Showing papers in "Geocarto International in 2022"

Assessing spatio-temporal mapping and monitoring of climatic variability using SPEI and RF machine learning models

Abstract: Abstract Droughts may inflict significant damage to agricultural and water supplies, resulting in substantial financial losses as well as the death of people and livestock. This study intends to anticipate droughts by studying the changes of an acceptable index using appropriate climatic factors. This study was divided into three phases, first being the determination of the Standardized Precipitation Evapotranspiration (SPEI) index for the Cholistan, Punjab, Pakistan area based on a dataset spanning 1980 to 2020. The indices are calculated at different monthly intervals which could to predict short-term periods for the Cholistan in Pakistan, we selected two distinctive time periods of one month (SPEI–1) and three months (SPEI–3). The second phase involved dividing the data into three sample sizes, which were used for training data from 1980 to 2010, testing data from 2011 to 2015 and validation data from 2016 to 2020. The utilization of the random forest (RF) algorithm to train and evaluate the data using a variety of climate variables e.g. potential evapotranspiration, rainfall, vapor pressure cloud cover, and mean, minimum and maximum, temperature. The final phase was to analyze the performance of the model based on statistical metrics and drought classes. Based on these considerations, statistical measures, such as the Coefficient of Determination (R2) and the Root Mean Square Error (RMSE) approach, were used to evaluate the performance of the test group throughout the testing period. The model's performance revealed the satisfactory results with R2 values of 0.80 and 0.78, for SPEI–1 and SPEI–3 situations, respectively. Following the data analysis, it was discovered that the validation period had a receiving operating curve and area under the Curve (ROC-AUC) of 0.87 for the SPEI–1 case and 0.85 for the SPEI–3 case. In this context, the results indicate that the SPEI may be useful as a prediction tool for drought prediction and the performances the RF model was suitable for both timescales. However, a more rigorous analysis with a larger dataset or a combination of datasets from different areas might be more beneficial for generalization over more extended time periods provide additional insights.

An interpretable model for the susceptibility of rainfall-induced shallow landslides based on SHAP and XGBoost

Abstract: Abstract The machine-learning “black box” models, which lack interpretability, have limited application in landslide susceptibility mapping. To interpret the black-box models, some interpretable machine learning algorithms have been proposed recently. Among them is SHaply Additive ExPlanation (SHAP), which has attracted much attention because of its ease of operation and comprehensiveness. In this study, a novel interpretable model based on SHAP and XGBoost is proposed to interpret landslides susceptibility evaluation at global and local levels. The established evaluation model provided 0.75 accuracy and 0.83 AUC value for the test sets. The global interpretation shows that the peak rainfall intensity and elevation are the dominant factors that influence the occurrence of landslides in the study area. The combination of local interpretation and field investigations can provide a comprehensive framework for evaluating designated landslides, and it can also be used as a reference for preventing and managing the hazards of landslides.

Land use/land cover and change detection mapping in Rahuri watershed area (MS), India using the google earth engine and machine learning approach

Abstract: Abstract The change detection and land use and land cover (LULC) maps are more important powerful forces behind numerous ecological systems and fallow land. The current research focuses on demarcating the spatiotemporal LULC changes, NDVI and change detections maps. These effects directly affect the ecosystem, land resources, cropping pattern and agriculture. LULC assessment and surveillance are essential for long-term planning and sustainable use of natural resources. However, we have developed the soft computing machine learning algorithm for mapping land use and land cover based on the Google earth engine (GEE) platform and change detection mapping done by SAGA GIS software. It is significantly used for ecological safety and planning under various climate variations. To accurately describe the land use and land cover classes with changes are identified in the area. This area exclusively uses the multitemporal Landsat-5 (30 m) and Sentinel-2 (10 m) imageries in LULC mapping. The GEE is a cloud-computing platform with the prevailing classification ability of random forest (RF) models to make five-year interval LULC maps for 2010, 2015 and 2020. To unique multiple RF models established as a classifier in the algorithm created by JavaScript and GEE. SAGA GIS has provided the best platform for detecting changes in land use and land cover classes. NDVI maps are created based on the cloud-based platform. These maps value ranges between −0.68 to −0.15, 0.76 to −0.29 and 0.66 to −0.11 in 2010, 2015 and 2020. Experimental outcomes indicate five classes such as water bodies, built up, barren, cropland and fallow land during 2010, 2015 and 2020. The overall accuracy of User and Producer for 2010, 2015 and 2019 years in between 86.23%, 88.34%, 85.53% and 92.51%, 94.34% and 91.54%, respectively. We have observed that (2010, 2015 − 2020) agriculture and built-up land increased by 1040.76 ha, 1246.32 ha, 1500.93 ha and 34.96 ha, 37.08 ha, 42.58 ha, respectively. Other side degraded land, fallow land, waterbodies areas (953.19 ha, 679.23 ha, 937.24 ha and 1385.73 ha, 1513.53 ha, 991.08 ha and 32.85 ha, 21.33 ha, 25.66 ha) are increased during the year of 2010, 2015 and 2020, respectively. While results have been done by GEE cloud platform and remote sensing data, this developed algorithm easily classified the land use maps from Landsat-5 and Sentinel-2 TM imagery in the machine learning approach. The determined 30-m and 10-m three-year LULC maps are made-up to deliver vital data on the changes, monitoring and understanding of which types of LULC classes and changes have occupied a place in the Rahuri area.

