Divyesh Manohar Varade
@iitjammu.ac.in
Assistant Professor, Department of Civil Engineering
Indian Institute of Technology Jammu
EDUCATION
Ph.D.
MSc
B.E.
RESEARCH INTERESTS
Remote Sensing, Geoinformatics, GIS, Cryosphere, Hydrology, Disaster Management
Scopus Publications
Scholar Citations
Scholar h-index
Scholar i10-index
Scopus Publications
Abdul Basir Mahmoodzada, Pragyan Das, Divyesh Varade, Mohd Arslaan Akhtar, and Sawahiko Shimada
Springer Science and Business Media LLC
Bhagyashree Chopade, Vikas Gupta, and Divyesh Varade
Wiley
Deepak Rawat, M. L. Sharma, Divyesh Varade, Roshan Kumar, Debi Prasanna Kanungo, Rayees Ahmed, S. C. Gupta, Hemant Singh, and Nishant Saxena
Springer Science and Business Media LLC
Abdul Basir Mahmoodzada, Divyesh Varade, Sawahiko Shimada, Hiromu Okazawa, Shafiqullah Aryan, Gulbuddin Gulab, Abd El-Zaher M. A. Mustafa, Humaira Rizwana, Yogesh K. Ahlawat, and Hosam O. Elansary
MDPI AG
In this study, we propose quantifying the Amu River riverbank erosion with the modelled river discharge in Kaldar District, Balkh Province of Afghanistan from 2004 to 2020. We propose a framework synergizing multi-source information for modelling the erosion area based on three components: (1) river discharge, (2) river width, and (3) erosion area. The total river discharge for the watershed shared by Afghanistan and Tajikistan was modelled using hydrological parameters from the European Centre for Medium-Range Weather Forecasts (ECMWF) Reanalysis v5 (ERA5) data through multivariate linear regression with ground station data. The river width was determined manually using the Normalized Difference Water Index (NDWI) derived from Landsat data. The riverbank erosion area was derived from the digital shoreline analysis using the NDWI. The digital shoreline analysis showed that, between 2008 and 2020, the average riverbank erosion area in Kaldar District is about 5.4 km2 per year, and, overall, 86.3 km2 during 2004–2020 due to flood events. The significantly higher land loss events occurred at 10 km2 bank erosion during the years 2008–2009 and 2015–2016, and 19 km2 peak erosion occurred during 2011–2012. A linear relation between the erosion area with respect to the discharge intensity and the specific stream power was observed with an R2 of 0.84 and RMSE of 1.761 for both.
Divyesh Varade, Hemant Singh, Abhinav Pratap Singh, and Shubham Awasthi
Springer Science and Business Media LLC
Hemant Singh, Divyesh Varade, Maximillian Van Wyk de Vries, Kirtan Adhikari, Manish Rawat, Shubham Awasthi, and Deepak Rawat
Elsevier BV
Shubham Awasthi, Divyesh Varade, Sutapa Bhattacharjee, Hemant Singh, Sana Shahab, and Kamal Jain
MDPI AG
Land deformation has become a crucial threat in recent decades, caused by various natural and anthropogenic activities in the environment. The seismic land dynamics, landslides activities, heavy rainfall resulting in flood events, and subsurface aquifer shrinkage due to the excessive extraction of groundwater are among the major reasons for land deformation, which may cause serious damage to the overall land surface, civil infrastructure, underground tunnels, and pipelines, etc. This study focuses on preparing a framework for estimating land deformation and analyzing the causes associated with land deformation. A time-series SAR Interferometry-based technique called PsInSAR was used to measure land deformation, using Sentinel-1 datasets from 2015 to 2021 by estimating land deformation velocities for this region. The obtained PSInSAR deformation velocity results ranged between −4 mm to +2 mm per year. Further, land use land cover (LULC) changes in the area were analyzed as an essential indicator and probable cause of land deformation. LULC products were first generated using Landsat-8 images for two time periods (2015, 2021), which were then evaluated in accordance with the deformation analysis. The results indicated an increase in the built-up areas and agricultural cover in the region at the cost of shrinkage in the vegetated lands, which are highly correlated with the land subsidence in the region, probably due to the over-extraction of groundwater. Further, the outer region of the study area consisting of undulating terrain and steep slopes also coincides with the estimated high subsidence zones, which could be related to higher instances of landslides identified in those areas from various primary and secondary information collected. One of the causes of landslides and soil erosion in the region is identified to be high-level precipitation events that loosen the surface soil that flows through the steep slopes. Furthermore, the study region lying in a high seismic zone with characteristic unstable slopes are more susceptible to land deformation due to high seismic activities. The approach developed in the study could be an useful tool for constant monitoring and estimation of land deformation and analysis of the associated causes which can be easily applied to any other region.
