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.     

Isotope hydrological study of mean transit time in the granitic Strengbach catchment (Vosges massif,France): application of the FlowPC model with modified input function

Measurements of ¹⁸O concentrations in precipitation, soil solution, spring and runoff are used to determine water transit time in the small granitic Strengbach catchment (0·8 km²; 883–1146 m above sea level) located in the Vosges Mountains of northeastern France. Water transit times were calculated by applying the exponential, exponential piston and dispersion models of the FlowPC program to isotopic input (rainfall) and output (spring and stream water) data sets during the period 1989–95. The input function of the model was modified compared with the former version of the model and estimated by a deterministic approach based on a simplified hydrological balance. The fit between observed and calculated output data showed marked improvements compared with results obtained using the initial version of the model. An exponential piston version of the model applied to spring water indicates a 38·5 month mean transit time, which suggests that the volume in the aquifer, expressed in water depth, is 2·4 m. A considerable thickness (>45 m) of fractured bedrock may be involved for such a volume of water to be stored in the aquifer. Copyright © 2005 John Wiley & Sons, Ltd.     

Transient‐state groundwater flow in various geometries with wells at arbitrary positions: analytical methods

We have developed a method for analytically solving the porous medium flow equation in many different geometries for horizontal (two‐dimensional), homogeneous and isotropic aquifers containing impermeable boundaries and any number of pumping or injection wells located at arbitrary positions within the system. Solutions and results are presented for rectangular and circular aquifers but the method presented here is easily extendible to many geometries. Results are also presented for systems where constant head boundary conditions can be emulated internal to the aquifer boundary. Recommendations for extensions of the present work are briefly discussed. Copyright © 2003 John Wiley & Sons, Ltd.     

利用Lageos 1 SLR资料解算1990-2001年的地球定向参数

利用 12年的Lageos 1卫星激光测距资料 (1990 - 2 0 0 1)解算得到了地球定向参数 (EOP) ,将该序列的结果与同期的EOP(IERS)C0 4进行比较 ,其外符精度为 :极移XP— 0 .4 0mas,YP— 0 .4 2mas ,日长变化Dr— 0 .0 35ms。     

青藏块体东北缘地壳水平运动状态

应用青藏块体东北缘1999～2003年多期GPS观测资料，计算了不同时段GPS点水平运动速率。通过分析发现：甘青块体可分为东部块体和西部块体，东、西部块体的运动状态存在明显的差异；受2001年11月14日昆仑Ms8．1地震的影响，震后地壳运动状态发生了明显的改变。     

Numerical prediction of blast‐induced stress wave from large‐scale underground explosion

This paper presents a numerical model for predicting the dynamic response of rock mass subjected to large‐scale underground explosion. The model is calibrated against data obtained from large‐scale field tests. The Hugoniot equation of state for rock mass is adopted to calculate the pressure as a function of mass density. A piecewise linear Drucker–Prager strength criterion including the strain rate effect is employed to model the rock mass behaviour subjected to blast loading. A double scalar damage model accounting for both the compression and tension damage is introduced to simulate the damage zone around the charge chamber caused by blast loading. The model is incorporated into Autodyn3D through its user subroutines. The numerical model is then used to predict the dynamic response of rock mass, in terms of the peak particle velocity (PPV) and peak particle acceleration (PPA) attenuation laws, the damage zone, the particle velocity time histories and their frequency contents for large‐scale underground explosion tests. The computed results are found in good agreement with the field measured data; hence, the proposed model is proven to be adequate for simulating the dynamic response of rock mass subjected to large‐scale underground explosion. Extended numerical analyses indicate that, apart from the charge loading density, the stress wave intensity is also affected, but to a lesser extent, by the charge weight and the charge chamber geometry for large‐scale underground explosions. Copyright © 2004 John Wiley & Sons, Ltd.     

APPLICATION OF GEOGRAPHICAL PARAMETER DATABASE TO ESTABLISHMENT OF UNIT POPULATION DATABASE

Now GIS is turning into a good tool in handling geographical, economical, and population data, so we can obtain more and more information from these data. On the other hand, in some cases, for a calamity, such as hurricane, earthquake, flood, drought etc., or a decision-making, such as setting up a broadcasting transmitter, building a chemical plant etc., we have to evaluate the total population in the region influenced by a calamity or a project. In this paper, a method is put forward to evaluate the population in such special region. Through exploring the correlation of geographical parameters and the distribution of people in the same region by means of quantitative analysis and qualitative analysis, unit population database (1km× 1km) is established. In this way, estimating the number of people in a special region is capable by adding up the population in every grid involved in this region boundary. The geographical parameters are obtained from topographic database and DEM database on the scale of     

Reliability assessment of serviceability performance of braced retaining walls using a neural network approach

In urban environments, one major concern with deep excavations in soft clay is the potentially large ground deformations in and around the excavation. Excessive movements can damage adjacent buildings and utilities. There are many uncertainties associated with the calculation of the ultimate or serviceability performance of a braced excavation system. These include the variabilities of the loadings, geotechnical soil properties, and engineering and geometrical properties of the wall. A risk‐based approach to serviceability performance failure is necessary to incorporate systematically the uncertainties associated with the various design parameters. This paper demonstrates the use of an integrated neural network–reliability method to assess the risk of serviceability failure through the calculation of the reliability index. By first performing a series of parametric studies using the finite element method and then approximating the non‐linear limit state surface (the boundary separating the safe and ‘failure’ domains) through a neural network model, the reliability index can be determined with the aid of a spreadsheet. Two illustrative examples are presented to show how the serviceability performance for braced excavation problems can be assessed using the reliability index. Copyright © 2005 John Wiley & Sons, Ltd.     

Possible chemical structures and biological precursors of different vitrinites in coal measure in Northwest China

Chemical structure of Jurassic vitrinites isolated from the coals in basins in NW China have been checked using solid state ¹³C NMR and flash pyrolysis-GC/MS. Study shows some Jurassic collodetrinites are rich in aliphatic products in pyrolysates, consisting with the high amount of methylene carbon in ¹³C NMR spectra. In contrast, pyrolysates of Jurassic collotelinites are rich in phenols and alkylbenzenes. Also one Pennsylvanian and one Permian vitrinite selected from the Ordos basin, NW China have been checked for comparison. The proportion of aliphatics is low in pyrolysates, and aliphatic carbon peak in ¹³C NMR spectrum of Permian vitrinite is mostly composed of gas-prone carbons compared with collodetrinites in those Jurassic basins. But both pyrolysis and ¹³C NMR data shows the Pennsylvanian vitrinite is not only gas-prone but also oil-prone. Relatively high proportion of long chain aliphatic structure of some Jurassic vitrinite in Junggar, Turpan-Hami basins may be due to the contribution of liptodetrinites, which may be included during the formation of vitrinites. And it seems that suberinite is the most possible precursor of long chain aliphatics in the structure of Jurassic collodetrinite.