Modeling catchment sediment yield: a genetic programming approach

Hydrologic processes are complex, non-linear, and distributed within a watershed both spatially and temporally. One such complex pervasive process is soil erosion. This problem is usually approached directly by considering the sediment yield. Most of the hydrologic models developed and used earlier in sediment yield modeling were lumped and had no provision for including spatial and temporal variability of the terrain and climate attributes. This study investigates the suitability of a recent evolutionary technique, genetic programming (GP), in estimating sediment yield considering various meteorological and geographic features of a basin. The Arno River basin in Italy, which is prone to frequent floods, has been chosen as case study to demonstrate the GP approach. The results of the present study show that GP can efficiently capture the trend of sediment yield, even with a small set of data. The major advantage of the GP analysis is that it generates simple parsimonious expression offering some possible interpretations to the underlying process.     

Temporal dynamics and spatial scales: Modeling deforestation in the southern Yucatán peninsular region

Two spatially explicit econometric land use change models are presented, focusing on tropical deforestation caused by agricultural expansion in the southern Yucatán peninsula, Mexico. The two models developed are both based on conceptually similar theoretical models of farmer behavior. However, there are different empirical specifications of this theoretical model according to the scale of the analysis as well as the availability of temporal data on the observation of deforestation. For both models, the unit of observation for the dependent variable of deforestation is the TM pixel from satellite data. However, the socio-economic explanatory variables are derived from different sources. The first econometric model links the satellite data for the entire study region with aggregate census data at the village level. This model is estimated using a discrete choice logit model over a single time period. The second econometric model uses individual household survey data for a small random sample of the region, linked to satellite data for the plots of each household over multiple time periods. This model is estimated using a dynamic hazard model that estimates the risk of a specific pixel converting from forest to agricultural use. Both estimated models are used to predict deforestation and the results of the two modeling approaches are compared.     

Two-dimensional dam break flooding simulation: a GIS-embedded approach

In the twenty-first century, around 200 notable dam and reservoir failures happened worldwide causing massive fatalities and economic costs. In order to reduce the losses, managers usually define mitigation strategies identifying flooding area due to dam break by using standalone hydrodynamic models and then importing the results within a GIS to perform risk analysis. This two-step procedure is time expensive, error prone due to export/import requirements and not user friendly. For this reason with this work, a new numerical model for the solution of the two-dimensional dam break problem has been implemented in the GRASS GIS with a GIS-embedded approach. The model solves the conservative form of the 2D shallow water equations using a finite volume method; the intercell flux is computed by one-side upwind conservative scheme extended to a two-dimensional problem. The newly developed GIS module, among others outputs, allows to derive maximum intensity maps that can be directly used for risk assessment. Finally, the model has been (1) tested against two standard synthetic problems referenced in literature showing differences in estimated water depth of 2, 3 and 15% and (2) verified against official flooding map of an existing dam (Verzasca) detecting 75% of similarity. The problem formulation, the new GRASS module and its validation is presented.     

Modeling of unsaturated granular flows by a two-layer approach

Flows of partially saturated grain-fluid mixtures over complex curved topography are commonly observed in nature. However, comprehensive understanding of the physics behind them is to date out of reach. To investigate their dynamic process, a two-layer approach is proposed, in which the fluid-saturated granular layer is overlaid by the pure granular material. More specifically, the lower layer is described by a two-phase mixture theory of density preserving solid and fluid constituents. For the upper layer, the single-phase granular mass is treated as a frictional Coulomb-like continuum, and the dilation effect and the influence of the interstitial air are ignored. The capillarity effects and grains-size segregation are not considered in both the layers. The lower and upper layers interact at an interface which is a material surface for the fluid phase, but across which the mass exchange for the granular phase may take place. The granular mass exchange across the layer interface is parameterized by an entrainment type postulate. In addition, the classical jump conditions are employed to connect both layers at the interface dynamically. Furthermore, we perform the depth-averaged technique for the saturated grain-fluid mixture lower layer and the pure granular upper layer, respectively, to simplify the governing equations established. It is demonstrated that the resulting model equations can be reduced to most of the existing single-layer pure granular flow models and saturated two-phase single-layer debris flow models. Numerical solutions demonstrate that the present two-layer model can describe flows of partially saturated grain-fluid mixtures and the transition process of a saturated grain-fluid mixture into an under-saturated state.     

