The use of logistic model tree (LMT) for pixel- and object-based classifications using high-resolution WorldView-2 imagery

Logistic model tree (LMT), a new method integrating standard decision tree (DT) induction and linear logistic regression algorithm in a single tree, have been recently proposed as an alternative to DT-based learning algorithms. In this study, the LMT was applied in the context of pixel- and object-based classifications using high-resolution WorldView-2 imagery, and its performance was compared with C4.5, random forest and Adaboost. Results of the study showed that the LMT generally produced more accurate classification results than the other methods for both pixel- and object-based classifications. The improvement in classification accuracy reached to 3% in pixel-based and 5% in object-based classifications. It was also estimated that the LMT algorithm produced the most accurate results considering the allocation and overall disagreement errors. Based on the Wilcoxon’s Signed-Ranks tests, the performance differences between the LMT and the other methods were statistically significant for both pixel- and object-based image classifications.     

An experimental comparison between KELM and CART for crop classification using Landsat-8 OLI data

The operational land imager (OLI) is the latest instrument in the Landsat series of satellite imagery, which officially began normal operations on 30 May 2013. The OLI includes two bands that are not on the thematic mapper series of sensors aboard Landsat-5 and 7; a cirrus band and a coastal/aerosol band. This paper compares the classification and regression tree and the kernel-based extreme learning machine (KELM) for mapping crops in Hokkaido, Japan, using OLI data, except the cirrus band and the pan band. The OLI data acquired on 8 July 2013 was used for crop classification of beans, beets, grassland, maize, potatoes and winter wheat. The KELM algorithm performed better in this study and achieved overall accuracies of 90.1%. According to the Jeffries–Matusita (J–M) distances, the short wavelength infrared band provides the greater contribution (the highest value was observed for band 6 in OLI data).     

A Finite Volume Method with Unstructured Triangular Grids for Numerical Modeling of Tidal Current

The finite volume method （FVM） has many advantages in 2-D shallow water numerical simulation. In this study, the finite volume method is used with unstructured triangular grids to simulate the tidal currents. The Roe scheme is applied in the calculation of the intercell numerical flux, and the MUSCL method is introduced to improve its accuracy. The time integral is a two-step scheme of forecast and revision. For the verification of the present method, the Stoker＇s problem is calculated and the result is compared with the mathematically analytic solutions. The comparison indicates that the method is feasible. A sea area of a port is used as an example to test the method established here. The result shows that the present computational method is satisfactory, and it could be applied to the engineering fields.     

Mean and Temporal Variability in Kuroshio Geostrophic Transport South of Taiwan (1993–2001)

Observations of the Kuroshio south of Taiwan have been carried out on a quarterly basis since late 1992 as part of the basin-wide High Resolution expendable bathythermograph/expendable conductivity-temperature-depth (XBT/XCTD) network. Mean geostrophic transport in the Kuroshio, 0–800 m, from 34 cruises is 22.0 Sv ± 1.5, consistent with previous results from moorings and geostrophic calculations in the upstream Kuroshio region. The mean core of the current has speed about 90 cm s⁻¹ and is located close to Taiwan. At this location the Kuroshio appears to be confined mainly to the upper 700 m, and there is no evident tight recirculation of the current. Eddy variability is substantial, and large eddies can be seen propagating westward for thousands of kilometers in TOPEX/Poseidon altimetric data, impinging on the current and altering its structure and transport. The annual range in transport is about 8 Sv ± 6, with maximum in summer. Interannual variability is about 12 Sv ± 6, with transport maxima in 1995 and 2000 and a minimum in 1997–1998. Interannual variability in the upstream Kuroshio may be uncorrelated with that in the downstream region south of Japan, where the transport is much greater. Our quarterly sampling aliases high frequency variability of the current, and an improved boundary-current observation program would include more frequent transects and occasional deeper measurements. This revised version was published online in August 2006 with corrections to the Cover Date.     

基于区域分割合并的建筑物半自动提取方法

提出了一种遥感图像半自动提取建筑物的方法。首先通过对建筑物边缘进行折线逼近和感知编组，可以得到其顶部大体结构。然后按照“二叉树”的数据结构，利用“超线”对建筑物影像进行区域分割。最后通过对分割后的区域进行相似性的统计并按照“二叉树”记录的分割顺序进行生长合并，从而得到了屋顶的边缘，实现了对平顶直角建筑物的提取。该方法能够充分利用区域的性质进行合并，稳定精确地提取建筑物。     

利用最小填充树分解方法实现最短路径查询

随着社会的快速发展,道路网的规模越来越大,传统的最短路径算法已不能满足当前的实时要求,本文将基于最小度的树分解查询算法扩展至有向有权图中,提出了效果更好的基于最小填充的树分解最短路径查询算法,并对查询算法求解集合的过程进行了优化,实验结果表明,随着数据规模的增长,算法的时间效率相对于采用二叉堆的Dijkstra算法得到数量级提高。     

