditions in such areas, it is important to define the distributions of low rises such as natural levees and gentle fans. We searched for geometric signatures which did not over-emphasize the influences of artificial unevenness and were thus useful for revealing low rises on alluvial plains. Based on our findings using the highly accurate DEM for the Nobi plain, we concluded that Height Above Nearest Drainage (HAND) was most useful for detecting natural levees and gentle fans. Then we used HAND and slope gradient to produce terrain classification data for the Central Plain of Thailand using MERIT DEM in consideration to future development of global data.

In the publication of this article (Iwahashi et al. 2018), there was an error in Table 3 in terrain group 10, 11 and 12.

Nishinoshima volcano in Ogasawara Islands has erupted since November, 2013. This volcanic eruption formed and enlarged a new island, and fused the new island with the old Nishinoshima Island. We performed automated aerial photographing using an Unmanned Aerial Vehicle (UAV) over the joined Nishinoshima Island on March 22 and July 4, 2014. We produced ortho-mosaic photos and digital elevation model (DEM) data by new photogrammetry software with computer vision technique, i.e. Structure from Motion (SfM) for estimating the photographic position of the camera and Multi-view Stereo (MVS) for generating the 3-D model. We also estimated the area and volume of the new island via analysis of ortho-mosaic photo and DEM data. Transition of volume estimated from the UAV photographing and other photographing shows the volcanic activity still keeps from initial level. The ortho-mosaic photos and DEM data were utilized to create an aerial photo interpretation map and a 3-D map. These operations revealed new knowledge and problems to be solved on the photographing and analysis using UAV and new techniques as this was first case in some respects.

The spatial distribution of shallow landslides in the Izumozaki area, Niigata, caused by the heavy rainfalls in August 1961 and July 2004 was investigated using high-resolution ortho-photoimagery and a 2-m DEM. We found that the number of the August 1961 landslides is more than twice the July 2004 landslides. More than half of the July 2004 landslides (about 70% as the number ratio and 54% as the area ratio) were primary landslides. These primary landslides seem to occur randomly regardless of the geological structure of the stratum. The large landslides which occurred in July 2004 were often expanded landslides immediately higher than the August 1961 landslides. These expanded landslides often occurred on daylighting dip slopes. Among the July 2004 landslides, the ratio of landslides along roads is very high compared to 1961 (1961: 4.2%; 2004: 16.4%).

We undertake digital terrain analyses of rainfall- and earthquake-induced landslides in Japan, using high-resolution orthoimagery and Light Detection and Ranging (LiDAR) DEMs. Our aims are twofold: to demonstrate an effective method for dealing with high-resolution DEMs, which are often too detailed for landslide assessments, and to evaluate the topographic differences between rainfall- and earthquake-induced landslides. The study areas include the Izumozaki (1961 and 2004 heavy rainfalls), Niihama (2004 heavy rainfalls), Houfu (2009 heavy rainfalls), and Hanokidachi/Kurikoma-dam regions (the 2008 M 7.2 Iwate–Miyagi Nairiku earthquake). The study areas include 7,106 landslides in these five regions. We use two topographic attributes (the slope gradient and the Laplacian) calculated from DEMs in varying window sizes. The hit rates for statistical prediction of landslide cells through discriminant analyses are calculated using the two topographic attributes as explanatory variables, and the landslide inventory data as the dependent variable. In cases of surface failure, the hit rates are found to diminish when the window size of the topographic attributes is too large or too small, indicating that an optimal scale factor is key in assessing shallow landslides. The representative window sizes are approximately 30 m for shallow landslides; the optimal window size may be directly related to the average size of landslides in each region. We also find a stark contrast between rainfall- and earthquake-induced landslides. Rainfall-induced landslides are always most common at a slope gradient of 30°, but the frequency of earthquake-induced landslides increases exponentially with slope gradient. We find that the Laplacian, i.e., the attributes of surface convexity and concavity, and the slope gradient are both important factors for rainfall-induced landslides, whereas earthquake-induced landslides are influenced mainly by slope steepness.

The average shear wave velocity for the top 30 m (Vs30) is one of the simple predictors of earthquake ground motion amplification. We performed a multiple linear regression analysis of the logarithm of observed Vs30 at 1646 locations in Japan with three topographic attributes (slope gradient, surface texture, and the logarithm of elevation) calculated from a 50-m DEM. A Vs30 map was created using the partial regression coefficients. The determination coefficient was 0.318. The development of the estimation method of Vs30 from a DEM will be useful for earthquake vulnerability assessment.