复杂碳酸盐岩油气藏建模及储量计算方法:以潜山油气储量计算为例

油气藏地质建模是储量计算的基础和前提.针对复杂碳酸盐岩储量计算中油气藏建模、参数求取等具体难点,通过石油地质、地球物理等多学科新技术、新理论的综合运用,创新性地建立了潜山准层状油气藏模式,为计算单元的正确划分、计算方法的适当选择、储量参数的合理取值提供了可靠依据.首次提出并实现了该类油气藏孔洞型和裂缝型储集体油气储量的分别计算,且对含油体积、有效厚度下限、孔隙度、饱和度等参数的求取进行了计算方法创新和软件开发,促成了轮南古隆起复杂碳酸盐岩油气储量的大幅度上升,并为类似油气储量计算提供了依据.     

Restoring groundwater levels after tunneling: a numerical simulation approach to tunnel sealing decision-making

Tunneling is often unpopular with local residents and environmentalists, and can cause aquifer damage. Tunnel sealing is sometimes used to avoid groundwater leakage into the tunnel, thereby mitigating the damage. Due to the high cost of sealing operations, a detailed hydrogeological investigation should be conducted as part of the tunneling project to determine the impact of sealing, and groundwater modeling is an accurate method that can aid decision-making. Groundwater-level drawdown induced by the construction of the Headrace water-conveyance tunnel in Sri Lanka dried up 456 wells. Due to resulting socio-environmental problems, tunnel sealing was decided as a remedy solution. However, due to the expectation of significant delays and high costs of sealing, and because the water pressure in the tunnel may prevent groundwater seepage into the tunnel during operation, there was another (counter) decision that the tunnel could remain unsealed. This paper describes groundwater modeling carried out using MODFLOW to determine which option—sealed or unsealed tunnel—is more effective in groundwater level recovery. The Horizontal Flow Barrier and River packages of MODFLOW were used to simulate sealed and unsealed tunnels, respectively. The simulation results showed that only through tunnel sealing can the groundwater level be raised to preexisting levels after 18 years throughout the study area. If the tunnel remains unsealed, about 1 million m³/year of water conveyed by the tunnel will seep into the aquifer, reducing the operational capacity of the tunnel as a transport scheme. In conclusion, partial tunnel sealing in high-impact sections is recommended.

     

Recent iron ore formation in Lake Malawi,Africa

Nontronite, limonite (with opal) and vivianite are forming at present in the aereated, shallower parts (water depths <250 m) of Lake Malawi. They overlie diatomite or coarse grained clastic sediments. Our investigations indicate a precipitation of nontronite and limonite (and opal) at the sediment/oxic water interface from geothermal solutions rich in SiO₂ percolating through the sedimentary fill of the basin. Under reducing conditions and a pH less than 7 (as occurring in the deeper parts of our sediment cores) iron and manganese are leached and discharged into the lake. In the anoxic parts of the lake (water depth greater than 250 m) Fe³⁺- and Mn³⁺-hydroxide are precipitated within the lake's water in the mixing zone of the aerobic with the anaerobic water bodies. High dispersion and a strong supply of detrital material, however, prevent a stronger enrichment of the hydroxides in the sediment. The formation of vivianite can be explained by a dissolution of Ca-phosphate (fish debris) within the sediment and a re-deposition as Fe-phosphate in the uppermost sediment layers under reducing but slightly alkaline conditions. The results of our investigations on Recent iron formation in Lake Malawi offer an explanation for the genesis of certain sedimentary iron ores in the geological past. They indicate that a formation of iron-rich sediments — including the silicate facies — is not restricted to the marine environment.