Assessment of PTEs in water resources by integrating HHRISK code, water quality indices, multivariate statistics, and ANNs

Abstract: Abstract The use of contaminated water for drinking and sanitary purposes can be detrimental to human health. In this article, the Human Health Risk (HHRISK) code was applied, alongside the modified heavy metal index (MHMI), synthetic pollution index (SPI), and entropy-weighted water quality index (EWQI), to investigate the pollution status, ingestion, and dermal health risks of potentially toxic elements (PTEs) (Fe, Zn, Mn, Pb, Cr, and Ni) in water resources from the Umunya area, Nigeria. Physicochemical measurements followed standard methods. Results of the MHMI, SPI, and EWQI revealed that about 60% of the water samples had low pollution and were considered suitable for human consumption, while 40% were unsuitable. Further, cumulative non-carcinogenic health risk scores indicated that 60% of the water samples pose low-medium risks while 40% pose high risks to child and adult populations. Contrarily, results of cumulative carcinogenic health risk showed that 6.67% of the samples expose water users to low risks, whereas 93.33% expose them to high risks. Although there are agreements between the results for both adult and child populations (regarding non-carcinogenic and carcinogenic risks), it is worth highlighting that the risk scores for children were higher. Therefore, children in the study area are more vulnerable to both carcinogenic and non-carcinogenic health risks. Also, it was revealed that the risk due to ingestion was higher than that due to dermal contact. Linear regression analysis showed strong agreement between the indexical models and the cumulative health risks. While artificial neural networks and multiple linear regression models accurately predicted the water quality indices, hierarchical dendrograms efficiently classed the water samples into various spatiotemporal water quality groups.

Impacts of land use/land cover patterns on groundwater quality in the Guanzhong Basin of northwest China

Abstract: Abstract Land use/land cover (LULC) significantly affects groundwater quality. This study analyzed the spatial distribution characteristics of groundwater hydrochemical parameters in the Guanzhong Basin and assessed the relationships between the groundwater parameters and LULC using statistical models and a curved streamline searchlight-shaped model (CS-SLM). The results showed that higher groundwater parameter concentrations were observed in the north of the plain than in the south. Forest and water bodies showed a positive effect on most hydrochemical parameters (Na+, Cl−, SO4 2−, F−, Cr6+), while bare land and cropland revealed a negative effect on them. In addition, the Patch-generating Land Use Simulation (PLUS) model was used to predict the LULC patterns for 2030 and to qualitatively analyze the potential changes in groundwater quality associated with LULC changes. The simulation results highlighted a significant decrease in forest areas in the southern part of Guanzhong Plain by 2030, resulting in potential groundwater quality deterioration.

Distribution, exposure, and human health risk analysis of heavy metals in drinking groundwater of Ghayen County, Iran

Abstract: Abstract The concentrations of arsenic (As), chromium (Cr), lead (Pb) and zinc (Zn) and the human health risk of the metals were estimated through drinking water consumption. Totally, 48 samples of groundwater were collected and analyzed during 2019. Health risk indices were estimated using chronic daily intake (CDI), hazard quotient (HQ) and, hazard index (HI), as well as cancer risk (CR). Except for As (66.6% of the areas were above permissible limits), the concentrations of other three heavy metals were within the US EPA and WHO permissible limits. values in 100% and 75% of the areas were above 1 for children and adults, implying non-carcinogenic risk from sum of heavy metals. Children total cancer risk (CRt) in all areas (except one area with acceptable risk) fell within unacceptable risk. Adults CRt in all areas (except one area with unacceptable risk) were in acceptable risk category. As was the highest contributor to non-carcinogenic and carcinogenic risks both for children and adults. Our results showed that children were more exposed to health risks, due to intake of drinking water containing As, Cr, and Pb in the study areas. The study can provide a basis for local governments to properly manage drinking water quality and provide a reference for water quality management in neighborhood areas.

