A Tool for Classification and Regression Using Random Forest Methodology: Applications to Landslide Susceptibility Mapping and Soil Thickness Modeling

Classification and regression problems are a central issue in geosciences. In this paper, we present Classification and Regression Treebagger (ClaReT), a tool for classification and regression based on the random forest (RF) technique. ClaReT is developed in Matlab and has a simple graphic user interface (GUI) that simplifies the model implementation process, allows the standardization of the method, and makes the classification and regression process reproducible. This tool performs automatically the feature selection based on a quantitative criterion and allows testing a large number of explanatory variables. First, it ranks and displays the parameter importance; then, it selects the optimal configuration of explanatory variables; finally, it performs the classification or regression for an entire dataset. It can also provide an evaluation of the results in terms of misclassification error or root mean squared error. We tested the applicability of ClaReT in two case studies. In the first one, we used ClaReT in classification mode to identify the better subset of landslide conditioning variables (LCVs) and to obtain a landslide susceptibility map (LSM) of the Arno river basin (Italy). In the second case study, we used ClaReT in regression mode to produce a soil thickness map of the Terzona catchment, a small sub-basin of the Arno river basin. In both cases, we performed a validation of the results and a comparison with other state-of-the-art techniques. We found that ClaReT produced better results, with a more straightforward and easy application and could be used as a valuable tool to assess the importance of the variables involved in the modeling.     

Application of Adjoint Approach to Oil Spill Problems

Adjoint method is applied to various oil spill problems. A three-dimensional model for describing the dispersion of a quasi-passive substance (a pollutant or a nutrient) and its adjoint model are considered in a limited sea region. Direct and adjoint estimates are used to get dual (equivalent) estimates of the mean concentration of the substance in important zones of the region. The role of dual estimates is illustrated with a few examples. They include such oil spill problems as the search of the most dangerous point of the oil tanker route, the oil dispersion with a climatic velocity, and the dependence of the oil concentration estimates on the oil spill rate. One more example is the application of optimal bioremediation strategy for cleaning a few zones polluted by oil. In this case, instead of oil, the model describes the dispersion of a nutrient released to marine environment. Balanced, unconditionally stable second-order finite-difference schemes based on the splitting method for the solution of the dispersion model and its adjoint are suggested. The main and adjoint difference schemes are compatible in the sense that at every fractional step of the splitting algorithm, the one-dimensional split operators of both schemes satisfy a discrete form of Lagrange identity. In the special unforced and non-dissipative case, each scheme has two conservation laws. Every split one-dimensional problem is solved by Thomas’ factorization method.     

Monitoring water quality in reservoirs for human supply through multi-biomarker evaluation in tropical fish

Paraíba do Sul River is located at a very densely inhabited region of Brazil crossing the three most industrialized states of the country (S?o Paulo, Minas Gerais and Rio de Janeiro states). As a result, industrial and farming residues as well as urban sewage are frequently disposed without appropriate treatment. The current study aimed at investigating the water quality in three reservoirs along the Paraíba do Sul River (Ilha dos Pombos, Santa Cecília and Santa Branca), through physiological, morphological, biochemical, and genetic biomarkers. The bioindicator chosen was the catfish Pimelodus maculatus, sampled during the dry (June 2008) and rainy (February 2009) seasons. Also, some water physicochemical parameters were analyzed from the sampling sites, but displayed no alterations according to the Brazilian Agency for Water Quality Legislation. Branchial carbonic anhydrase activity was inhibited in the dry season, while renal carbonic anhydrase activity was inhibited in the rainy season in the Santa Branca reservoir, indicating disturbance of osmoregulatory and acid-base regulation processes. Histopathological alterations were observed in the gills (neoplasic and tissue hyperplasia processes) and liver (necrosis), indicating serious damage to the functional integrity of these organs. A high incidence of melanomacrophage centers was observed in the liver, suggesting an intense immune response in all reservoirs. Acetylcholinesterase and catalase activity showed also differences corroborating some morphological results. Likewise, the induction of the micronucleus and DNA damage indicate genotoxicity, but mainly in the Santa Branca reservoir. Thus, the health status of P. maculatus warrants caution in the use of the water from the 3 reservoirs for direct human consumption, particularly after the accidental spill of endosulfan in November 2008, three months before the rainy season sampling.     