---

Zusammenfassung Nontronit, Limonit (und Opal) sowie Vivianit werden an zahlreichen Stellen des südlichen Malawi-Sees in Wassertiefen bis 250 m (aerober Bereich) abgeschieden und bilden dort die jüngste sedimentäre Einheit über Diatomit oder grobklastischem Sediment. Es muß angenommen werden, daß die Bildung der Eisenmineralien aus geothermalen Lösungen erfolgt, die reich an gelöster Kieselsäure und Fe²⁺ (untergeordnet Mn²⁺) sind, wobei Eisen und Mangan aus dem Sediment selber stammen, da heiße Quellen außerhalb des Malawi-Sees diese Elemente nicht enthalten. Die Auslösung von Eisen und Mangan erfolgt durch saure Lösungen im reduzierenden Milieu unter Bedingungen, wie sie im tieferen Teil der Sediment-Kerne angetroffen wurden (pH bis 3,6!). Beim Zusammentreffen dieser Lösungen mit dem sauerstoffhaltigen, schwach alkalischen Wasser des Malawi-Sees kommt es zur Ausfällung von Nontronit oder — bei wahrscheinlich höherem Redox-Potential — von Limonit und Opal. Beim Austritt der Fe²⁺, Mn²⁺- und SiO₂-reichen Lösungen in den anaeroben Teil des Malawi-Sees (Wassertiefen größer als 250 m) unterbleibt zunächst eine Ausfällung. Erst in der Mischzone des anoxischen mit dem oxischen Wasserkörper können Fe³⁺- und Mn³⁺-Hydroxide ausgeschieden werden, die jedoch infolge hoher Dispergierung und starker Zufuhr von klastischem Material eine starke Verdünnung erfahren. Für die Vivianit-Bildung kann angenommen werden, daß ein in den Sedimenten häufig vorhandener Apatit-Anteil (Fischreste) durch die sauren Lösungen aufgelöst und in den obersten Sedimentschichten (reduzierendes Milieu, pH jedoch >7) zusammen mit Fe²⁺ der Lösung als Vivianit abgeschieden wird. Die vorliegenden Untersuchungen geben Hinweise auf die Bildung bestimmter fossiler sedimentärer Eisenerze und zeigen gleichzeitig, daß die Genese eisenreicher Sedimente — auch von silikatischem Fazies-Typ — nicht an den marinen Ablagerungsraum gebunden ist.

     

Modeling erosion and sediment control practices with RUSLE 2.0: a management approach for natural gas well sites in Denton County,TX, USA

Sediment yields from natural gas well sites in Denton County, TX, USA can be substantial and warrant consideration of appropriate erosion and sediment control best management practices (BMPs). Version 2 of the revised universal soil loss equation (RUSLE 2.0) was used to predict sediment yields and evaluate the efficiency of BMPs for multiple combinations of different land surface conditions (soil erodibility and slope) commonly found at gas well sites in the area. Annual average sediment yield predictions from unprotected site conditions ranged from 12.1 to 134.5 tonnes per hectare per year (t/ha/yr). Sediment yield predictions for 1, 2, 5, and 10-year design storms ranged from 8.1 to 20.6 t/ha. When site conditions were modeled with BMPs, predicted sediment yields were 52–93% less. A comparison of modeled efficiency values to a review of laboratory and field data suggests that modeled (theoretical) sediment yield results with BMPs are likely best case scenarios. This study also evaluated BMPs in the context of site management goals and implementation cost, demonstrating a practical approach for the application of RUSLE 2.0 for managing soil loss and understanding the importance of selecting appropriate site-specific BMPs for disturbed site conditions.     

Life-time prediction for rocks under static compressive and tensile loads: a new simulation approach

The life time or time to failure of rocks under load is governed by microstructural defects, like microcracks, voids etc. The life time can be predicted either by empirical exponential laws or physical laws based on damage and fracture mechanics. The proposed numerical model is based on subcritical crack growth using the linear elastic fracture mechanical approach and is implemented as a numerical cellular automate. The algorithm considers both tensile and shear fracturing. Each cell contains a microcrack of random length according to a given probability function. Fracture growth is controlled by the Charles equation. Macroscopic cracks are the results of the coalescence of growing microcracks. Within the numerical approach elasto-plastic stress redistributions take place. If the stress intensity factors have reached the critical values or the microcrack has reached the zone dimension, the zone is considered as fractured and residual strength values are assigned. The proposed approach was applied to rock samples under uniaxial compressive and tensile loads (creep tests). Successful results were obtained in respect to the predicted life time, damage evolution and the fracture pattern. Conclusions for further improvements and extensions of this methodology were drawn.     

Modeling of chemo-hydromechanical behavior of unsaturated porous media: a nonlocal approach based on integral equations

Acta Geotechnica - Unsaturated clay is a heterogeneous porous medium consisting of three phases, namely solid soil skeleton, pore water, and pore air. It has been well recognized that the variation...     