Evaluation of groundwater quality and its suitability for drinking purposes in semi-arid region of Southern India: an application of GIS

Abstract: Abstract In the present study, the quality of groundwater for drinking purposes was evaluated. The following parameters were analyzed using American Public Health Association standard method: pH, total hardness (TH), total dissolved solids (TDS), electrical conductivity (EC), bicarbonate (HCO3 −), chloride (Cl−), sulphate (SO4 2−), fluoride (F−), calcium (Ca2+), magnesium (Mg2+), sodium (Na+) and potassium (K+). These values were compared with limits recommended by the World Health Organization (WHO) and the Bureau of Indian Standards (BIS) for drinking purposes. The cation and anion dominance of the study region groundwater samples were Na+ > Ca2+ > Mg2+ > K+ and HCO3 − > Cl− > SO4 2− > F−, respectively. Bicarbonate was the dominant anion and Na+ was identified as the dominant cation in the groundwater of the study region. The Na+ and Cl− concentration of 43% and 37% of groundwater samples were found to be more than the acceptable limit of WHO in the study region. Most of the groundwater samples in the study region were categorized as a very hard category. The groundwater was highly affected by the fluoride and about 60% of groundwater samples were unfit for drinking purposes in the study region.

Comparative study of convolutional neural network (CNN) and support vector machine (SVM) for flood susceptibility mapping: a case study at Ras Gharib, Red Sea, Egypt

Abstract: Abstract Geohazard risk is high in Arab countries due to ineffective disaster preparedness measures, mismanagement, lack of public awareness, inadequate funding and lack of stakeholder support. One such country is Egypt, which is hit by floods every year that cost lives and bring the economy to a standstill. Moreover, not much has been done to map flood-prone areas. In this paper, flood susceptibility modelling was evaluated in the Ras Gharib region of Egypt using two effective techniques machine learning technique-MLT (Support Vector Machine (SVM)) and deep learning method-DL (Convolutional Neural Networks (CNN)). Thirteen flood related factors and flood inventory layer were prepared to construct these models. Validation was performed with 30% of the flood locations where receiver operating characteristic (ROC) curves showed that the deep learning technique (CNN) gave a prediction accuracy of 86.5% (high performance), while the MLTs (SVM) gave 71.6% (medium performance). The results show that CNN provides 17% better than SVM which indicates a powerful and accurate model in flood susceptibility mapping. Results were confirmed using the Astro Digital images shortly after the 2016 flood, in which the CNN model provides a good agreement.

Furthering the precision of RUSLE soil erosion with PSInSAR data: an innovative model

Abstract: Abstract Soil erosion is a severe environmental problem worldwide, especially in tropical regions. The Revised Universal Soil Loss Equation (RUSLE), one of the universally accepted empirical soil erosion models, is quite commonly used in tropical climatic conditions to estimate the magnitude and severity of soil erosion. This study, apart from identifying the role of individual parameters in influencing the results of the RUSLE, also aims at refining the RUSLE results by incorporating the state-of-the-art technique Persistent Scatterer Interferometric Synthetic Aperture Radar (PSInSAR) in a GIS environment by utilizing its ability to measure minute surface changes in millimetre levels. Apart from this novel approach of prioritising soil erosion classes using PSInSAR, the eroding surface conditions were also studied using low coherence value (<0.75 in this study). The spatially and temporally averaged annual soil loss and net soil erosion (2015–2019), derived through RUSLE and transport limited sediment delivery (TLSD) approach, respectively, was improved by spatially integrating the PSInSAR velocity map. The integrated methodological framework is demonstrated for a tropical river basin in South India (Muvattupuzha River Basin [MRB]), which shows a mean rate of net soil loss of 6.8 ton/ha/yr, and nearly 8% of the area experiences deposition. Our approach to improve the accuracy of RUSLE-based soil erosion classes using PSInSAR techniques clearly demarcated the areas that call for utmost priority in implementing management practices. The corollary results show that the very severe soil erosion class is characterized by PSI velocity with higher negative values, followed by the successively lower classes. Results strongly suggest that RUSLE output can be improved as well as validated using a velocity map derived from radar data.

