Prediction and monitoring of LULC shift using cellular automata-artificial neural network in Jumar watershed of Ranchi District, Jharkhand

Impact of wetland health and provisioning services on the livelihood of the fishing community

The present study aimed to explore the linkage between wetland health, provisioning service value (PsV) and livelihood vulnerability of the dependent fishermen community taking examples from the Moribund deltaic wetlands of India. Wetland health including hydrological strength, habitat state, and the water quality of the wetland was assessed using a random forest (RF) and XGBoosing machine learning approach, and the Livelihood Vulnerability Index (LVI) was computed using balanced weighted approach. Ordinary least square (OLS) regression, correlation matrix, and rank correlation matrix were done to make the relationship between wetland health and LVI. Wetland health was found poor (28.38%) in the isolated, smaller, and peripheral parts of the wetland with agricultural and built-up area proximity. Hydrological strength (with r2 of 0.67) was found as the most dominant health determinant followed by habitat state (with r2 of 0.62). OLS reported that in most of the cases, the standard regression residual is low (0.5 to - 0.5) which indicates that there is a strong relation between wetland health and LVI. KDE plot and correlation matrix also figured out the same. From the field survey, it was found that the wetlands with good habitat health are promising for providing more provisioning services like fish which in turn supports the livelihood of the dependent communities. The findings of this study have a deeper insight into livelihood management through wetland management. Hence, it would inspire policymakers and stakeholders to conserve wetlands not only for the sake of ecology but also for society.

Spatiotemporal variations, sources, pollution status and health risk assessment of dissolved trace elements in a major Arabian Sea draining river: insights from multivariate statistical and machine learning approaches

River Mahi drains through semi-arid regions (Western India) and is a major Arabian Sea draining river. As the principal surface water source, its water quality is important to the regional population. Therefore, the river water was sampled extensively (n = 64, 16 locations, 4 seasons and 2 years) and analyzed for 11 trace elements (TEs; Sr, V, Cu, Ni, Zn, Cd, Ba, Cr, Mn, Fe, and Co). Machine learning (ML) and multivariate statistical analysis (MVSA) were applied to investigate their possible sources, spatial-temporal-annual variations, evaluate multiple water quality parameters [heavy metal pollution index (HPI), heavy metal evaluation index (HEI)], and health indices [hazard quotient (HQ), and hazard index (THI)] associated with TEs. TE levels were higher than their corresponding world average values in 100% (Sr, V and Zn), 78%(Cu), 41%(Ni), 27%(Cr), 9%(Cd), 8%(Ba), 8%(Co), 6%(Fe), and 0%(Mn), of the samples. Three principal components (PCs) accounted for 74.5% of the TE variance: PC-1 (Fe, Co, Mn and Cu) and PC-2 (Sr and Ba) are contributed from geogenic sources, while PC-3 (Cr, Ni and Zn) are derived from geogenic and anthropogenic sources. HPI, HEI, HQ and THI all indicate that water quality is good for domestic purposes and poses little hazard. ML identified Random forest as the most suitable model for predicting HEI class (accuracy: 92%, recall: 92% and precision: 94%). Even with a limited dataset, the study underscores the potential application of ML to predictive classification modeling.

Prediction of soil erosion and sediment yield in an ungauged basin based on land use land cover changes

