Conference Papers
Permanent URI for this collectionhttps://idr.nitk.ac.in/handle/123456789/28506
Browse
7 results
Search Results
Item Spatio-temporal Analysis and Modeling of Coastal areas for Water Salinity Prediction(Institute of Electrical and Electronics Engineers Inc., 2023) Sudhakara, B.; Priyadarshini, R.; Bhattacharjee, S.; Kamath S․, S.; Umesh, P.; Gangadharan, K.V.; Ghosh, S.K.Salinity is an important parameter affecting the quality of water, and excessive amounts adversely affect vege-tation growth and aquatic organism populations. Natural factors like tidal waves, natural calamities etc., and man-made factors like unchecked disposal of industrial wastes, domestic/ urban sewage, and fish hatchery activities can cause significant increases in water salinity. In this article, an approach that utilizes multimodal data like Landsat 8 optical observations and the SMAP salinity data product for predicting water salinity indices in the coastal region is proposed. Machine Learning models such as K-nearest neighbor (KNN), Gradient Boost (GB), Extremely Randomized Tree (ERT), Random Forest Regression (RFR), Decision Tree (DT), Multiple Linear Regression (MLR), Lasso Regression (LR), and Ridge Regression (RR) are used for salinity prediction. Empirical experiments revealed that the ERT model outperformed other ML models, with a R2 of 0.92 and RMSE of 0.25 psu. © 2023 IEEE.Item Water Salinity Assessment Using Remotely Sensed Images—A Comprehensive Survey(Springer Science and Business Media Deutschland GmbH, 2023) Priyadarshini, R.; Sudhakara, B.; Kamath S․, S.; Bhattacharjee, S.; Umesh, P.; Gangadharan, K.V.In the past few years, the problem of growing salinity in river estuaries has directly impacted living and health conditions, as well as agricultural activities globally, especially for those rivers which are the sources of daily water consumption for the surrounding community. Key contributing factors include hazardous industrial wastes, residential and urban wastewater, fish hatchery, hospital sewage, and high tidal levels. Conventional survey and sampling-based approaches for water quality assessment are often difficult to undertake on a large-scale basis and are also labor and cost-intensive. On the other hand, remote sensing-based techniques can be a good alternative to cost-prohibitive traditional practices. In this article, an attempt is made to comprehensively assess various approaches, datasets, and models for determining water salinity using remote sensing-based approaches and in situ observations. Our work revealed that remote sensing techniques coupled with other techniques for estimating the salinity of water offer a clear advantage over traditional practices and also is very cost-effective. We also highlight several observations and gaps that can be beneficial for the research community to contribute further in this significant research domain. © 2023, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.Item SOIL MOISTURE PREDICTION USING MULTI-SOURCE DATASETS AND GRAPH NEURAL NETWORK(Institute of Electrical and Electronics Engineers Inc., 2024) Sudhakara, B.; Bhattacharjee, S.In this study, we introduce a novel method for predicting soil moisture (SM) in the Bidar district of Karnataka by leveraging a combination of advanced Machine Learning (ML) techniques and remote sensing data. The study utilizes multi-source inputs, such as MODIS satellite data, soil properties, elevation, and precipitation, to predict the SM. Our methodology implements state-of-the-art graph neural networks (GNNs), particularly Graph Convolutional Networks (GCNs), in combination with Long Short-Term Memory (LSTM) networks to capture both spatio-temporal dependencies in the data. The proposed model is compared with LSTM and CNN (Convolutional Neural Network)LSTM models and experimental results demonstrate that the GCN-LSTM model outperforms other approaches, achieving an R2 value of 0.9152, a mean absolute error (MAE) of 0.0240 m3/m3, and a root mean square error (RMSE) of 0.0322 m3/m3. These findings highlight the potential of graph neural networks in enhancing the accuracy of SM predictions, providing valuable insights for agricultural and water resource management in the region. © 2024 IEEE.Item Prediction of High-Resolution Soil Moisture Using Multi-source Data and Machine Learning(Springer Science and Business Media Deutschland GmbH, 2024) Sudhakara, B.; Bhattacharjee, S.Soil moisture (SM) stands as a critical meteorological element influencing the dynamic interplay between the land and the atmosphere. Its comprehension, modeling, and examination hold key significance in unraveling this interaction. Information about the surface SM is necessary for predicting crop yield, future disasters, etc. Ground-based SM measurement is accurate but