Shubham Awasthi, Kamal Jain, Sutapa Bhattacharjee, Vivek Gupta, Divyesh Varade, Hemant Singh, Avadh Bihari Narayan, and Alessandra Budillon
Elsevier BV
Shubham Awasthi, Divyesh Varade, Praveen Kumar Thakur, Ajeet Kumar, Hemant Singh, Kamal Jain, and Snehmani
Elsevier BV
Abdul Basir Mahmoodzada, Divyesh Varade, Sawahiko Shimada, Farid Ahmad Rezazada, Abdul Saboor Mahmoodzada, Abdul Nasir Jawher, and Mohammadi Toghyan
Elsevier BV
Divyesh Varade, Surendar Manickam, and Gulab Singh
Wiley
Shubham Awasthi and Divyesh Varade
Informa UK Limited
ABSTRACT Seasonal alpine snow contributes significantly to the water resource. It plays a crucial role in regulating the environmental feedback and from the perspective of socio-economic sustainability in the alpine regions. While most nations are pursuing renewable energy sources, hydropower generated from snowmelt runoff is one of the primary sources. Additionally, alpine regions with snow cover are major tourist destinations that are often affected by natural disasters such as avalanches. The snowmelt runoff and early avalanche warning require timely information on the spatio-temporal aspects of the snow geophysical parameters. In this regard, advances in remote sensing of snow have been observed to be significant. Recent developments in remote sensing technology in the visible, infrared, and microwave spectrum have significantly improved our understanding of snow geophysical processes. This paper provides a review concerning the qualitative and quantitative studies of alpine snow. The electromagnetic characteristics of the alpine snow are largely dependent upon its inherent geophysical structure and the properties of the snow. Snow behaves differently with respect to the wavelength of the incident radiation. In this paper, we provide a categorical review of the remote sensing techniques for estimating the snow geophysical properties, inclusive of permittivity, density, and wetness corresponding to the wavelength used in the remotely sensed data: (1) visible-infrared spectrum including multispectral/hyperspectral, (2) active and passive microwave spectrums. We also discuss the recent advancements in the remote sensing techniques for approximating the volumetric snowpack parameters such as the snow depth and the snow water equivalent based on active and passive microwave remote sensing. This review further discusses the limitations of the techniques reviewed and future prospects for the retrieval of snow geophysical parameters (SGP) corresponding to the recent progress in remote sensing technology. In summary, the recent advances have laid down a foundation for rigorous assessment of seasonal snow using spaceborne remote sensing, particularly at a regional scale. Yet, the scope for improvements in the methods and payload design exists.
Divyesh Varade, Ajay K. Maurya, and Onkar Dikshit
Informa UK Limited
Abstract Information on the spatial and temporal extent of snow cover distribution is a significant input in hydrological processes and climate models. Although hyperspectral remote sensing provides significant opportunities in the assessment of land cover, the applications of such data are limited in the snow-covered alpine regions. A major issue with hyperspectral data is the larger dimensionality. Feature selection methods are often used to derive the most informative subset of bands from the hyperspectral data. In this study, a band selection technique is proposed which utilizes the mutual information (MI) between hyperspectral bands and a reference band. The first principal component of the hyperspectral data is selected as the reference band. Two variants of this approach are proposed involving preclustering of bands using: (1) the k-means and (2) the fuzzy k-means algorithms. The MI is derived from weighted entropy of the hyperspectral band and the reference band. The weights are computed from the cluster distance ratio and the cluster membership function for the k-means and fuzzy k-means algorithm, respectively. The selected bands were classified using random forest classifier. The proposed methods are evaluated with four datasets, two Hyperion datasets corresponding to the geographical locations of Dhundi and Solang in India, corresponding to snow covered terrain and two benchmark AVIRIS datasets of Indian Pines and Salinas. The average classification accuracy (0.995 and 0.721 for Dhundi and Solang datasets, respectively) for the proposed approach were observed to be better as compared with those from other state of the art techniques.
