Please use this identifier to cite or link to this item: http://localhost:8081/xmlui/handle/123456789/1313
Full metadata record
DC FieldValueLanguage
dc.contributor.authorJoshi, Himanshu-
dc.date.accessioned2014-09-23T04:43:46Z-
dc.date.available2014-09-23T04:43:46Z-
dc.date.issued1992-
dc.identifierPh.Den_US
dc.identifier.urihttp://hdl.handle.net/123456789/1313-
dc.guidePuri, N.-
dc.guidePande, P .K.-
dc.guideMathur, R. P.-
dc.description.abstractEnvironmental degradation in the context of human activities has necessitated the consideration of water quality evaluation and improvement as an integral part of river ecosystem management. This requires an insight into the structural and functional components of the ecosystem alongwith their interaction with the forcing functions. Thus, comprehensive physical monitoring of river ecosystem, the influencing processes and subsequent data analysis and interpretation assume great significance. Ganga river (under study) is one of the largest in the Indian subcontinent traversing a total distance of about 2525 km. During its journey, it receives waste inputs from diverse point and non point sources. The zone of the river presently under study extends from its origin to about 500 km. downstream (Narora). Almost 242 km. of this zone from Badrinath to Rishikesh falls in the mountainous region, and the remaining in upper plains region. This zone is relatively thinly populated, its economy being mainly agriculture oriented, and the industrial activity is nearly absent. Major abstractions from the river are for Irrigation and power generation. It lias not been monitored regularly and comprehensively in the past as it was regarded to be least polluted and hence, very little information on the attributes of the river ecosystem is available. In recent years, tracer studies have been conducted in rivers of many countries of the world to explore the nature of transport and dispersion of dissolved conservative substances. Due to the difficulties in handling and use of the radioactive tracers,fluorescent dyes have found much favour. However, dye dispersion studies on Indian rivers in ii general and especially on Ganga have not been reported. Water quality modelling is considered as a tool for arriving at sound management decisions as it provides a mathematical framework for simulating the response of a system under different situations. However, involvement of modelling in operational management of river quality problems in India has been lacking. A limited number of studies have been reported on Indian rivers and they are also deficient in adequate treatment of steps like calibration, validation and sensitivity analysis and/or in elaborate characterization of the ecosystem. It was against this background that the present study was planned during 1984 to 1988 in the overall framework of the Coordinated Research Project on Ganga river launched by Department of Environment, Government of India, in 1983. The study relied heavily on the field measurements and the use of the available statistical and interpretation techniques to extract meaningful information to fill in the gaps. Following objectives were envisaged : *• To conduct regular monitoring of the river at several locations comprising a macro network, subsequent analysis of samples for physical, chemical and biological variables and subjecting the resulting information to suitable statistical analyses. 2- To conduct dye dispersion studies in two selected stretches and analysis of the generated data to explore the magnitude of dispersion and nature of dispersion process i.e. Fickian or non Fickian. 3- To employ water quality modelling by adapting and applying a widely acclaimed and used river quality model of reasonable iii complexity and exploring the possibility of its further application on other Indian rivers. Twelve stations were identified in the macro monitoring network considering abstractions, additions, confluences, mixing, accessibility and the topography. Samples were collected during Dec. 1984 to Dec. 1986 and analysed for several physical, chemical and biological variables. A functionally simple Data Storage and Retrieval (DSR) system was designed. Input data were stratified under separate heads and allotted different code numbers. Further, retrieval was organized in two different formats. A good status of water quality was exhibited displaying low buffering, low organics, high dissolved oxygen (DO), nutrient deficiency and low productivity. Diurnal variation in DO was found to be insignificant. The generated data on various variables was tested for normality. Log normal, Square root, Pearson Type 3 and Log Pearson Type 3 transformations were employed. Principal Component Analysis (PCA) and Factor Analysis (FA) were employed to identify and quantify underlying patterns of variation in the data set, explaining the variation in fewer and simpler number of column vectors than the original input. FA appeared to account for more variance among the variables than PCA as