Conference Papers
Permanent URI for this collectionhttps://idr.nitk.ac.in/handle/123456789/28506
Browse
17 results
Search Results
Item Hydrologic modelling of flash floods and their effects(Springer Science and Business Media Deutschland GmbH, 2021) Paul, A.R.; Kundapura, S.Flash floods are ranked on top in the number of people that are affected and amount of damages caused. The sudden nature of this disaster gives less time for the victims to prepare, thereby ending up with a disrupted social, economic and political stature. The study aims at analysing the possible peak discharges with the help of a rainfall–runoff model for the flood events that have occurred in the Harangi River basin in the Kodagu district and estimate the economic damage induced. The SCS curve number method is used for simulating the runoff. Fourteen peak events over the months of July and August of two different years are chosen for the validation and calibration of the model. Discharges are simulated using the HEC-HMS extension in WMS software. The effect of variation of rainfall and land-use practices in the runoff volume is studied. It is observed that the changes in land-use practices have more effect on the runoff volume than the rainfall volume. Rapid urbanization and industrialization has increased the intensity of flood damages. The largest flow was recorded when a natural bund of water collected in the upstream was collapsed. © Springer Nature Singapore Pte Ltd 2021.Item Flood inundation mapping of harangi river basin, kodagu, using gis techniques and hec-ras model(Springer Science and Business Media Deutschland GmbH, 2021) Dev Anand, M.R.; Kundapura, S.Flood is the most common hydrologic event frequently experienced in India. The states of Kerala, UP, West Bengal, Karnataka and Assam were the mainly affected by flood in 2018. In Kodagu, the southern district of Karnataka, many people have been affected by heavy rains. Landslides in hilly terrain and flooding have worsened the lives of people and led to the destruction of 800 homes, 240 bridges collapsed, road networks of 2225 km damaged and 65 government buildings affected. The cost of rebuilding road infrastructure and buildings is approximately Rs. 3000 crores. While developing flood mitigation measures, flood inundation maps are an essential component, which will be useful for the planning stage. The mapping is expected to estimate the prone flood zone based on river flood stage without performing additional simulations and quantification of the flood risk with respect to different vulnerability parameters giving a clear picture of the planning stage. These are going to be achieved by both 1D hydrodynamic models and GIS environment. This study gives an insight about how unscientific development activities may increase the negative impacts of natural disasters. It can support the planners to correctly identify the non-vulnerable places while rebuilding the damaged infrastructure. This can help people to resettle permanently in a safer place, so that they will not be affected in the case of future disasters. Depending on the severity of the water levels, we can identify the area for construction of hydraulic structures for flood protection. © Springer Nature Singapore Pte Ltd 2021.Item High-resolution mapping of soil properties using aviris-ng hyperspectral remote sensing data—a case study over lateritic soils in mangalore, india(Springer Science and Business Media Deutschland GmbH, 2021) Chitale, M.M.; Kundapura, S.Quick and accurate mapping of properties of soil is considered to be critical for agriculture and environmental management. Rapid assessment of soil properties is a daunting task in monitoring the environment. The conventional field sampling is a laborious as well as time-consuming job. The conventional methods are restricted to a specific region but there is a need to analyses the soil properties at landscape levels. Hence, this study emphasises on hyperspectral remote sensing which to some extent helps in rapid assessment of the properties. The hyperspectral data used for the study is AVIRIS-NG data. The study explored the potential of AVIRIS-NG hyperspectral data in mapping soil properties which were analysed by in situ laboratory methods and compared with them by geostatistical method of spatial interpolation. Hence, the method adopted for this purpose is the study on spatial variability of soil properties by using Kriging interpolation technique. Also, a review study is carried out on the visible and near-infrared analysis (VNIRA), multiple regression analysis approach and spectral angle mapper supervised classification technique on the high-resolution AVIRIS-NG Hyperspectral data, which will yield as an empirical model for predicting the soil property in question from both wet chemistry and spectral information of a representative set of samples and classifies the data accordingly. © Springer Nature Singapore Pte Ltd 2021.Item A statistical approach for comparison of secondary precipitation products(Springer Science and Business Media Deutschland GmbH, 2021) Kommu, R.; Kundapura, S.; Venkatesh, V.Meteorological data retrieval is the fundamental process for any hydrological research. Precipitation data collection from some constrained territories like high slant geography and inaccessible areas is exceptionally troublesome. Setting the rain gauges is a matter of expense and timely maintenance. To overcome these issues, satellite sensors producing high spatial and temporal resolution datasets can be utilized in the studies