We apply the Support Vector Regression (SVR) machine learning model to estimate surface roughness on a large alluvial fan of the Kosi River in the Himalayan Foreland from satellite images. To train the model, we used input features such as radar backscatter values in Vertical–Vertical (VV) and Vertical–Horizontal (VH) polarisation, incidence angle from Sentinel-1, Normalised Difference Vegetation Index (NDVI) from Sentinel-2, and surface elevation from Shuttle Radar Topographic Mission (SRTM). We generated additional features (VH/VV and VH–VV) through a linear data fusion of the existing features. For the training and validation of our model, we conducted a field campaign during 11–20 December 2019. We measured surface roughness at 78 different locations over the entire fan surface using an in-house-developed mechanical pin-profiler. We used the regression tree ensemble approach to assess the relative importance of individual input feature to predict the surface soil roughness from SVR model. We eliminated the irrelevant input features using an iterative backward elimination approach. We then performed feature sensitivity to evaluate the riskiness of the selected features. Finally, we applied the dimension reduction and scaling to minimise the data redundancy and bring them to a similar level. Based on these, we proposed five SVR methods (PCA-NS-SVR, PCA-CM-SVR, PCA-ZM-SVR, PCA-MM-SVR, and PCA-S-SVR). We trained and evaluated the performance of all variants of SVR with a 60:40 ratio using the input features and the in-situ surface roughness. We compared the performance of SVR models with six different benchmark machine learning models (i.e., Gaussian Process Regression (GPR), Generalised Regression Neural Network (GRNN), Binary Decision Tree (BDT), Bragging Ensemble Learning, Boosting Ensemble Learning, and Automated Machine Learning (AutoML)). We observed that the PCA-MM-SVR perform better with a coefficient of correlation (R = 0.74), Root Mean Square Error (RMSE = 0.16 cm), and Mean Square Error (MSE = 0.025 cm2). To ensure a fair selection of the machine learning model, we evaluated the Akaike’s Information Criterion (AIC), corrected AIC (AICc), and Bayesian Information Criterion (BIC). We observed that SVR exhibits the lowest values of AIC, corrected AIC, and BIC of all the other methods; this indicates the best goodness-of-fit. Eventually, we also compared the result of PCA-MM-SVR with the surface roughness estimated from different empirical and semi-empirical radar backscatter models. The accuracy of the PCA-MM-SVR model is better than the backscatter models. This study provides a robust approach to measure surface roughness at high spatial and temporal resolutions solely from the satellite data.
The Ganga River ecosystem is under severe anthropogenic stress. Flow regulations through structural barriers alter the geomorphic and hydraulic geometry of riverine habitats. Determining the ecological health of river habitats under the contemporary modification is detrimental to its restoration and management. This study evaluates reach averaged hydraulic habitat of the endangered Ganga River dolphin (Platanista gangetica) in a stretch between Bijnor and Narora barrages. We consider an optimal minimum flow depth as the determining factor of habitat suitability. Field measurement of the hydraulic geometry and flow characteristics show that the optimal flow depth is available in the study reach during the monsoon period, while in the premonsoon, the minimum depth is present only in the reach upstream to Narora barrage. We use a geomorphic instream flow tool (GIFT) and satellite altimeter water level data to simulate reach averaged hydraulic habitat in varying flow conditions in area upstream of Narora barrage. We observe that to maintain the minimum flow depth which supports the dolphin habitat in the study reach, an optimal discharge of about > 280 m 3 s À1 is essential. Furthermore, we develop a water-level (altimeter) and discharge (simulated from GIFT) rating curve for the study reach. It can be used to get a first-order estimate of discharge for a given water level or vice versa. This study indicates that the altimetry datasets are good precursors for estimating averaged hydraulic habitat of rivers in the data-scarce regions. The application of altimeter data can be a boon in the effective management of river habitat health over a reach scale.
<p>We coupled a hydrologic model Variable Infiltration Capacity (VIC) with the Hydrologic Engineering Center River Analysis System (HEC-RAS-2D) to model the compound impact of flood drivers in the Tapi River basin, India. Our modelling framework consists of two distinct phases; firstly, we calibrate and validate the VIC simulated daily stream flow of the Tapi River using the data observed at the Sarangkheda gauge (upstream of Ukai Reservoir) during the 2005-2012 and 2013-2016, respectively. Secondly, to simulate the high and low flow events, a separate HEC-RAS 2D model is forced with flood hydrograph (Ukai dam release) and stage hydrograph (Tidal level at Hazira) as upstream and downstream boundary conditions, respectively. We calibrated this hydrodynamic model for the 2012 flood event and validated it for the 2006 and 2014 flood events with the observed discharge and water level at the five gauges (Kakrapar Weir, Ghala, Kathor, Singanpur Weir and Nehru Bridge) located along the Tapi River in the Lower Tapi Basin (LTB). We observed that the VIC simulated daily stream flow accords well with the observed in-situ measurements. The Kling-Gupta and Nash Sutcliffe Efficiency values for calibration are 0.84 and 0.86, while, for validation, the values are 0.78 and 0.71, respectively. Furthermore, the hydrodynamic model analysis indicates satisfactory performance with the Root Mean Square Error (RMSE) for discharge and water levels in the range of 300-325 m<sup>3</sup>s<sup>-1 </sup>and 0.12&#8211;0.43 m, respectively. Finally, we prepare the flood hazard maps to provide critical insights for effective flood management and to enhance the flood resilience of the flood-prone regions of the LTB.</p>
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.