A bibliometric analysis of the landslide susceptibility research (1999–2021)

Abstract: Abstract Landslide susceptibility assessment (LSA) is a significant part of landslide research, which plays an important role in preventing landslide disasters. It has gained an increasing attention in both the academic and practice fields for the past two decades. However, there have been few bibliometric analyses on this topic, although bibliometric analysis can inspire future researchers by exploring the overall characteristics of the published literature. This article aims at collecting and analyzing the information of the abstracts, authors, institutions, countries, journals, funds, and keywords of the recent 4,732 papers published from 1999 to 2021 in Web of Science (www.webofscience.com). In particular, latent Dirichlet allocation (LDA), a machine learning and text analysis method, is utilized to analyze the abstract of each article to identify the hottest research topics related to LSA. The results revealed that: (1) The amount of annual publications related to LSA generally shows an increasing trend, which accounts for about 22% of the total landslide publications in 2021; (2) The author of Pradhan B, the institution of the Chinese Academy of Sciences, the country of China, the journal of Natural Hazards and the funding agency of the National Natural Science Foundation of China, are the productive performers in each aspect of LSA; and (3) Machine learning methods have gained a rapid increase in LSA in recent five years, which have become the most popular research topic.

Performances of MLR, RBF-NN, and MLP-NN in the evaluation and prediction of water resources quality for irrigation purposes under two modeling scenarios

Abstract: Abstract One of the pivotal decision-making tools for sustainable management of water resources for various uses is accurate prediction of water quality. In the present paper, multiple linear regression (MLR), radial basis function neural network (RBF-NN), and multilayer perceptron neural network (MLP-NN) models were developed for the monitoring and management of irrigation water quality (IWQ) in Ojoto area, southeastern Nigeria. This paper is the first to integrate and simultaneously implement these predictive methods for the modeling of seven IWQ indices. Moreover, two modeling scenarios were considered. Scenario 1 represents predictions that utilized the specific physicochemical parameters for calculating the IWQ indices as input variables while Scenario 2 represents predictions that utilized pH, EC, Na+, K+, Mg2+, Ca2+, Cl-, SO4 2-, and HCO3 - as inputs. In terms of salinity hazard, most of the water resources are unsuitable/poor for irrigation. However, in terms of carbonate and bicarbonate impact and magnesium hazard, majority of the samples have good and excellent IWQ. Seven agglomerative Q-mode dendrograms spatiotemporally classified the water resources based on the IWQ indices. Model validation metrics showed that the MLR, RBF-NN, and MLP-NN models developed in the two scenarios performed well in both scenarios, with minor variations.

Polymetallic mineralization prospectivity modelling using multi-geospatial data in logistic regression: The Diapiric Zone, Northeastern Algeria

Abstract: Abstract Prospecting and exploring minerals present major challenges in tectonically complex regions for sustainable development as in Northeastern Algeria. This area is promising for its mineral potential, especially the metallogenic province ‘The Diapiric Zone’. This study concerns mapping and predicting potential polymetallic mineralization locations by integration of remote sensing, gravity, and magnetic datasets. Several enhancement and processing methods have been applied on Landsat8_OLI and ASTER_1T remote sensed data to reduce uncertainty for achieving the best detection of hydrothermal alteration zones and lithological mapping. Furthermore, the Centre for Exploration Targeting grid analysis technique, the contact occurrence density and entropy orientation tools were employed on ground-gravity and aeromagnetic data to understand and visualize the pathways for hydrothermal fluids circulation of mineral deposits. The polymetallic mineralization prospective areas were produced using a logistic regression model on the resulting multifactor. High zones of lead-zinc cover most the area that has been confirmed by field investigation.

