A Hierarchical Hybrid SVM Method for Classification of Remotely Sensed Data

The focus of this work is on developing a new hierarchical hybrid Support Vector Machine (SVM) method to address the problems of classification of multi or hyper spectral remotely sensed images and provide a working technique that increases the classification accuracy while lowering the computational cost and complexity of the process. The paper presents issues in analyzing large multi/hyper spectral image data sets for dimensionality reduction, coping with intra pixel spectral variations, and selection of a flexible classifier with robust learning process. Experiments conducted revealed that a computationally cheap algorithm that uses Hamming distance between the pixel vectors of different bands to eliminate redundant bands was quite effective in helping reduce the dimensionality. The paper also presents the concept of extended mathematical morphological profiles for segregating the input pixel vectors into pure or mixed categories which will enable further computational cost reductions. The proposed method’s overall classification accuracy is tested with IRS data sets and the Airborne Visible Infrared Imaging Spectroradiometer Indian Pines hyperspectral benchmark data set and presented.     

从雷电灾害浅谈弱电设备的雷电防护

通过近年哈尔滨市多起雷电灾害的特点,对弱电设备防雷进行分析,重点探讨雷电浪涌对弱电设备的危害及防雷措施,并提出合理的建议.     

In-cloud and below-cloud scavenging of aerosol ionic species over a tropical rural atmosphere in India

The temporal variation in concentrations of major water soluble ionic species has been studied from several rain events occurred over Gadanki (13.5 °N, 79.2 °E), located in tropical semi arid region in southern India. The contribution from rain-out (in cloud) and wash-out (below cloud) processes to the total removal of ionic species by rain events is also estimated using the pattern of variations of ionic species within an individual event. A number of rain samples were collected from each rain event during June–November in 2006, 2007 and 2008. On average, nearly 20% of the total NH ₄ ⁺ and non-sea SO ₄ ^2? is removed by in-cloud scavenging, suggesting that their removal by “below cloud” washout is relatively dominant. In contrast Na⁺, Ca²⁺, Mg²⁺, NO ₃ ^? and sea-SO ₄ ^2? are mainly removed by below-cloud scavenging or wash-out process. A significant variation in the acidity was observed within rain events with successive precipitation showing higher acidity at the final stage of the precipitation due to partial neutralization of non-sea SO ₄ ^2? . Overall, greater influence of both terrestrial and anthropogenic sources is recorded in the rain events compared to that from marine sources.     

Aerosol characteristics during winter fog at Agra,North India

Simultaneous measurements on physical, chemical and optical properties of aerosols over a tropical semi-arid location, Agra in north India, were undertaken during December 2004. The average concentration of total suspended particulates (TSP) increased by about 1.4 times during intense foggy/hazy days. Concentrations of SO₄ ²⁻, NO₃ ⁻, NH₄ ⁺ and Black Carbon (BC) aerosols increased by 4, 2, 3.5 and 1.7 times, respectively during that period. Aerosols were acidic during intense foggy/hazy days but the fog water showed alkaline nature, mainly due to the neutralizing capacity of NH₄ aerosols. Trajectory analyses showed that air masses were predominantly from NW direction, which might be responsible for transport of BC from distant and surrounding local sources. Diurnal variation of BC on all days showed a morning and an evening peak that were related to domestic cooking and vehicular emissions, apart from boundary layer changes. OPAC (Optical properties of aerosols and clouds) model was used to compute the optical properties of aerosols. Both OPAC-derived and observed aerosol optical depth (AOD) values showed spectral variation with high loadings in the short wavelengths (<1 μm). AOD value at 0.5 μm wavelength was significantly high during intense foggy/hazy days (1.22) than during clear sky or less foggy/hazy days (0.63). OPAC-derived Single scattering albedo (SSA) was 0.84 during the observational period, indicating significant contribution of absorbing aerosols. However, the BC mass fraction to TSP increased by only 1% during intense foggy/hazy days and thereby did not show any impact on SSA during that period. A large increase was observed in the shortwave (SW) atmospheric (ATM) forcing during intense foggy/hazy days (+75.8 W/m²) than that during clear sky or less foggy/hazy days (+38 W/m²), mainly due to increase in absorbing aerosols. Whereas SW forcing at surface (SUF) increased from −40 W/m² during clear sky or less foggy/hazy days to −76 W/m² during intense foggy/hazy days, mainly due to the scattering aerosols like SO₄ ^2-.     

Geochemistry of carbonate precipitation from the groundwater in a coastal region

Groundwater samples were collected from a coastal region of Andhra Pradesh to assess the possible conditions of the formation of carbonates. The area experiences a semi-arid climate and is underlain by khondalites, over which the Quaternary sediments occur. The study of the geochemistry of groundwater indicates that groundwater is mostly of fresh, with alkaline nature. The study further suggest that the breakdown of feldspars as kaolinite during rock-water interaction, releases Ca²⁺. Soils/weathered products contribute high CO₂ under the open system. The Ca²⁺ and CO₂ are added to the groundwater through the infiltrating recharge water. They subsequently precipitate as fine-grained carbonates in the weathering profile due to evapotranspiration under a freshwater environment.     