Ajay K. Maurya, Divyesh M. Varade, and Onkar Dikshit
Informa UK Limited
Seasonal dynamics of snow cover is an essential area of research for hydrological modelling and water resource management. With the increased availability of remote sensing data, the timely information of the spatiotemporal distribution of snow cover is feasible at regular intervals. The primary objective of this study is to assess the effect of pansharpening in the accuracy of snow cover change detection in mountainous regions using freely available Landsat-8 multispectral data. In mountainous regions at the medium resolution, the changes at the mountain ridges are seldom identified. The incorporation of pansharpening in the change detection framework facilitates an improvement in the snow cover change detection at the ridges. For pansharpening, the PanNet architecture based on convolutional neural networks was adopted. A study area around Dhundi in the state of Himachal Pradesh in India was selected for the analysis. The experiments were carried out using a subset of Landsat-8 multispectral data acquired in the autumn and the winter seasons of 2017 and 2018, respectively. An improvement of 0.184 and 0.267 in the kappa coefficient was observed for the overall changes in the snow cover and at the ridges, respectively, based on the results from the proposed approach.
Abdul Basir Mahmoodzada, Divyesh Varade, and Sawahiko Shimada
MDPI AG
The Pamir ranges of the Hindu Kush regions in Afghanistan play a substantial role in regulating the water resources for the Middle Eastern countries. Particularly, the snowmelt runoff in the Khanabad watershed is one of the critical drivers for the Amu River, since it is a primary source of available water in several Middle Eastern countries in the off monsoon season. The purpose of this study is to devise strategies based on active microwave remote sensing for the monitoring of snow depth during the winter and the melt season. For the estimation of snow depth, we utilized a multi-temporal C-band (5.405 GHz) Sentinel-1 dual polarimetric synthetic aperture radar (SAR) with a differential interferometric SAR (DInSAR)-based framework. In the proposed approach, the estimated snowpack displacements in the vertical transmit-vertical receive (VV) and vertical transmit-horizonal receive (VH) channels were improved by incorporating modeled information of snow permittivity, and the scale was enhanced by utilizing snow depth information from the available ground stations. Two seasonal datasets were considered for the experiments corresponding to peak winter season (February 2019) and early melt season (March 2019). The results were validated with the available nearest field measurements. A good correlation determined by the coefficient of determination of 0.82 and 0.57, with root mean square errors of 2.33 and 1.44 m, for the peak winter and the early melt season, respectively, was observed between the snow depth estimates and the field measurements. Further, the snow depth estimates from the proposed approach were observed to be significantly better than the DInSAR displacements based on the correlation with respect to the field measurements.
Divyesh Varade and Onkar Dikshit
Informa UK Limited
Abstract In this study, we propose a modified thresholds method for the determination of land surface emissivity (LSE) for snow covered mountainous areas. The conventional Normalized Differenced Vegetation Index (NDVI) thresholds method (NDVITHM) does not discriminate the snow covered pixels with soil pixels in assigning the LSE based on NDVI thresholds. In the proposed approach, we incorporate different thresholding rules based on the Normalized Differenced Snow Index and the S3 index for incorporating separability in the LSE for the snow covered pixels. The LSE thus derived is used to determine the land surface temperature using the Single Channel Method. The approach was evaluated for a study area around the Kullu Valley in the lower Indian Himalayas for a dataset of the winter season of Landsat-8 multispectral data. The observed coefficient of determination values indicated that the proposed method yielded better results with respect to the conventional NDVITHM approach.