Changes in crop type distribution in Zhangye City of the Heihe River Basin,China

With rapid economic development in China, crops have undergone remarkable changes in both their type and spatial pattern. Timely and accurate information of crop type distribution will help government and agricultural producers quickly understand regional agricultural production conditions to better facilitate appropriate adjustments in planting patterns and policies. Another benefit of acquiring such knowledge of crops is that it should enhance regional agricultural competitiveness, optimize resource allocations, and further guarantee national food security. Towards this end, and using the Zhangye City in the Heihe River Basin as a study area, the present research elaborated upon a methodology to classify crop type distribution based on multi-temporal Thematic Mapper and Enhanced Thematic Mapper Plus (TM/ETM+) images. Using this methodology we achieved the spatial distributions of crop types in Zhangye City in 2007 and 2012, and analyzed changes in their distributions over this period. In addition, some landscape indices were calculated to clarify the landscape pattern of crops. The crop conversion potentials in 2017 were modeled using four conversion sub-models of the Multi-Layer Perceptron (MLP) neural network. Generally, the overall accuracy of crop classification in Zhangye was high, at 89.38%. From 2007 to 2012, the cultivated land area in Zhangye increased from 463.81 × 10³ ha to 493.89 × 10³ ha. The sowing area of corn and oilseed rape increased by 39.21 × 10³ ha and 5.99 × 10³ ha, respectively, while for wheat and barley the sowing area decreased by 3.61 × 10³ ha and 9.14 × 10³ ha, respectively. Considering other crop types as a group, their sowing area decreased by only 2.37 × 10³ ha. The increase in corn sowing area mainly came from the conversion of other crops to corn, which accounted for 43.09% of its total sowing area in 2012. Furthermore, corn and oilseed rape showed a tendency of intensive sowing, whereas for wheat and barley the tendency was towards scattered sowing. For the future, corn has high conversion potential in Linze and Gaotai counties of Zhangye, while wheat, barley and oilseed rape have high conversion potentials in Minle and Shandan counties.     

The use of SfM-photogrammetry to quantify and understand gully degradation at the temporal scale of rainfall events: an example from the Ethiopian drylands

Abstract

With the recent technological advances offered by SfM-photogrammetry, we now have the possibility to study gully erosion at very high spatial and temporal scales from multi-temporal DEMs, and thus to enhance our understanding of both gully erosion processes and controls. Here, we examine gully degradation and aggradation at a gully headcut and at four re-incisions along a gully reach in Northern Ethiopia. Environmental controls recorded are topography rainfall, runoff, land use and cover, land management, and soil characteristics. The overall vulnerability of the catchment to erosion is low as calculated from the RUSLE (average 11.83 t ha^?1 y^?1). This reflects the successful land management of the past years. The runoff coefficient was on average 7.3% (maximum 18.2%). Runoff events caused most geomorphic change in the gully, but slumping of the gully bank also occurred on dry days. Most geomorphic change was caused by one major rainfall event of 54.8 mm d^?1, and smaller runoff events caused both degradation and aggradation, often asynchronous between studied sites. Although most research focuses on gully heads alone, re-incisions at lower locations can still cause important gully degradation, which ultimately will reach the gully head and cause instability.     

The influence of topographic variables on treeline trees under different environmental conditions

Aiming to analyze the site conditions of treeline trees at the highest elevations, we investigated 360 km of treelines in the Upper Valtellina, Italian Alps. We analyzed approximately five trees per km and determined the environmental factors limiting treeline elevation by distinguishing between geomorphologic constraints (mean elevation 2355 m), climatic constraints (2530 m), and human impacts (2335 m). Up to 82% of the 1814 analyzed treeline trees were influenced by geomorphological constraints, whereas human impacts influenced only 3% of the trees. Climatic treelines (15% of trees) were most common in the western sector. Moreover we analyzed the frequency distributions of elevation, slope, and aspect. Elevation was the most important variable that was also strongly associated with climatic treelines. The slope variable was more strongly associated with treelines limited by geomorphology than by climate. By reconstructing the altitudinal dynamics at the Mt. Confinale study site, we found rates of an ongoing treeline upward shift of up to 2.6 m/y in the period 2000–2009. Our results indicate that climatic conditions related to the ongoing air temperature rise in this region will likely enhance the treeline shift, especially at high elevations (>2400 m a.s.l.) and on non-extreme slopes (<45°).     

基于完全三叉树的快速推进法地震波走时计算

快速推进法(简称FMM)在地震波走时计算中有着精度高、效率高的特点,但窄带扩展每次都要寻找最小走时。当网格节点较多时,寻找最小走时非常耗时。在保证精度的前提下,为了提高计算效率,笔者对堆排序的排序方式做了改进,将完全三叉树排序方法引入到快速推进法地震波走时计算中。模型试算结果表明,基于完全三叉树快速推进法计算出的地震波走时与用完全二叉树方法的精度一致,且前者比后者效率提高约10%。