Differences in groundwater response to deforestation — a continuum of interactions between hydroclimate,landscape characteristics and time

After several decades of catchment research, most forest hydrologists agree that deforestation leads to increased annual catchment yields and augment groundwater recharge and dry-season flow. Nevertheless, millions of people in the semi-arid tropics have good reasons — a deep knowledge of the land after using it for many generations — to state the very opposite, ie that deforestation implies dry springs and streams. In this paper a model is presented which attemps to show that both of the above views are correct, but only if the hydrological consequences of deforestation are related to the diversity in land management, hydroclimate and landscape characteristics. The possibility of closing the dichotomy of opinion between people and professionals is discussed.     

典型森林生态系统碳交换的机理模拟及其与观测的比较研究

CEVSA模型是一个基于生理生态过程模拟植物—土壤—大气系统能量交换和水碳氮耦合循环及其对环境变化响应和适应的机理模型,在区域和全球尺度上得到广泛应用,但缺乏在生态系统尺度上的验证。本研究对CEVSA模型中碳循环关键过程的定量表达进行了重要改进：将模型的模拟时间步长由“旬”改为“日”;增加了物候参数化的子程序;调整了生物量的分配方案;基于碳平衡模拟叶面积指数的季节动态等。分别使用改进前后的CEVSA模型模拟了3类典型森林生态系统碳交换的季节动态,模拟的结果与通量观测进行了比较分析。结果表明,改进后的CEVSA模型能更好地模拟不同类型森林生态系统碳交换的季节动态,但在不同生态系统的不同时段,模型模拟的结果与通量观测相比还有一定的偏差,模型在碳交换关键过程对环境变化的响应和适应,尤其是针叶林光合作用对温度变化的响应和适应,以及生态系统对高温和水分亏缺的响应和适应等方面的模拟还需要进一步的改进和验证。     

Modeling and mapping of water erosion in northeastern Algeria using a seasonal multicriteria approach

Aurès region remains one of the most exposed areas to water erosion phenomenon in Algeria, because of the strong climatic aggressiveness, the rugged relief, the predominance of sensitive land, and a vegetative cover that does not play its protective role. This article is a part of studies performed to protect agricultural and water infrastructure in this region. The main objective of this study is the cartographic modeling of an erosion hazard at the Oued Chemoura watershed, representative of the Aurès. The modeling approach uses a geographic information system and incorporates the following six factors controlling erosion: slope, friability of substrate, erodibility of soils, land cover, rainfall erosivity, and support practices. Result shows a synthetic map of the soil erosion hazard which locates the most threatened areas and priorities for possible planning interventions. A statistical study on the relationship solid–liquid flow was developed. Measurements conducted at the station of Chemoura, over the period 1969–1994, were exploited for this purpose. The results show a high specific degradation varying between 50 and 360 tons km^?2 season^?1.     

Impacts of primary deforestation upon limestone slopes in northern Vancouver Island, British Columbia

Effects of deforestation upon slopes in limestones and in volcanic rocks in the Benson River valley, northern Vancouver Island, have been investigated quantitatively. Postlogging soil erosion and vegetal regeneration success were assessed by measuring soil depth, percent bare rock and moss cover, and the numbers and diversity of trees, shrubs, and plants on 25 sampling sites, each containing ten measuring quadrats selected at random. Sixteen sites were on the Quatsino Formation, a well-karstified limestone, and nine on the Karmutsen Formation of basaltic lavas. Eight sites were of virgin forest, 16 were logged between 1970 and 1983, and one (on limestone) was logged in 1911. Both bedrock types were significantly affected by the cutting. There was greater loss of soil and an increase in bare rock on the limestones. Erosion was increased significantly by burning on the limestones but not on the volcanics. Within-group comparisons on the limestones determined that steeper slopes and harder burned areas suffered the most and are slowest to regenerate. Volume of timber on the 1911 site was 19 percent of that in similar uncut forest sites. It appears that complete recovery on the barren limestone slopes will require at least some centuries.     