involving precipitation component. These satellite products are affected by biases, and hence, there is a need for calibration and verification by using ground observation data based on the statistical coefficients. In this study, the most accessible satellite data products, i.e., CHIRPS, PERSIANN-CDR and TRMM, are employed to check the accuracies against IMD gridded data for the years 2000–2012 using a statistical approach. Selecting the data product having a high coefficient of correlation and low PBIAS is utmost necessary. The current study was performed based on catchment-to-catchment (C-C) method by comparing IMD gridded data with satellite datasets obtained from Google Earth Engine. The results can highlight the data product which can conquer the issue of data inaccessibility in the investigation territory and can be utilized as reference precipitation dataset for different hydrological applications. © Springer Nature Singapore Pte Ltd 2021.Item Relative wave run-up parameter prediction of emerged semicircular breakwater(Springer Science and Business Media Deutschland GmbH, 2021) Kundapura, S.; Rao, S.; Arkal, V.H.Relative wave run-up parameter (Ru/Hi) on breakwaters is a vital component in fixing the elevation of the breakwater crest. In the present study, several soft computing methods has been employed to predict the wave run-up on the emerged seaside perforated semicircular breakwater for the prevailing Arabian sea wave climate, off Mangaluru coast in India. Unlike the mathematical modeling techniques, the soft computing tools have no complexity involved about understanding the nature of underlying process and prediction consumes less time when proper physical model data is available. The soft computing methods like artificial neural network (ANN), adaptive neuro fuzzy inference system (ANFIS), genetic algorithm based adaptive neuro fuzzy inference system (GA-ANFIS) and particle swarm based adaptive neuro fuzzy inference system (PSO-ANFIS) are the four models employed in the study. The ANN predicted well for the set architecture of (5-7-1). The ANFIS is used to predict the wave run-up on semicircular breakwater models using the hybrid efficiency of fuzzy logic and neural network. An initial FIS is generated for input variables by mapping the input-output data; the training is done using ANN; and the objective of GA and PSO is set to find the best FIS, reducing the root mean square error in the prediction of wave run-up. The most influencing input parameters (Hi/gT2, d/gT2, S/D, hs/d, R/Hi) are taken in non-dimensional form. The data required has been acquired from the physical model experiments conducted in the Marine structures laboratory of National Institute of Technology Karnataka (NITK), Surathkal, India. The GA-ANFIS prediction of wave run-up is found to be better than that of ANFIS prediction in terms of Correlation coefficient (R), Root mean square error (RMSE), Nash sutcliffe efficiency (NSE), Bias and Scatter index (SI). However, among the four models developed the ANN prediction outperformed the other three considered models with a higher R = 0.9467. © Springer Nature Singapore Pte Ltd 2021.Item Hydrological modeling of stream flow over netravathi river basin(Springer Science and Business Media Deutschland GmbH, 2021) Ashish, S.; Kundapura, S.; Kaliveeran, V.Riverine resources which are the basis of life are being transformed through urbanization. This has to be analyzed effectively in order to rejuvenate riverine ecosystems. The effects of land-use dynamics are a factor to be analyzed, and using hydrological modeling which is adopted in this study aids for the same. Soil and Water Assessment Tool (SWAT) is used as an effective tool in modeling the river basin due to its ability to quantify the alternate input data provided to the model. 14-year daily data was simulated in the model provided; the warm-up period for the model is 2 years. Coefficient of determination value of 0.74 and Nash–Sutcliffe efficiency (NSE) to be 0.71 were obtained from the analysis which indicate that the simulated values fall within a good range. The parameters which influence most are found to be curve number, available water capacity in the soil, groundwater delay, Manning’s n and plant uptake compensation. The fitted range was obtained, and this was used to increase the accuracy in SWAT Calibration and Uncertainty Procedures (SWAT-CUP). Sequential Uncertainty Fitting ver.2 (SUFI2) was found to be effective because of its uncertainty consideration criteria, and it accounts for all uncertainties that may occur in the mode. Hydrological modeling of a river basin can help us to assess the impact of alternative input data on the stream flow. © Springer Nature Singapore Pte Ltd 2021.Item Selection of Suitable General Circulation Model Outputs of Precipitation for a Humid Tropical Basin(Springer Science and Business Media Deutschland GmbH, 2022) Abraham, A.; Kundapura, S.Climate change has an observed effect on the environment and ecosystem. Climate change study aids in visualisation of changes in the environmental processes as well as development of planning and adopting the strategies. General circulation models (GCMs) are widely used in understanding present and projecting future climate change. The uncertainties associated with the climate projections is a major risk in impact assessment studies and could be reduced with considering GCMs suitable for the region. The study attempts to select a suitable subset of GCMs for precipitation simulation in the humid tropical