Soil erosion is a significant problem in the agriculture sector and the environment globally. Susceptible soil erosion zones must be identified and erosion rates evaluated to decrease land degradation problems and increase crop productivity by protecting soil fertility. Therefore, a research study has been carried out in the Ponnaniyar River basin, an ungauged tributary of the Cauvery basin in India, primarily used for agriculture. The main purpose of this study is to assess soil erosion (SE) and sediment yield (SY) for the future in an ungauged basin by utilizing the projected land use/land cover (LULC) map of the study area. Additionally, Landsat 8 satellite dataset was only used for the classification and prediction of LULC to eliminate the variation between the resolution, bands and its wavelength of different satellites datasets. To achieve the goals of this study, three phases were followed. First, the LULC of the study area was classified using a Random Trees Classifier (RTC), a machine learning technique, followed by the projection of land cover using a Cellular Automata-based Artificial Neural Network (CA-ANN) model. The driving factors for this model include digital elevation model (DEM), slope, distance to roads, settlements, and water bodies. The accuracy level of the projected LULC map was determined by comparing it with the classified LULC map of the study area, and the results showed an overall accuracy (OA) of 85.35 percentage and a kappa coefficient (K) of 0.74, respectively. Second, the projected LULC map was used in the land management factor (C) and conversation practice factor (P) of the Revised Universal Soil Loss Equation (RUSLE) model to assess soil erosion. The model was integrated with the sediment delivery ratio (SDR) to estimate sediment yield within the study area. The accuracy of the generated erosion map based on the classified and projected LULC for the year 2022 was determined using the receiver operating characteristic curve (ROC) curve, and it was found to be in satisfactory agreement. Finally, for effective soil and water conservation measures, the basin was divided into 13 sub-watersheds (SWs) using terrain analysis in geographical information system (GIS). The SWs were prioritized based on the mean soil loss in the 4-year interval from 2014 to 2030 and integrated using the weighted average method to determine the final prioritization. From these findings, SW 11, SW 9, SW 12, and SW 1 are extremely affected by soil erosion, and immediate implementation of water harvesting structures is required for soil conservation. Also, this research might be useful for decision-makers and policymakers in land management.

LULC change detection using support vector machines and cellular automata-based ANN models in Guna Tana watershed of Abay basin, Ethiopia

Recurrent changes recorded in LULC in Guna Tana watershed are a long-standing problem due to the increase in urbanization and agricultural lands. This research aims at identifying and predicting frequent changes observed using support vector machines (SVM) for supervised classification and cellular automata-based artificial neural network (CA-ANN) models for prediction in the quantum geographic information systems (QGIS) plugin MOLUSCE. Multi-temporal spatial Landsat 5 Thematic Mapper (TM) imageries, Enhanced Thematic Mapper plus 7 (ETM+), and Landsat 8 Operational Land Imager (OLI) images were used to find the acute problem the watershed is facing. Accuracy was assessed using the confusion matrix in ArcGIS 10.4 produced from ground truth data and Google Earth Pro. The results acquired from kappa statistics for 1991, 2007, and 2021 were 0.78, 0.83, and 0.88 respectively. The change detection trend indicates that urban land cover has an increasing trend throughout the entire period. In the future trend, agriculture land may shoot up to 86.79% and 86.78% of land use class in 2035 and 2049. Grassland may attenuate by 0.03% but the forest land will substantially diminish by 0.01% from 2035 to 2049. The increase of land specifically was observed in agriculture from 3128.4 to 3130 km2. Judicious planning and proper execution may resolve the water management issues incurred in the basin to secure the watershed.

Spatiotemporal change and prediction of land use in Manasi region based on deep learning

The Manasi region is located in an arid and semi-arid region with fragile ecology and scarce resources. The land use change prediction is important for the management and optimization of land resources. We utilized Sankey diagram, dynamic degree of land use, and landscape indices to explore the temporal and spatial variation of land use and integrated the LSTM and MLP algorithms to predict land use prediction. The MLP-LSTM prediction model retains the spatiotemporal information of land use data to the greatest extent and extracts the spatiotemporal variation characteristics of each grid through a training set. Results showed that (1) from 1990 to 2020, cropland, tree cover, water bodies, and urban areas in the Manasi region increased by 855.3465 km2, 271.7136 km2, 40.0104 km2, and 109.2483 km2, respectively, whereas grassland and bare land decreased by 677.7243 km2 and 598.5945 km2, respectively; (2) Kappa coefficients reflect the accuracy of the mode's predictions in terms of quantity. The Kappa coefficients of the land use data predicted by the MLP-LSTM, MLP-ANN, LR, and CA-Markov models were calculated to be 95.58%, 93.36%, 89.48%, and 85.35%, respectively. It can be found that the MLP-LSTM and MLP-ANN models obtain higher accuracy in most levels, while the CA-Markov model has the lowest accuracy. (3) The landscape indices can reflect the spatial configuration characteristics of landscape (land use types), and evaluating the prediction results of land use models using landscape indices can reflect the prediction accuracy of the models in terms of spatial features. The results indicate that the model predicted by MLP-LSTM model conforms to the development trend of land use from 1990 to 2020 in terms of spatial features. This gives a basis for the study of the Manasi region to formulate relevant land use development and rationally allocate land resources.