Structural mapping of potential field sources using BHG filter

Abstract: Abstract Edge detection is one of the most commonly used interpretation techniques applied to potential data for detecting subsurface susceptibility/density discontinuities. Several edge detection techniques utilize the first, second, or even third-order gradients of the magnetic field to delineate the susceptibility contrasting zones. A new filter called Balanced Horizontal Gradient (BHG) filter is designed for enhancing the edges of the subsurface bodies that overcomes the limitations of the conventional filters. The BHG filter is based on the arctangent function of derivatives of horizontal gradients. The main advantage of the proposed filter is that it seems to detect all the source edges with no false edges generated around the true edges and effectively balances the edges arising from the shallow and deep-seated sources. We applied the filter on the aeromagnetic data from central India and identified the extensions of the Godavari rift and central Indian shear underneath the Deccan volcanic province.

Osdes_net: oil spill detection based on efficient_shuffle network using synthetic aperture radar imagery

Abstract: Abstract Synthetic Aperture Radar (SAR) imagery can be beneficial for segmenting oil spills, which are a common environmental hazard. Oil spill detection in SAR imagery faces several challenges, including speckle noise, heterogeneous backgrounds, blurred edges, and a lack of comprehensive datasets with multiple images. ShuffleNet is one of the deep networks, which has never been used for oil spill segmentation. In this article, ShuffleNet blocks are used to detect oil spills in SAR images, which is more effective than other methods. Besides, the main network design, six other blocks were evaluated, and the most valuable one was selected. We use group convolutions, shuffle channels, and atrous convolutions in this model with a minimum number of layers of ReLU. The methods are evaluated based on the Intersection Over Union (IoU) parameter so that the proposed method improved the mIoU by 7.1% over the best results of some previous methods.

Assessment of flood susceptibility mapping using support vector machine, logistic regression and their ensemble techniques in the Belt and Road region

Abstract: Abstract Floods have occurred frequently all over the world. During 2000–2020, nearly half (44.9%) of global floods occurred in the Belt and Road region because of its complex geology, topography, and climate. Therefore, providing an insight into the spatial distribution characteristics of flood susceptibility in this region is essential. Here, a database was established with 11 flood conditioning factors, 1500 flooded points, and 1500 non-flooded points selected by an improved method. Subsequently, a rare combination of logistic regression and support vector machine, integrated by heterogeneous framework, was applied to generate an ensemble flood susceptibility map. Based on it, the concept of ecological vulnerability synthesis index in the ecological field was introduced into this study, and the flood susceptibility comprehensive index (FSCI) was proposed to quantify the degree of flood susceptibility of each country and sub-region. At the results, the ensemble model has an excellent accuracy, with the highest AUC value of 0.9342. The highest and high flood susceptibility zones are mainly located in the southeastern part of Eastern Asia, most of Southeast Asia and South Asia, account for 12.22% and 9.57% of the total study area, respectively. From the regional perspective, it can be found that Southeast Asia had the highest flood susceptibility with the highest FSCI of 4.69, while East Asia and Central and Eastern Europe showed the most significant spatial distribution characteristics. From the national perspective, of the 66 countries in this region, 20 of the countries have the highest flood susceptibility level (FSCIn > 0.8), which face the greatest threat of flooding. These results are able to facilitate reasonable flood mitigation measures develop at the most critical locations in the Belt and Road region and lays a theoretical basis for quantifying flood susceptibility at national or regional scale.

Structural interpretation of potential field data using the enhancement techniques: a case study

Abstract: Abstract The enhancement techniques of potential field data are commonly used to detect the boundary locations of geological structures. There are many different techniques for estimating the source boundaries. Through synthetic examples and Bouguer data from the southern Red Sea, we have evaluated the performance of 15 enhancement techniques. The findings show that the tilt angle of horizontal gradient (TAHG) and fast sigmoid (FSED) techniques perform better than other techniques under almost all scenarios. Moreover, these two techniques can avoid producing false structures or connected structures as other techniques. The extracted lineaments from the TAHG and FSED were compared with surface faults of the study area. As a result, major differences are caused by rifting effect on the oceanic crust. The obtained results provide valuable information to better understand the structural features of the southern Red Sea and to introduce a more reliable structural interpretation.