basin, Achencoil, Kerala, India. Three statistical indicators, correlation coefficient (CC), normalised root mean square deviation (NRMSD) and absolute normalised mean bias deviation (ANMBD), are considered to evaluate the GCMs with the historical observations. The entropy technique is assigned to determine the weight of each performance indicator. Multi-criterion decision-making approaches, compromise programming (CP) and preference ranking organisation method of enrichment evaluation (PROMETHEE-2) are applied individually to rank the GCMs. The ranking is then integrated with group decision-making approach. The GCMs, MIROC-ESM-CHEM, MIROC-ESM, BCC-CSM1-1 and NorESM1-M occupied the first four positions in replicating the historical rainfall in the basin. © 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.Item Mechanical characterization of adhesive layer using Double strap joint specimens(Elsevier Ltd, 2022) Sahana, T.S.; Kaliveeran, V.; Kundapura, S.In practical applications, most of our mechanical and civil structures are subjected to either oscillatory or static load. To design those structures, stress analysis is needed to withstand the design loads. In experimental stress analysis, strain gauges are mounted on those structures to determine stress at a point. Usually, strain gauges are mounted on the substrate using proper adhesives. The quality of strain data is a function of proper mounting. The thickness of adhesives used for mounting techniques should be optimum and very thin to ensure quality results. The underlying adhesive layer acts as an interface between the substrate and the strain gauge. This adhesive layer receives the maximum shear stress when the member is loaded in tension. Therefore, in this research work, the main focus is on characterizing the Araldite adhesive material. A tension test was conducted on the prepared bulk specimen to characterize the adhesive. The Double Strap Joint (DSJ) tests were conducted to investigate the interfacial shear strength and failure modes according to ASTM D 3528 standards. Two sets of specimens, such as plain and knurled straps, were used in this research work. The failure mode in the plain strap joint shows cohesive failure, and the knurled strap joint shows adhesive failure. © 2022Item Machine Learning Ensemble Model for Flood Susceptibility Mapping(Institute of Electrical and Electronics Engineers Inc., 2022) Kundapura, S.; Soman, A.; Kuruvilla, E.Floods can be considered the most dangerous natural disaster, given their unpredictability and capacity to wipe out valuable life and property. The timely and efficient prediction of floods and flood susceptible or risk zones has been of utmost importance and can help with risk assessment, long-term management, and future preparedness. Over the years, Machine Learning (ML) has evolved as a powerful tool to build accurate flood models, inundation maps, and warning systems. This study uses the Support Vector Machine (SVM) and Random Forest (RF) algorithms and a Bagging Ensemble Model for Flood Susceptibility Mapping of Ernakulam, Kerala, India using multi-source geospatial data and the WEKA software. A total of twelve Flood Conditioning Factors (FCFs) are considered as variables, namely elevation, slope, curvature, Topographic Roughness Index (TRI), Topographic Wetness Index (TWI), Stream Power Index (SPI), rainfall, Land Use Land Cover (LULC), distance to the river, drainage density, and geology. The contribution of factors is assessed using the OneR Feature Selection method. The models are compared using the Receiver Operating Characteristics (ROC) curve and the Area Under the Curve (AUC) method. The Flood Susceptibility Map provides an opportunity for planners and authorities to flood preparation and long-term planning against flood impacts. © 2022 IEEE.Item Performance Comparison of Machine Learning Algorithms in Groundwater Potability Prediction(Institute of Electrical and Electronics Engineers Inc., 2022) Kuruvilla, E.; Kundapura, S.Rising global water demand has resulted in the overuse of groundwater resources and a decline in groundwater quality. Physical and chemical characteristics significantly impacted by geological formations and human activities determine how groundwater quality varies. An accurate and reliable assessment of groundwater resource information is the key element for effective management and enhancement of groundwater quality. The utilization of modern Machine Learning (ML) techniques in groundwater quality assessment provides insights for policymakers in suggesting remedies and management approaches for groundwater quality issues. Machine Learning models outperform other simulation models, using input and output datasets without considering the intricate relationship of the model to be analyzed and decreasing computational efforts. Comparison of various ML techniques, including Ensemble, Nonlinear, and Linear models for the prediction of groundwater potability is the main objective of this study. The presence of potable groundwater suggests that the water is fit for human consumption. The proposed approach makes use of eight ML algorithms i.e. Gradient Boosting Classifier (GB), Random Forest (RF), Decision Tree (DT), K-Nearest Neighbors (KNN), Naïve Bayes (NB), Support Vector Machine (SVM), Linear Regression (LR) and Stochastic Gradient Descent (SGD) algorithm. According to the results, the Ensemble ML models outperformed well followed by the Nonlinear models, and Linear classification ML models have comparatively less accuracy and reliability. © 2022 IEEE.