Divyesh Varade, Ajay K. Maurya, Onkar Dikshit, Gulab Singh, and Surendar Manickam
Informa UK Limited
ABSTRACT The studies on snow depth comprise a crucial area of research in the Indian Himalayas, where the seasonal snow cover primarily drives the rivers and significant water resources. In this paper, the initial estimates of the line of sight displacement obtained using differential interferometric phase in VV and VH polarizations using Sentinel-1 bi-temporal dual polarimetric SAR data corresponding to snow covered and snow free land cover, are improved by applying bias corrections for the snow phase and for residual errors in displacement derived from the corrected snow phase. The bias for the snow phase is computed from the observed phase in VV and VH polarizations for the snow free area and the bias for the residual errors is computed by observing the stationary pixels identified in the snow free area using a digital elevation model. The snow depth is computed as a weighted sum of the corrected displacements in the VV and VH polarization, with the weights derived using the local incidence angle. The snow depth results based on the proposed approach was evaluated with respect to field measurements and a coefficient of determination of 0.628 was observed with an improvement of ~0.4 as compared to the displacement observed in the VV and VH channel using the conventional method.
Divyesh Varade, Surendar Manickam, Onkar Dikshit, Gulab Singh, and Snehmani
Elsevier BV
Divyesh Varade, Gulab Singh, Onkar Dikshit, and Surendar Manickam
American Geophysical Union (AGU)
Recently, with the extensive availability of fully polarimetric synthetic aperture radar (SAR) data, methods that are simple and efficient, and involve lesser computation and data processing, are needed to be explored for snow cover mapping. This paper analyzes different polarimetric parameters such as entropy, anisotropy, and the mean scattering angle for the identification of snow cover area. We present a novel index for mapping snow cover based on the assessment of entropy (H) and anisotropy (A) using fully polarimetric SAR data and refer to it as the radar snow fraction (RSF). The RSF is proposed as an extension of the H(1‐A) metric by applying a sigmoidal function to this metric. The experiments to evaluate the applicability of the proposed RSF are carried out using fully polarimetric SAR data of L‐band ALOS‐2/PALSAR‐2 and C‐band RADARSAT‐2 data sets corresponding to different geographical locations in the Indian Himalayas. The developed snow cover maps from the proposed method were validated with respect to reference snow cover maps derived by thresholding the Normalized Differenced Snow Index developed from multispectral data (e.g., Landsat‐8 imagery). These maps were also statistically compared with those obtained from the conventional radar snow index, which is based on the polarization fraction. We determined a mean overall accuracy of 0.8 between the developed snow cover maps and the reference maps for the different data sets used for experiments. The results showed that, in general, the RSF outperformed the other polarimetric parameters for snow cover detection.
Divyesh Varade and Onkar Dikshit
Informa UK Limited
ABSTRACT Snow geophysical parameters such as wetness, density and permittivity are a significant input in hydrological models and water resource management. In this paper, we utilize the triangle method based on a feature space developed with the near-infrared (NIR) reflectance and the Normalized Differenced Snow Index (NDSI) for the estimation of surface snow wetness, permittivity and density. The triangular feature space based on NIR reflectance and NDSI is parameterized to yield a linear relationship between the snow wetness and the NIR reflectance. Snow density and permittivity are derived based on the least squares solution of empirical relations based on the observations of surface snow wetness. The proposed methodology was evaluated using Sentinel-2 data, and the modeled snow geophysical parameters were validated with respect to field measurements. Based on the results, it was inferred that the NIR reflectance varies linearly with the liquid water content in the snow. A good agreement was determined between the modeled and measured parameters for wet snow conditions as observed by the coefficient of determination of 0.968, 0.521 and 0.969 for the snow wetness, density and permittivity (real part), respectively. The proposed approach can be significantly utilized with unmanned aerial sensors for monitoring of physical properties of fresh or wet snow and is thus expected to contribute considerably in hydrological applications and avalanche studies.
Divyesh Varade, Anudeep Sure, and Onkar Dikshit
Informa UK Limited
Abstract This study proposes the development of a multi-sensor, multi-spectral composite from Landsat-8 and Sentinel-2A imagery referred to as ‘LSC’ for land use land cover (LULC) characterisation and compared with respect to the hyperspectral imagery of the EO1: Hyperion sensor. A three-stage evaluation was implemented based on the similarity observed in the spectral response, supervised classification results and endmember abundance information obtained using linear spectral unmixing. The study was conducted for two areas located around Dhundi and Rohtak in Himachal Pradesh and Haryana, respectively. According to the analysis of the spectral reflectance curves, the spectral response of the LSC is capable of identifying major LULC classes. The kappa accuracy of 0.85 and 0.66 was observed for the classification results from LSC and Hyperion data for Dhundi and Rohtak datasets, respectively. The coefficient of determination was found to be above 0.9 for the LULC classes in both the datasets as compared to Hyperion, indicating a good agreement. Thus, these three-stage results indicated the significant potential of a composite derived from freely available multi-sensor multi-spectral imagery as an alternative to hyperspectral imagery for LULC studies.