新潟県出雲崎地域の1961年8月豪雨と2004年7月豪雨による豪雨の崩壊地の空間分布について, 高解像度のオルソ画像と2mDEMを用いて再検討した。その結果, 次の事が明らかになった。1961年豪雨の崩壊箇所の方が倍以上多かった。2004年崩壊は道路法面周辺の崩壊の比率が高かった (1961年豪雨4. 2%に対して16. 4%) 。2004年崩壊地の半分以上 (箇所数で70%, 面積率で54%) は, 同じ谷壁斜面に1961年崩壊がない新規の崩壊であった。新規の崩壊は, 同じ岩相内で地形条件が同じなら, 斜面内部の受け盤・流れ盤等の構造に関わらず, ランダムに発生する傾向があると推測された。1961年の崩壊地が拡大する形で崩壊した場合, 大部分は斜面上方の崩壊であった。西山層を除き, 拡大型の崩壊には柾目盤が目立つ傾向があった。

2004年 (平成16年) 7月の新潟豪雨および10月の新潟県中越地震によって, 傾斜5度以上の丘陵地・山地に相当する中新世～更新世堆積岩類の斜面で起きた崩壊について, 25mグリッドのレベルで, 傾斜, 雨量, 最大加速度, 地質, 曲率, 地質構造, 斜面方位のGISデータを用いて判別分析を行った。判別分析によって, 各パラメータの崩壊に対する寄与を評価した。その結果, 傾斜は重要なファクターであるが, 豪雨による斜面崩壊では, 傾斜以上に日雨量の寄与が大きいこと, 西山階泥岩優勢タービダイト層と魚沼層では豪雨と地震に於ける崩れやすさが異なっていたこと, 曲率の寄与は豪雨で大きいこと, 地質と地質構造の寄与は地震で起きた大崩壊の場合大きいこと, 斜面方位の寄与には地域差があることなどが分かった。

2004年7月13日に新潟県中越地域では強い前線の活動による集中豪雨が発生した。その雨量は24時間で400mmに達し, 堤防の破堤による河川氾濫をもたらし, アジア航測による空中写真判読では3, 359箇所以上の崩壊も発生させた。写真判読と現地調査により二つのタイプの崩壊が確認された。つまり, ひとつは表層崩壊であり, 2つは深層崩壊で, 後者は特に泥流をともなうものであった。本研究ノートでは, 出雲崎地域に焦点をあてて議論する。なぜなら, この地域は2004年7月豪雨による代表的な崩壊発生地域のひとつであることと, 1961年8月と1976年, 1978年にも豪雨崩壊が発生したためである。そこで, 筆者らは, 出雲崎地域の空中写真の1962年撮影 (1961年豪雨の1年後) , 1982年撮影 (1976-1978年豪雨崩壊を含む) , および2004年撮影 (7月豪雨直後) のそれぞれを判読し, GISを使用して, それぞれの年代の崩壊の分布を比較し, その分布と地形や地質との関連を検討した。

An earthquake-fault-like structure was reported in the north of the Aratozawa dam after the Iwate-Miyagi Nairiku Earthquake in 2008. Horizontal displacements around the structure were measured by photogrammetry. We found 2 rigid blocks on both side of the structure. Comparing with the nearest GPS-based control station (permanent GPS receiving station), we judged that relative displacement between the rigid blocks was caused by the movement of the earthquake fault. Displacements are smoothly varying between the structure and the rigid blocks. Therefore we concluded that the structure is based on the movement of the earthquake fault.

On July 13, 2004, heavy rainfalls because of the intensive activities of the rain front occurred in the Mid-Niigata Region, Japan. They were as much as 400 mm in 24 h, bringing about serious flooding by breaking the river banks. The heavy rainfalls also triggered more than 3,500 landslides. Three months later, the southern region of Mid-Niigata was attacked by an earthquake of magnitude 6.8 on the Richter scale on October 23, 2004. The main earthquakes were followed by intensive aftershocks, which continued until December 2004. By these earthquakes, variable landslides of more than 4,400 also occurred in the hilly and mountainous areas. Namely, different triggers brought about the variable landslides in the hilly mountains whose features are very similar in geological and geomorphological points of view. Therefore, these two events are very useful for clearing the difference in features of the landslides between the two. We have been researching on both landslides in the field just after both occurred and later analyzing air photographs using the geographic information system (GIS). In this paper, we describe the comparison in the distribution features using GIS analytical data between the heavy rainfall-induced and the intensive earthquake-induced landslides.

