An evaluation of Guided Regularized Random Forest for classification and regression tasks in remote sensing

New Earth observation missions and technologies are delivering large amounts of data. Processing this data requires developing and evaluating novel dimensionality reduction approaches to identify the most informative features for classification and regression tasks. Here we present an exhaustive evaluation of Guided Regularized Random Forest (GRRF), a feature selection method based on Random Forest. GRRF does not require fixing a priori the number of features to be selected or setting a threshold of the feature importance. Moreover, the use of regularization ensures that features selected by GRRF are non-redundant and representative. Our experiments based on various kinds of remote sensing images, show that GRRF selected features provides similar results to those obtained when using all the available features. However, the comparison between GRRF and standard random forest features shows substantial differences: in classification, the mean overall accuracy increases by almost 6% and, in regression, the decrease in RMSE almost reaches 2%. These results demonstrate the potential of GRRF for remote sensing image classification and regression. Especially in the context of increasingly large geodatabases that challenge the application of traditional methods.     

The bivalve Lembulus bicuspidatus may enhance denitrification in shelf sediment at the Angola–Benguela Frontal Zone

We collected living individuals of the bivalve Lembulus bicuspidatus, which shows an unusual preference for the oxygen-deficient habitat found at the Angola–Benguela Frontal Zone of the southeastern Atlantic. With a series of incubation experiments with ¹⁵N-labelled nitrate as a tracer in combination with membrane-inlet mass spectrometry, we studied the potential contribution of L. bicuspidatus to nitrate reduction in the upper sediment layer. Our preliminary results suggest that L. bicuspidatus enhances nitrate reduction if the oxygen concentration is sufficiently low. The Lembulus-mediated nitrate reduction rate is then similar to the rate of microbial nitrate reduction in the surrounding sediment.     

1970-2016年青藏高原岗扎日冰川变化与物质平衡遥感监测研究

可可西里处于青藏高原腹地,是青藏高原自然环境的交接与过渡地带。近年来该区域冰川物质平衡可能有从西向东由正转负的趋势,但是其过渡地带岗扎日地区冰川状态未知。本研究利用地形图、SRTM、ASTER和Landsat等资料分析了岗扎日地区冰川面积变化和物质平衡变化,并对可可西里地区冰川变化空间规律进行了探讨,结果表明：①1970-2016年岗扎日冰川总面积年均缩小率为0.08±0.02%。2006年后冰川退缩趋势减缓。②1970-2012年岗扎日冰川平均减薄-8.64±0.30 m,体积减少1.45±0.06 km³,平均物质平衡为-0.21±0.01 m w.e. a^-1。冰川物质平衡趋势由负转正（1970-1999年：-0.34±0.01 m w.e. a^-1;1999-2012：0.16±0.02 w.e. a^-1）。③东南、南、西南朝向作为迎风坡,1970年以来其冰川物质亏损较小,1999-2012年呈现强烈的正平衡。冰川面积变化滞后于物质平衡变化,东朝向和东南朝向冰川面积缩小率最大,主要是因为冰川冰舌较长,末端所处的海拔较低。④气温升高是岗扎日冰川1970-1999年呈现负物质平衡状态的主因,降水增多是1999-2012年正平衡状态的主因。⑤可可西里地区冰川1970s以来面积年均缩小率从西向东不断增大、物质平衡下降,与西风环流和季风环流相关,但局地气候也影响冰川变化和物质平衡。     

波段选择协同表达高光谱异常探测算法

高光谱遥感影像具有光谱分辨率极高的特点,承载了大量可区分不同类型地物的诊断性光谱信息以及区分亚类相似地物之间细微差别的光谱信息,在目标探测领域具有独特的优势。与此同时,高光谱遥感影像也带来了数据维数高、邻近波段之间存在大量冗余信息的问题,高维度的数据结构往往使得高光谱影像异常目标类和背景类之间的可分性降低。为了缓解上述问题,本文提出了一种基于波段选择的协同表达高光谱异常探测算法。首先,使用最优聚类框架对高光谱波段进行选择,获得一组波段子集来表示原有的全部波段,使得高光谱影像异常目标类与背景类之间的可分性增强。然后使用协同表达对影像上的像元进行重建,由于异常目标类和背景类之间的可分性增强,对异常目标像元进行协同表达时将会得到更大的残差,异常目标像元的输出值增大,可以更好地实现异常目标和背景类的分离。本文使用了3组高光谱影像数据进行异常目标探测实验,实验结果表明,该方法与其他现有高光谱异常目标探测算法对比,曲线下面积AUC(Area Under Curve)值更高,可以更好地实现异常目标与背景分离,能够更有效地对高光谱影像进行异常目标探测。     

Muon density measurements with the KASCADE central detector

Frequency distributions of local muon densities in high-energy extensive air showers (EAS) are presented as signature of the primary cosmic ray energy spectrum in the knee region. Together with the gross shower variables like shower core position, angle of incidence, and the shower sizes, the KASCADE experiment is able to measure local muon densities for two different muon energy thresholds. The spectra have been reconstructed for various core distances, as well as for particular subsamples, classified on the basis of the shower size ratio N_μ/N_e. The measured density spectra of the total sample exhibit clear kinks reflecting the knee of the primary energy spectrum. While relatively sharp changes of the slopes are observed in the spectrum of EAS with small values of the shower size ratio, no such feature is detected at EAS of large N_μ/N_e ratio in the energy range of 1–10 PeV. Comparing the spectra for various thresholds and core distances with detailed Monte Carlo simulations the validity of EAS simulations is discussed.     

关于高速滑坡的几个问题

高速滑坡问题是当前工程地质和环境地质学界极为关注的问题。本文对诸如滑体高速滑动所对应的时空界限、滑体滑动过程所产生的热效应以及空气浮托力等高速滑坡产生机制问题进行了初步探讨。     

EUV and coronagraphic observations of coronal mass ejections

The Large Angle Spectrometric Coronagraph (LASCO) and Extreme-ultraviolet Imaging Telescope (EIT) onboard Solar and Heliospheric Observatory (SOHO) provide us with unprecedented multi-wavelength observations helping us to understand different dynamic phenomena on the Sun and in the corona. In this paper we discuss the association between post-eruptive arcades (PEAs) detected by EIT and white-light coronal mass ejections (CMEs) detected by LASCO/C2 telescope.     

The Roche Limit

The role of tides in deforming and possibly disrupting a secondary body orbiting about a primary body has been known for a considerable time. This was first inspired by the observations of ocean tides on Earth and then seen as playing an important role in the formation and evolution of the Earth–Moon system. Finally, in the beginning of the 20th century it was generally thought to have a significant role in the formation of the solar system through the tidal disruption of the Sun. Here, an overview of the historical developments of the ideas concerned with tidal disruption of a secondary body that can lead to mass loss is given. Some discussion of possible extensions to consider more realistic situations where the secondary body may not be moving on a circular orbit and may not rotate so as to maintain the phase-on configuration to the primary body is also given.