Electrical Resistivity Imaging technique to delineate coal seam barrier thickness and demarcate water filled voids

Exploration and exploitation of coal seams is one of the major resources for the energy sector in any country but at the same time water filled voids/water logged areas in the old workings of these seams are very critical problems for the coal mining industry. In such situations, disasters like inundation, landslides, collapsing of the old seams may occur. In this regard, it is necessary to find out the water saturated/water filled voids and zones in the mining areas. Since no established technique is available to find such zones, an experimental study using Electrical Resistivity Imaging (ERI) has been carried out in one of the coal mining areas near Dhanbad, to find out the feasibility of finding the barrier thickness and the water logged area in underground coal mines. The area under study forms part of Jharia coalfield in Dhanbad district, Jharkhand state. The coal bearing rocks of Barakar Formation of Lower Permian age (Gondwana period) occur in the area under a thin cover (10 m to15 m) of soil and or alluvium. Coal bearing Barakar Formations consist mainly of sandstone of varying grain size, intercalation of shale and sandstone, grey and carbonaceous-shale and coal seams. Since the water saturation reduces the resistivity of a formation to a large extent, water filled voids and old coal workings are expected to have significant resistivity contrast with the surrounding host rock. Hence, ERI technique was applied in such an environment as this technique uses high-density data acquisition both laterally and vertically by using multiple number of electrodes. Along with ERI, mise-à-la-masse (also called charged body) technique was also employed at one of the promising sites to find out the connectivity of water logged areas and also detection of these old workings from the surface measurements was analyzed. The interpreted 2D resistivity sections have clearly indicated the water bearing zone(s) along the profile which was well confirmed with the existing water level in the nearby borewells. On the other hand, this technique did not identify the size of the coal pillar and gallery (air filled voids), which might be due to the small size of the voids (i.e. about 2 m × 2 m) below a depth of 15m and more but have indicated altogether as a high resistive zone ranging from 600–1000 Ohm-m.     

Wind-driven estuarine turbidity maxima in Mandovi Estuary,central west coast of India

Systematic studies on the suspended particulate matter (SPM) measured on a seasonal cycle in the Mandovi Estuary, Goa indicate that the average concentrations of SPM at the regular station are ∼20mg/l, 5mg/l, 19mg/l and 5mg/l for June–September, October–January, February–April and May, respectively. SPM exhibits low-to-moderate correlation with rainfall indicating that SPM is also influenced by other processes. Transect stations reveal that the SPM at sea-end stations of the estuary are at least two orders of magnitude greater than those at the river-end during the monsoon. Estuarine turbidity maximum (ETM) of nearly similar magnitude occurs at the same location in two periods, interrupted by a period with very low SPM concentrations. The ETM occurring in June–September is associated with low salinities; its formation is attributed to the interactions between strong southwesterly winds (5.1–5.6ms⁻¹) and wind-induced waves and tidal currents and, dominant easterly river flow at the mouth of the estuary. The ETM occurring in February–April is associated with high salinity and is conspicuous. The strong NW and SW winds (3.2–3.7ms⁻¹) and wind-driven waves and currents seem to have acted effectively at the mouth of the estuary in developing turbidity maximum. The impact of sea breeze appears nearly same as that of trade winds and cannot be underestimated in sediment resuspension and deposition     

Magnetic anomalies of offshore Krishna-Godavari basin,eastern continental margin of India

The marine magnetic data acquired from offshore Krishna-Godavari (K-G) basin, eastern continental margin of India (ECMI), brought out a prominent NE-SW trending feature, which could be explained by a buried structural high formed by volcanic activity. The magnetic anomaly feature is also associated with a distinct negative gravity anomaly similar to the one associated with 85°E Ridge. The gravity low could be attributed to a flexure at the Moho boundary, which could in turn be filled with the volcanic material. Inversion of the magnetic and gravity anomalies was also carried out to establish the similarity of anomalies of the two geological features (structural high on the margin and the 85°E Ridge) and their interpretations. In both cases, the magnetic anomalies were caused dominantly by the magnetization contrast between the volcanic material and the surrounding oceanic crust, whereas the low gravity anomalies are by the flexures of the order of 3–4 km at Moho boundary beneath them. The analysis suggests that both structural high present in offshore Krishna-Godavari basin and the 85°E Ridge have been emplaced on relatively older oceanic crust by a common volcanic process, but at discrete times, and that several of the gravity lows in the Bay of Bengal can be attributed to flexures on the Moho, each created due to the load of volcanic material.     

Exact Bianchi type-II, VIII and IX perfect fluid cosmological models in Saez-Ballester theory of gravitation

Exact Bianchi type-II, VIII and IX cosmological models are obtained in a scalar tensor theory proposed by Saez and Ballester (Phys. Lett. A 113:467, 1986) with perfect fluid as a source. Some physical and geometrical properties of the models are studied. It is observed that the models are free from initial singularities and they are expanding with time.