An iterative procedure that implements the classification of continuous topography as a problem in digital image-processing automatically divides an area into categories of surface form; three taxonomic criteria–slope gradient, local convexity, and surface texture–are calculated from a square-grid digital elevation model (DEM). The sequence of programmed operations combines twofold-partitioned maps of the three variables converted to greyscale images, using the mean of each variable as the dividing threshold. To subdivide increasingly subtle topography, grid cells sloping at less than mean gradient of the input DEM are classified by designating mean values of successively lower-sloping subsets of the study area (nested means) as taxonomic thresholds, thereby increasing the number of output categories from the minimum 8 to 12 or 16. Program output is exemplified by 16 topographic types for the world at 1-km spatial resolution (SRTM30 data), the Japanese Islands at 270 m, and part of Hokkaido at 55 m. Because the procedure is unsupervised and reflects frequency distributions of the input variables rather than pre-set criteria, the resulting classes are undefined and must be calibrated empirically by subsequent analysis. Maps of the example classifications reflect physiographic regions, geological structure, and landform as well as slope materials and processes; fine-textured terrain categories tend to correlate with erosional topography or older surfaces, coarse-textured classes with areas of little dissection. In Japan the resulting classes approximate landform types mapped from airphoto analysis, while in the Americas they create map patterns resembling Hammond's terrain types or surface-form classes; SRTM30 output for the United States compares favorably with Fenneman's physical divisions. Experiments are suggested for further developing the method; the Arc/Info AML and the map of terrain classes for the world are available as online downloads.

The 2005 northern Pakistan earthquake (magnitude 7.6) of 8 October 2005 occurred in the northwestern part of the Himalayas. We interpreted landslides triggered by the earthquake using black-and-white 2.5-m-resolution System Pour l’Observation de la Terre 5 (SPOT 5) stereo images. As a result, the counts of 2,424 landslides were identified in the study area of 55 by 51 km. About 79% or 1,925 of the landslides were small (less than 0.5 ha in area), whereas 207 of the landslides (about 9%) were large (1 ha and more in area). Judging from our field survey, most of the small landslides are shallow rock falls and slides. However, the resolution and whitish image in the photos prevented interpreting the movement type and geomorphologic features of the landslide sites in detail. It is known that this earthquake took place along preexisting active reverse faults. The landslide distribution was mapped and superimposed on the crustal deformation detected by the environmental satellite/synthetic aperture radar (SAR) data, active faults map, geological map, and shuttle radar topography mission data. The landslide distribution showed the following characteristics: (1) Most of the landslides occurred on the hanging-wall side of the Balakot–Garhi fault; (2) greater than one third of the landslides occurred within 1 km from the active fault; (3) the greatest number of landslides (1,147 counts), landslide density (3.2 counts/km2), and landslide area ratio (2.3 ha/km2) was found within Miocene sandstone and siltstone, Precambrian schist and quartzite, and Eocene and Paleocene limestone and shale, respectively; (4) there was a slight trend that large landslides occurred on vertically convex slopes rather than on concave slopes; furthermore, large landslides occurred on steeper (30° and more) slopes than on gentler slopes; (5) many large landslides occurred on slopes facing S and SW directions, which is consistent with SAR-detected horizontal dominant direction of crustal deformation on the hanging wall.

Landslide mass size frequency distributions and mean slope-angle frequency distributions were calculated for slump, slide, and creep type landslides in the Higashikubiki area. Mean slope-angle frequency distributions closely approximated Weibull distributions. Size frequency distributions show power-law dependencies. Both can be explained by modeling landslides as linked uniform blocks in tensile force. Power coefficients for size frequency distributions were 2.01–2.32 (approximation to power functions) or 2.10–2.24 (approximation to Pareto distributions).

Various numerical analyses using 25 m-grid DEMs were compared with rasterized slope movement distribution maps and geological maps of the Higashikubiki area in Niigata Prefecture, Japan. Slope movement distribution maps included five classes: “stable slope”, “landslide mass”, “landslide scarp”, “collapse scarp” and “crack”. Their statistical characteristics were clarified by numerical landform analyses for each lithology. Frequency distributions of slope angle were nearer to a normalized distribution for landslide masses than for stable slopes. From the difference of mean slope degrees of landslide masses and that of stable slopes, it was estimated that the smallest slope angle at which landslides occur is approximately 10° for sandy formations. Moreover, mean values of slope angle in the total study area and total landslide masses were both approximately 14°. For sedimentary rocks, graphs of mean slope for each elevation zone of landslide mass and stable slope were parallel, and slope angle did not increase above approximately 100–160 m. For landslide masses and collapse scarps, the relationship between slope and convexity or roughness showed a regular relation closely resembling linear or quadratic function.

Detailed landform classification is important if effective measures against landslides are to be taken. Conventional techniques can only measure the detailed terrain in vegetated areas with difficulty. Airborne light detection and ranging (LIDAR) is a promising tool to precisely and directly measure a digital elevation model (DEM). Using a two-meter-grid DEM we attempted to understand landslide characteristics, namely, we produced manual and automated landform classification maps in Tomari-no-tai area in Shirakami Mountains, Japan. In advance, 1 : 2500-scale two-meter-interval contour map was newly printed using the LIDAR-DEM. It was found that valleys and other geomorphological features could be seen in better detail in the airborne LIDAR contour map than in the existing photogram-metric contour map. The map and 1 : 8000-scale aerial photographs were interpreted, and manual landform classification map was produced. As a result, 17 classifications were identified in the map.