Sequential Gaussian simulation for geosystems modeling: A machine learning approach

Sequential Gaussian Simulation(SGSIM)as a stochastic method has been developed to avoid the smoothing effect produced in deterministic methods by generating various stochastic realizations.One of the main issues of this technique is,however,an intensive computation related to the inverse operation in solving the Kriging system,which significantly limits its application when several realizations need to be produced for uncertainty quantification.In this paper,a physics-informed machine learning(PIML)model is proposed to improve the computational efficiency of the SGSIM.To this end,only a small amount of data produced by SGSIM are used as the training dataset based on which the model can discover the spatial correlations between available data and unsampled points.To achieve this,the governing equations of the SGSIM algorithm are incorporated into our proposed network.The quality of realizations produced by the PIML model is compared for both 2D and 3D cases,visually and quantitatively.Furthermore,computational performance is evaluated on different grid sizes.Our results demonstrate that the proposed PIML model can reduce the computational time of SGSIM by several orders of magnitude while similar results can be produced in a matter of seconds.     

Population and poverty in Jamaican deforestation: Integrating satellite and household census data

This study uses MSS data to derive sub-national level deforestation rates at the constituency administrative level for Jamaica for 1987 and 1992. It then investigates the role of poverty and population in driving forest loss during this period by linking these estimates in a GIS with constituency level demographic and socioeconomic census data for the island. OLS regression results support the importance of population pressures and poverty in driving the destruction of Jamaica's forests and the relative contribution to deforestation of their various measures are noted and discussed. In addition to providing information on Jamaica's deforestation attributes, the study demonstrates how remotely sensed data can be used in conjunction with household census data to derive information on human-forest interactions at the sub-national level. A small simulation experiment based on regression results using key variables suggests that under any scenario, the impacts of key social and demographic changes on Jamaica's remaining forest cover may be substantial by the year 2010. This revised version was published online in July 2006 with corrections to the Cover Date.     

Modeling of a permeate flux of cross-flow membrane filtration of colloidal suspensions: A wavelet network approach

Although traditional artificial neural networks have been an attractive topic in modeling membrane filtration, lower efficiency by trial-and-error constructing and random initializing methods often accompanies neural networks. To improve traditional neural networks, the present research used the wavelet network, a special feedforward neural network with a single hidden layer supported by the wavelet theory. Prediction performance and efficiency of the proposed network were examined with a published experimental dataset of cross-flow membrane filtration. The dataset was divided into two parts: 62 samples for training data and 329 samples for testing data. Various combinations of transmembrane pressure, filtration time, ionic strength and zeta potential were used as inputs of the wavelet network so as to predict the permeate flux. Through the orthogonal least square alogorithm, an initial network with 12 hidden neurons was obtained which offered a normalized square root of mean square of 0.103 for the training data. The initial network led to a wavelet network model after training procedures with fast convergence within 30 epochs. Futher the wavelet network model accurately depicted the positive effects of either transmembrane pressure or zeta potential on permeate flux. The wavelet network also offered accurate predictions for the testing data, 96.4 % of which deviated the measured data within the ± 10 % relative error range. Moreover, comparisons indicated the wavelet network model produced better predictability than the back-forward backpropagation neural network and the multiple regression models. Thus the wavelet network approach could be employed successfully in modeling dynamic permeate flux in cross-flow membrane filtration.     

Modeling humic acids transport in a bauxite profile: Los Pijiguaos, Venezuela

Los Pijiguaos bauxite deposit is located in Southern Venezuela, in a tropical rain forest environment. Among the various impurities during the bauxite processing (Bayer process), organic matter compounds present in Bayer liquor can slow the rate of alumina precipitation, and reduce product yield and quality. Soil samples were taken from a 37-m-deep core and the distribution of organic carbon (OC) was investigated as well as its relationship with major oxides Al₂O₃, Fe₂O₃, SiO₂ and with the mineral phases gibbsite, hematite, kaolinite and quartz. Each sample was subjected to an adsorption process of humic acids by interaction with solutions of known concentrations. Maximum adsorption capacity was also studied in each mineral separately, which allowed calculation of the theoretical adsorption capacity of each soil sample. Even if concentrations of OC were higher in the first 5 m of the profile (0.2–0.5%), samples exhibited a low adsorption capacity (0.02%), while at depth between 10 and 37 m, samples showed lower OC concentrations (0.03–0.05%), and a larger OC adsorption capacity (0.08%). The model for the adsorption capacity only works between 10 and 30 m of depth, since in the first 10 m the bauxitic profile seems to be saturated in OC, because the concentration of OC is higher than the maximum predicted by adsorption experiments.