Knowledge discovery from multispectral satellite images

A new approach to extract knowledge from multispectral images is suggested. We describe a method to extract and optimize classification rules using fuzzy neural networks (FNNs). The FNNs consist of two stages. The first stage represents a fuzzifier block, and the second stage represents the inference engine. After training, classification rules are extracted by backtracking along the weighted paths through the FNN. The extracted rules are then optimized by use of a fuzzy associate memory bank. We use the algorithm to extract classification rules from a multispectral image obtained with a Landsat Thematic Mapper sensor. The scene represents the Mississippi River bottomland area. In order to verify the rule extraction method, measures such as the overall accuracy, producer's accuracy, user's accuracy, kappa coefficient, and fidelity are used.     

A hydrometeorological analysis of the severe rainstorm of 20–22 September 1900 over the Gangetic West Bengal

A hydrometeorological study is made of the September, 1900 severe rainstorm which led up to the record rain-falls over Gangetic West Bengal with subsequent disastrous flooding in the Damodar and the Hooghly rivers. The spatial extent of the rainstorm for different durations has been examined by constructing the isohyetal patterns based on rainfall records of stations affected by the storm. Areal rainfalls for 1,2 and 3-day periods are calculated and the values have been compared with similar values from other major rainstorms of the region. The comparison revealed that the September, 1900 rainstorm was the heaviest for 1,2 and 3-day durations for all the areas. The storm contrib-uted rainfalls of 33.0 cm, 52.0 cm and 62.0 cm over an area of 10,000 km2 in 1,2 and 3 days respectively. This rainstorm could, therefore, be considered as an important input in flood and design storm studies in the Gangetic West Bengal region. A relationship between point to areal rainfall has also been developed with a view to evaluate areal PMP estimates.     

Modeling elasto-visco-bio-plastic mechanical behavior of municipal solid waste in landfills

Acta Geotechnica - This paper presents a mechanical model for the prediction of short-term and long-term settlement of municipal solid waste in landfills. The load-induced volumetric compression,...     

Geospatial and MCDM tool mix for identification of potential groundwater prospects in a tropical river basin,Kerala

The efficiency of GIS, RS and multi-criteria tools in isolating potential groundwater (GW) zones in the Kuttiyadi River basin (KRB), Kerala, has been robustly demonstrated by analysis of relevant data. To infer geohydrological makeup and consequent behavior of the KRB in respect of GW potential, firstly, various thematic layers viz. geomorphology, geology, slope, soil, lineament density and drainage density, were created. Secondly, thematic layers and their features were assigned suitable weights on the Saaty’s scale according to their relative significance for the presence and potential of GW. The assigned weights of the layers and their features were normalized using analytic network process method, and then the selected thematic maps were integrated in GIS using weighted overlay method to create the final groundwater prospect zone map. From the outcomes, the groundwater prospect zones of the KRB basin was found to be very good (166.21 km²), good (92.01 km²), moderate (180.33 km²), poor (237.25 km²), which constitute 24, 15, 26 and 35% of the study area, respectively. The GW prospect zone map was finally validated using geohydrology of area and GW level data from 43 phreatic wells in the study area. This study showed that groundwater prospect zone demarcation along with multi-criteria decision making is a powerful tool for proper utilization, planning and management of the precious groundwater resource.     

Significance of glacio-morphological factors in glacier retreat: a case study of part of Chenab basin,Himalaya

A study has been carried out in part of Chenab basin,Himalaya to understand the relationship between glacio-morphological factors and change in glacial area. Initially change in areal extent of glaciers was derived for two time frames(1962-2001/02 and 2001/02-2010/11). The study comprised of 324 glaciers for the monitoring period of 1962-2001/02 for,which 11% loss in glacial area was observed. Two hundred and thirty-eight glaciers were further monitored between 2001/02 and 2010/11. These glaciers showed an area loss of 1.1%. The annual deglaciation has been found to be higher during the period of 1962-2001/02 compared to 2001/02-2010/11. The spatial and temporal variability in deglaciation was also addressed usingglacio-morphic parameters. Area,length,percentage of debris cover,and various elevation parameters of glaciers were observed to have significant controls on relationships to the rate of glacial shrinkage. Largerarea and longer glaciers show a lower percentage of retreat than smaller and shorter ones. Moreover,glaciers located at lower altitudes and having gentle slopes show more area retreat. The results of area retreat in debris covered and debris free glaciers supports that the glaciers covered by debris retard ice melting at some extent. 158 glaciers were observed having no debris cover,and these exhibit 14% of loss in surface area. In glaciers having 40% debris cover,8% of deglaciation was observed. The glaciers located below equilibrium line altitude(ELA) have experienced 4.6% of deglaciation for the time frame 2001/02 – 2010/11 whereas it was found to be 1.1% for the glaciers occurring above ELA. However,theorientation of glaciers did not show any considerable influence on glacial change based on hypothesis.     