Integrated approach for demarcating subsurface pollution and saline water intrusion zones in SIPCOT area: a case study from Cuddalore in Southern India

This paper deals with a systematic hydrogeological, geophysical, and hydrochemical investigations carried out in SIPCOT area in Southern India to demarcate groundwater pollution and saline intrusion through Uppanar River, which flows parallel to sea coast with high salinity (average TDS 28, 870 mg/l) due to back waters as well as discharge of industrial and domestic effluents. Hydrogeological and geophysical investigations comprising topographic survey, self-potential, multi-electrode resistivity imaging, and water quality monitoring were found the extent of saline water intrusion in the south and pockets of subsurface pollution in the north of the study area. Since the area is beset with highly permeable unconfined quaternary alluvium forming potential aquifer at shallow depth, long-term excessive pumping and influence of the River have led to lowering of the water table and degradation of water quality through increased salinity there by generating reversal of hydraulic gradient in the south. The improper management of industrial wastes and left over chemicals by closed industries has led surface and subsurface pollution in the north of the study area.     

Denitrification potential of the Northern Arabian Gulf—an experimental study

The study aims to establish denitrification potential of the Northern Arabian Gulf (NAG), as nitrogen critically affects the ocean productivity, obliterates acidity, oxidative capacity and radiative transfer capability of atmosphere. The experimental study was conducted by taking cores from intertidal zones from two different sites in North and South, referred as sites N and S; representing two distinct environmental milieu. The experiment was conducted in controlled laboratory conditions simulating the tidal cycles. Multiple cores were taken and loaded with seawater with different N concentrations, the redox potential was established for each condition. Redox potential was significantly lower at 10?cm depth compared to the surface in all cores (P?<?0.001). The redox potential at surface and at 10?cm depth was significantly lower at site S compared to site N (P?<?0.001; F?=?714.2), suggesting anaerobic sediments at site S. Effects of nitrate spiked seawater on denitrification under nonflooded and flooded conditions at the two sites were also studied. Three-way ANOVA analysis indicated that site, nitrate concentration, and flooding had significant main and interactive effects on the rate of denitrification. The results suggest that under ambient nitrate concentrations (0.03?mg NO₃-N?l^?1), 6.3?±?2.1?g NO₃-N?ha?day can be denitrified by inter-tidal zone sediments. At a nitrate concentration of 1?mg NO₃-N?l^?1, 92?±?16?g NO₃-N?ha?day may be denitrified whilst at a very high nitrate load of 10?mg NO₃-N?l^?1, the sediments may attain a rate of denitrification close to 404?±?78?g NO₃-N?ha?day.     

Bioaccumulation surveillance in Milford Haven Waterway

Biomonitoring of contaminants (metals, organotins, polyaromatic hydrocarbons (PAHs), PCBs) was undertaken in Milford Haven Waterway (MHW) and a reference site in the Tywi Estuary (St Ishmael/Ferryside) during 2007–2008. Bioindicator species encompassed various uptake routes—Fucus vesiculosus (dissolved contaminants); Littorina littorea (grazer); Mytilus edulis and Cerastoderma edule (suspension feeders); and Hediste (=Nereis) diversicolor (sediments). Differences in feeding and habitat preference have subtle implications for bioaccumulation trends though, with few exceptions, contaminant burdens in MHW were higher than the Tywi reference site, reflecting inputs. Elevated metal concentrations were observed at some MHW sites, whilst As and Se (molluscs and seaweed) were consistently at the higher end of the UK range. However, for most metals, distributions in MH biota were not exceptional. Several metal-species combinations indicated increases in bioavailability upstream, which may reflect the influence of geogenic/land-based sources—perhaps enhanced by lower salinity. TBT levels in MH mussels were below OSPAR toxicity thresholds and in the Tywi were close to zero. Phenyltins were not accumulated appreciably in M. edulis, whereas some H. diversicolor populations appear subjected to localized (historical) sources. PAHs in H. diversicolor were distributed evenly across most of MHW, although acenaphthene, fluoranthene, pyrene, benzo(a)anthracene and chrysene were highest at one site near the mouth; naphthalenes in H. diversicolor were enriched in the mid-upper Haven (a pattern seen in M. edulis for most PAHs). Whilst PAH (and PCB) concentrations in MH mussels were mostly above reference and OSPAR backgrounds, they are unlikely to exceed ecotoxicological thresholds. Bivalve Condition indices (CI) were highest at the Tywi reference site and at the seaward end of MH, decreasing upstream—giving rise to several significant (negative) relationships between CI and body burdens. Despite the possible influence of salinity gradient as a complicating factor, multivariate analysis indicated that a combination of contaminants could influence the pattern in condition (and the biomarkers metallothionein and TOSC). Integrating bioaccumulation data with biological and biochemical endpoints is seen as a useful way to discriminate environmental quality of moderately contaminated areas such as MHW and to prioritise cause and effect investigations.     