time-consuming and costly. An alternate approach for measuring SM using satellite images is becoming more popular in recent years. Surface SM retrieval with a fine-resolution still poses challenges. The proposed work considers multi-satellite data for predicting high-resolution SM of Oklahoma, USA using multiple Machine Learning (ML) algorithms, such as K-nearest neighbour (KNN), Decision tree (DT), Random forest (RF), and Extra trees regressor (ETR). A high-resolution SM map for the study region is also reported, considering the Soil Moisture Active Passive (SMAP) SM data as the base, Landsat 8 bands, and normalized difference vegetation index (NDVI) data as the reference datasets. The ETR model performed the best with a mean absolute error (MAE) of 0.940 mm, a root mean square error (RMSE) of 1.303 mm and a coefficient of determination (R2 ) of 0.965. The external validation is carried out with ground-based SM data from the International Soil Moisture Network (ISMN). Both the actual SMAP SM and predicted SM values demonstrate a comparable correlation with the ISMN data. © 2024, The Author(s), under exclusive license to Springer Nature Switzerland AG.Item Drought Detection in India using Spatio-Temporal Graph Neural Networks(Institute of Electrical and Electronics Engineers Inc., 2025) Sudhakara, B.; Maheshwari, A.; Palakuru, S.; Bhattacharjee, S.Droughts, which are defined by extended periods of water scarcity, offer significant difficulties to agriculture, ecosystems, and human populations. Drought detection that requires timely and precise assessment is critical for effective mitigation and resource planning. This work proposes a novel technique for drought detection using satellite imagery with the capabilities of Graph Neural Networks (GNNs). The proposed GNN-based model captures spatio-temporal dependencies by representing 671 districts across India as nodes, connected based on geographical proximity. The spatio-temporal model achieved its best performance with an RMSE of 6.849, MAE of 4.367, and R2 of 0.903 for the Normalized Vegetation Supply Water Index (NVSWI). This work is one of the initial attempts to predict the drought over the Indian region using graph neural networks. © 2025 IEEE.Item High-resolution Soil Moisture Prediction from SMOS using Machine Learning Models(Institute of Electrical and Electronics Engineers Inc., 2025) Sudhakara, B.; Maheshwari, A.; Periasamy, M.; Bhattacharjee, S.Soil moisture is essential for the land carbon cycle, surface and groundwater circulation, heat transport, energy exchange between these systems and other processes. SMOS's (Soil Moisture and Ocean Salinity) 36-kilometer spatial resolution and 3-day temporal resolution offer valuable insights into soil moisture dynamics. This research paper introduces an innovative approach to enhance our understanding and prediction of SMOS values by applying advanced machine learning models. Our research focuses on developing and implementing advanced downscaling techniques, leveraging advanced machine learning algorithms. The primary objective is to establish a robust framework for estimating soil moisture levels at multiple geographic locations within the study region of Oklahoma, USA. To achieve this, three years of SMOS (Soil Moisture and Ocean Salinity) data was integrated with remotely captured images spanning the full range of the electromagnetic spectrum, from visible to infrared wavelengths. The LSTM model performed significantly better in predicting soil moisture values with 0.041 RMSE (m3/m3) and 0.869 (R2) than the other models. © 2025 IEEE.Item Predicting High-Resolution Soil Moisture Using MODIS Bands: A Fusion-Regression Approach(Institute of Electrical and Electronics Engineers Inc., 2025) Sudhakara, B.; Pais, S.M.; Bhattacharjee, S.Soil moisture (SM) is a crucial biophysical parameter that plays a significant role in predicting food security. Monitoring SM is essential as it provides valuable insights into agricultural productivity and the impacts of climate change. Remote sensing platforms retrieve SM data from various satellites, enabling global monitoring. However, the coarse resolution of these datasets limits their ability to capture fine-scale variations in SM. To address this challenge, this study aims to generate a high-resolution SM product using data from the Soil Moisture Active Passive (SMAP) mission. The SMAP data is fused with MODIS spectral bands to create a multi-source dataset, which is then used as input for different regression models to achieve downscaling of SM. Experimental results demonstrate that the Extremely Randomized Trees (ERT) model achieves the minimum error, with RMSE 0.0113 m3/m3, MAE 0.0172 m3/m3, and R2 0.9725. The study focuses on Karnataka, covering the temporal window from 2015 to 2022. © 2025 IEEE.