Divyesh Varade, Onkar Dikshit, Surendar Manickam, Gulab Singh, and Snehmani
IEEE
The snow water equivalent is a significant parameter in the determination of the hydrological potential of snow, which requires timely information on snow density and snow depth. For shallow snow depth, the C-band polarimetric synthetic aperture radar (SAR) data can be effectively utilized for the estimation of snow density. Although methods exist in the literature for the estimation of snow density, they are largely based on the exploitation of fully polarimetric SAR, which is mostly commercial and highly expensive. In this paper, we assess the capability of the Sentinel-1 polarimetric SAR data for determination of snow density using two bi-temporal dataset pairs based on a novel framework. The method is based on the derivation of incremental/decremental factors of the scattering matrix, which are related to the Fresnel transmission coefficients (FTC) and the attenuation constant. The snow density is determined using the permittivity derived from the inversion of the FTC. The experiments are also performed using RADARSAT2 datasets synchronized with field measurements. The results are observed to be in agreement with respect to the available in-situ measurements as determined by the coefficient of determination of 0.658 and 0.525 for the RADARSAT-2 and Sentinel-1 datasets, respectively.
Divyesh Varade, Ajay K. Maurya, and Onkar Dikshit
Informa UK Limited
ABSTRACT In this paper, we focus on utilizing the image denoising method for ranking of significant bands in hyperspectral imagery. We make use of the fact that the denoising error of bands varies with the significant information content of the bands in hyperspectral imagery. The denoising error is computed for each band individually and compared using a matching parameter with the denoising error of a reference image. The reference image is selected to be the first principal component corresponding to the maximum information. Three matching parameters including mutual information (MI), correlation coefficient (r) and the structural similarity index (SSIM) were used for ranking the bands based on the match with the denoising error of the reference image. The proposed algorithm is tested using three datasets, namely, Indian Pines, Salinas and Dhundi. The Indian Pines and Salinas datasets were acquired from the Airborne Visible Infrared Imaging Spectrometer (AVIRIS) sensor and comprised rural and agricultural area. The Dhundi dataset of Hyperion comprises mostly of features corresponding to snow-covered mountainous regions. To assess the accuracy of the proposed method, a supervised classification was carried out using a random forest classifier with 20% training pixels selected randomly from the ground reference. The proposed method yielded significantly better results determined by the kappa coefficient (κ) of 0.756, 0.910 and 0.996 for the Indian Pines, Salinas and Dhundi datasets, respectively, over several other state of the art methods. The classification results of the proposed method also yielded better results than those obtained by the state-of-the-art methods for hyperspectral band selection.
Divyesh M. Varade, Ajay K. Maurya, and Onkar Dikshit
Informa UK Limited
ABSTRACT Spectral indexes (SI) are widely used for land cover characterization and also in several physical models for the study of land surface processes. For example, the normalized differenced vegetation index (NDVI) is used in the characterization of soil moisture along with shortwave infrared reflectance. However, for hyperspectral imagery (HSI) comprising many bands within a single spectrum, it is significant to identify the optimal bands for the development of SI. In this paper, we study the potential of band selection in specific bandwidths for the determination of SI. The proposed methodology includes two strategies for development of SI: direct SI determined by the best band within specific spectrums and fused SI determined by fusion of two best bands within specific spectrums. The experiments are conducted using three datasets, two corresponding to snow-covered areas, studied using the normalized differenced snow index (NDSI) and one comprising the agricultural area, studied using NDVI. The developed SI are evaluated through a comparison with the supervised classification maps from the corresponding HSI. A kappa coefficient of 0.693, 0.726 and 0.803 was observed between the results obtained from histogram slicing of SI with respect to the classification maps for the three datasets, respectively.
Divyesh Varade, Onkar Dikshit, and Surendar Manickam
Springer Science and Business Media LLC