Mixed response in bacterial and biochemical variables to simulated sand mining in placer-rich beach sediments,Ratnagiri, West coast of India

We investigated the influence on bacterial community and biochemical variables through mechanical disturbance of sediment-akin to small-scale mining in Kalbadevi beach, Ratnagiri, a placer-rich beach ecosystem which is a potential mining site. Changes were investigated by comparing three periods, namely phase I before disturbance, phase II just after disturbance, and phase III 24 h after disturbance as the bacterial generation time is ≤7 h. Cores from dune, berm, high-, mid-, and low-tide were examined for changes in distribution of total bacterial abundance, total direct viability (counts under aerobic and anaerobic conditions), culturability and biochemical parameters up to 40 cm depth. Results showed that bacterial abundance decreased by an order from 10⁶ cells g^− 1 sediment, while, viability reduced marginally. Culturability on different-strength nutrient broth increased by 155% during phase II. Changes in sedimentary proteins, carbohydrates, and lipids were marked at berm and dune and masked at other levels by tidal influence. Sedimentary ATP reduced drastically. During phase III, Pearson’s correlation between these variables evolved from non-significant to significant level. Thus, simulated disturbance had a mixed effect on bacterial and biochemical variables of the sediments. It had a negative impact on bacterial abundance, viability and ATP but positive impact on culturability. Viability, culturability, and ATP could act as important indicators reflecting the disturbance in the system at short time intervals. Culturability, which improved by an order, could perhaps be a fraction that contributes to restoration of the system at bacterial level. This baseline information about the potential mining site could help in developing rational approach towards sustainable harnessing of resources with minimum damage to the ecosystem.     

Biological and economic optima in a tropical grazing ecosystem in Australia

Our aim in this study was to identify the economically optimal level of use of resources for a grazing system in tropical Australia and to compare it with the biological optimum. Grazing management trials were conducted in the Ord River irrigation area of north-western Australia. Biologists at the Commonwealth Scientific and Industrial Research Organization provided the biological data. The liveweight gains of Kimberley shorthorn steers were recorded, after a 1-year fattening period, for different grazing management strategies. Five different stocking rates were used on irrigated pangola grass (Digitaria decumbens) fertilized with nitrogen at five different rates. All other were applied at fixed, non-limiting levels.The results of the trials were modelled by a liveweight gain production function. This production function was used in an income simulation model to identify economically optimal (profit maximizing) stocking rates. This was done in two stages. First, a base analysis was undertaken on the basis o the economic conditions prevailing in September 1987. Second, sensitivity analysis was applied to the results of the base analysis. The parameters varied were the rate of interest, store cattle prices, finished cattle prices, transport costs and the length of the planning horizon.The biological optimum, when defined as maximum liveweight gain ha⁻¹, corresponds to a much higher intensity of use of resources than the economic optimum, and yields substantially lower profits or higher losses. These discrepancies increase when economic conditions worsen, or when the planning horizon of graziers is decreased from 50 to 10 years. Therefore, the adoption of the biological optimum cannot result in an econonomically stable grazing system in the medium to long term, when it is highly likely that economic parameters will fluctuate. The economic stability of the system is increased, but is still quite weak, if the economically optimal input combination is utilized. It was not possible to determine whether this economic optimum would result in long-term ecological stability because the relevant data were not collected during the trials.The relevance of management recommendations made to graziers and land administrators would be increased if these recommendations were based on an analysis of both the economic and the ecological stability of grazing systems over the medium to long term. We show in the last part of this study how this could be accomplished.     

Sorption and Biodegradation of Vapor-Phase Organic Compounds with Wastewater Sludge and Food Waste Compost

ABSTRACT

To test the possible use of composted food waste and wastewater sludge as biofilters to treat gas-phase volatile organic compounds (VOCs), batch experiments were conducted with an isolated strain that could degrade aromatic compounds under aerobic conditions. A benzene and trichloroethylene (TCE) mixture was used as the gas-phase pollutant in experiments with composted food waste, sludge, and soil. Under aerobic conditions, benzene was degraded as a primary substrate and TCE was degraded cometabolically, with water contents varying from 6 to 60% (volume of water added/volume of solid). Optimal water content for VOC removal was 12% for the soil, 36% for the composted food waste, and 48% for the sludge.

The extent of VOC sorption and biodegradation at the optimal water content was different for each material. With the same initial VOC concentration, more VOCs were removed by sorption onto the composted food waste and the sludge, while less VOCs were biodegraded in comparison with the results using soil. The reason the biodegradation in the soil was greater may be partly attributed to the fact that, due to less sorption, the aqueous-phase concentration of VOCs, which microorganisms could utilize as a carbon source or cometabolize, was higher. We also speculate that the distribution of microorganisms in each medium affects the rate of biodegradation. A large number of microorganisms were attached to the composted food waste and sludge. Mass transfer of VOCs and oxygen to these microorganisms, which appear to have been heterogeneously distributed in clusters, may have been limited, resulting in hindered biodegradation.