Study of Dynamical Structure of an Unusual Monsoon Depression formed over the Bay of Bengal during August 2006

Monsoon depressions, the main rain-producing systems over the Indian region along and near their tracks, are found to intensify the monsoon circulation by organizing low-level convergence. The normal track of the monsoon depressions is along the position of the monsoon trough at the surface, i.e., northwestward from the Head Bay of Bengal. Most of the monsoon depressions dissipate within one or two days after landfall. An unusual monsoon depression formed in the Bay of Bengal during the 1^st week of August 2006 causing heavy to very heavy rainfall over Madhya Pradesh, Maharashtra and Gujarat States of India. The track of this depression was anomalously southward from the mean track of the August depressions. It maintained its intensity during its longer travel. This paper addresses some of the dynamical characteristics of the depression in relation to its southward/westward track and longer travel. It is observed that horizontal advection of absolute vorticity above 550 hPa (below 600 hPa) along west (east) of the depression and maximum divergence of absolute vorticity below 400 hPa dominated for the westward movement of the depression. Increased moisture supply from the Arabian Sea (after the landfall of the depression) helped to maintain the intensity of the system throughout its long travel. The energy conversion terms revealed the strengthening of the zonal flow at higher levels prior to the formation of the depression.     

Observed changes in extreme rain indices in semiarid and humid regions of Godavari basin,India: risks and opportunities

Natural Hazards - The present study examines the spatial and temporal changes in extreme rainfall events (EREs) across the Godavari river basin to develop adaptation strategies for vulnerable...     

Progress in Remediation of Groundwater at Petroleum Sites in California

Quantifying the overall progress in remediation of contaminated groundwater has been a significant challenge. We utilized the GeoTracker database to evaluate the progress in groundwater remediation from 2001 to 2011 at over 12,000 sites in California with contaminated groundwater. This paper presents an analysis of analytical results from over 2.1 million groundwater samples representing at least $100 million in laboratory analytical costs. Overall, the evaluation of monitoring data shows a large decrease in groundwater concentrations of gasoline constituents. For benzene, half of the sites showed a decrease in concentration of 85% or more. For methyl tert‐butyl ether (MTBE), this decrease was 96% and for TBE, 87%. At remediation sites in California, the median source attenuation rate was 0.18/year for benzene and 0.36/year for MTBE, corresponding to half‐lives of 3.9 and 1.9 years, respectively. Attenuation rates were positive (i.e., decreasing concentration) for benzene at 76% of sites and for MTBE at 85% of sites. An evaluation of sites with active remediation technologies suggests differences in technology effectiveness. The median attenuation rates for benzene and MTBE are higher at sites with soil vapor extraction or air sparging compared with sites without these technologies. In contrast, there was little difference in attenuation rates at sites with or without soil excavation, dual phase extraction, or in situ enhanced biodegradation. The evaluation of remediation technologies, however, did not evaluate whether specific systems were well designed or implemented and did not control for potential differences in other site factors, such as soil type.     

Time vs. Money: A Quantitative Evaluation of Monitoring Frequency vs. Monitoring Duration

The National Research Council has estimated that over 126,000 contaminated groundwater sites are unlikely to achieve low ug/L clean‐up goals in the foreseeable future. At these sites, cost‐effective, long‐term monitoring schemes are needed in order to understand the long‐term changes in contaminant concentrations. Current monitoring optimization schemes rely on site‐specific evaluations to optimize groundwater monitoring frequency. However, when using linear regression to estimate the long‐term zero‐order or first‐order contaminant attenuation rate, the effect of monitoring frequency and monitoring duration on the accuracy and confidence for the estimated attenuation rate is not site‐specific. For a fixed number of monitoring events, doubling the time between monitoring events (e.g., changing from quarterly monitoring to semi‐annual monitoring) will double the accuracy of estimated attenuation rate. For a fixed monitoring frequency (e.g., semi‐annual monitoring), increasing the number of monitoring events by 60% will double the accuracy of the estimated attenuation rate. Combining these two factors, doubling the time between monitoring events (e.g., quarterly monitoring to semi‐annual monitoring) while decreasing the total number of monitoring events by 38% will result in no change in the accuracy of the estimated attenuation rate. However, the time required to collect this dataset will increase by 25%. Understanding that the trade‐off between monitoring frequency and monitoring duration is not site‐specific should simplify the process of optimizing groundwater monitoring frequency at contaminated groundwater sites.