observed in similar earlier studies. Upon analysis of data for two stretches viz. mountainous (upper) and plain (lower), four out of the extracted factors for each of these appeared significant, their eigenvalues being higher than 1.0. Further, specific combinations of prominent variable/s represented by each factor could be ascertained, with the variables having factor loadings higher than 0.5 considered significant for incorporation as per the current practice. Factor scores were utilized to further calculate water quality indices. Resemblance iv among monitoring stations was also ascertained employing Cluster Analysis (CA). Dye experiments were also performed in two stretches, one of about 8 km. in mountainous region and the other of about 12 km. in plain region. Rhodamine-B was injected and concentration time (c-t) curves were measured at several downstream monitoring stations. Dispersion characteristics like travel times of peak, leading edge, centroid and passage of cloud; and maximum conservative concentration were estimated along with their Interrelationships. The estimate of Longitudinal dispersion coefficient (D^ was initially obtained by Change of moment method and further refined by Fischer's routing method. This was found to be different than the D^stimates obtained from various available empirical relationships, which displayed a wide range as also reported earlier. Analysis for exploring the nature of diffusion process was based on testing against the assumptions inherent in the Fickian process i.e. requirement of constant convective velocity, linear increase in concentration variance with time, attenuation of maximum concentration with square root of time and a constant DL> The results /established a strong possibility of the process to be non Fickian. Also, application of Similarity analysis model (non Fickian) was found to result In a better prediction of the downstream c-t curves in the lower stretch. QL2SMG (improvised version of QUALL II), a one-dimensional, deterministic water quality model was selected for application in this study. All variables were considered. Separate stretches (upper and lower) were identified for the calibration and validation of the model, in view of their different boundary conditions. Extensive field surveys for measurement of physical, chemical, biological, hydrological and meteorological inputs were conducted to provide major parameter V estimates whereas others were either taken from past records and established literature or were suitably adjusted. The field surveys were planned during low flow season to ensure stable hydraulic conditions in the basin in order to comply with the steady state criteria for model application. Calibration and validation were performed, for individual variables, in an order, which was governed by their interdependence on each other. The least interactive (i.e. conservative) constituent was taken up first and the most interactive constituent (DO) was taken up in the end. Point loads appeared to have a negligible impact on the river quality in the upper stretch, whereas impact of point and incremental (regeneration from ground water) loads appeared prominent in the lower. Median Relative Error (MRE), Student's t test (95% significance level) and Root Mean Square Error (RMSE) were evaluated for each calibration and validation run for ensuring statistical acceptability. The model was successfully calibrated and validated for the collected data. Sensitivity analysis was carried out for major inputs, perturbing each of them, one by one, by a constant amount ( + 1°C for temperature, + 50 % for others except benthic fluxes for which the upper limit was kept + 150 % ). Increase in sensitivity, in general, was noticed at the end of the stretches over the intermediate locations, reflecting the effect of travel time. The results appeared to be lesser sensitive to D in comparison to other hydraulic parameters. Contribution of DO in the river was primarily found to come from reaeration, and much less, from primary production. Considerable amount of uncertainty appeared to be associated with decay rates for coliforms and BOD as also the input loads. The model appeared to be well suited for application to other Indian rivers or zones with due precautions in the light of expressed uncertainties and sensitivities. vi The study, as a whole, appeared to be fruitful considering the available resources as it yielded information on various aspects of Ganga river ecosystem, which was, hitherto, unavailable.en_US
dc.language.isoenen_US
dc.subjectCIVIL ENGINEERINGen_US
dc.subjectECOSYSTEMen_US
dc.subjectWATER QUALITY STUDIESen_US
dc.subjectRIVERINE ECOSYSTEMen_US
dc.titleWATER QUALITY STUDIES ON A RIVERINE ECOSYSTEMen_US
dc.typeDoctoral Thesisen_US
dc.accession.number247219en_US
Appears in Collections:DOCTORAL THESES (Civil Engg)

Files in This Item:
File Description SizeFormat 
WATER QUALITY STUDIES ON A RIVERINE ECOSYSTEM.pdf8.72 MBAdobe PDFView/Open


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.