Novel ensemble of deep learning neural network and support vector machine for landslide susceptibility mapping in Tehri region, Garhwal Himalaya

Abstract: Abstract Over the years, landslide has become one of the most destructive events that can happen in hilly areas. Tehri, a region in the Himalayas is no different. Current research aids in the construction of ensemble models of DLNN and SVM, which are then compared with various SVM kernels. Landslide susceptibility mapping in the Tehri region of the Himalayas has been worked upon using a deep learning (DLNN), four machine learnings (SVM-RBF, SVM-Linear, SVM-Polynomial, SVM-Sigmoid), and their novel ensembles i.e., DLNN with SVM-RBF, DLNN with SVM-Linear, DLNN with SVM-Polynomial and DLNN with SVM-Sigmoid. 16 geo-environmental landslide conditioning factors (LCFs) have been considered. These models were trained using 70% of inventory landslides and tested using 30% of the same. The results revealed the superiority of DLNN, DLNN-SVM (RBF), DLNN-SVM (Linear) models which quantified 28.32, 26.96 and 22.41% of the area highly susceptible for landslide, respectively.

Flash flood susceptibility mapping in urban area using genetic algorithm and ensemble method

Abstract: Abstract Flooding is the main recurring natural disaster in Sungai Pinang catchment, Malaysia. Flash flood susceptibility mapping (FFSM) explains a key component of flood risk analysis and enables efficient estimation of the spatial extent of flood characteristics. The current study applied four machine learning models (i.e. Logistic Regression (LR), K-Nearest Neighbors (KNN), Random Forest (RF), and Extreme Gradient Boosting (XGBoost)) ensembled with the Statistical Index (SI) to develop flash flood susceptibility mapping (FFSM). 110 flash flood locations in the Sungai Pinang catchment were used in this study. Genetic algorithm (GA) was combined with Fuzzy Unordered Rules Induction Algorithm (FURIA), Rotation Forest, and Random Subspace for the feature selection method (FSM). The results showed that GA-FURIA outperformed the other two models in terms of accuracy based on the FSM. Twelve flash flood variables were selected by GA-FURIA. The FFSM results showed that the SI-RF model has the highest area under the receiver operating characteristics (AUROC) curve of success rate (0.978), whereas the SI-XGB has the best AUROC in terms of validation rate (0.997). The findings suggest that the twelve ideal conditioning variables may be used to optimize FFSM development.

Mapping of fracture zones and structural lineaments of the Gulf of Guinea passive margins using marine gravity data from CryoSat-2 and Jason-1 satellites

Abstract: Abstract Satellite gravity datasets have been widely used in understanding of the Earth's internal structure and processes. These datasets are effective in providing insights into the lithospheric structure of the Earth. Gravity data from the satellite products CryoSat-2 and Jason-1 have been used to investigate the Gulf of Guinea sedimentary basins and structural lineaments in this under-explored marginal sea region through the application of various edge filters. Present-day filtering methods such as the analytical signal, tilt angle of the gradient amplitude, NTilt gradient amplitude and softsign function were evaluated for their performance on synthetic gravity anomalies with and without noise prior to their application to gravity data covering the Gulf of Guinea. The softsign filter outputs obtained from synthetic examples result in higher resolution and more explicit edges while preventing fictitious edges production in the findings. The results of the data interpretation disclosed deep-seated and shallow structural features within the Gulf of Guinea oriented mainly in the WNW-ESE, NW-SE, NE-SW and NNE-SSW directions and seen to be concentrated in the upper slope domain. This observation is in good agreement with the occurrence of drag folds and the apparent concentration of wrench faulting features across the ridge top and bordering the upper slope. These results bring new facts ruling our understanding of this atypical regional tectonic setting in close agreement with the NE–SW-striking Pelusium Megashear Zone.

High resolution inventory and hazard assessment of potentially dangerous glacial lakes in upper Jhelum basin, Kashmir Himalaya, India

Abstract: Abstract In the Himalayan Mountain region, a large number of glacial lakes have developed in the recent past due to glacier recession under the influence of climate change. In this study, we used high resolution satellite data such as Indian Remote Sensing (IRS) LISS-IV (5.8 m) and Google earth images supplemented with field survey to generate an updated glacial lake inventory of Upper Jhelum Basin (UJB) of Kashmir Himalaya. The Sentinel-2A (10 m), Landsat-OLI (30 m) and MSS (60 m), and Cartosat-DEM (30 m) were additional data sources used for glacial lake mapping and change detection analysis. A total of 393 glacial lakes covering an area of 21.55 ± 3.8 km2 were identified, mapped and inventoried. The lake inventory includes 102 proglacial lakes, 13 supraglacial lakes and 278 unconnected glacial lakes. Using the weighted index-based method, 21 glacial lakes were found as Potentially Dangerous Glacial Lakes (PDGLs). Out of these, 7 lakes were classified as High, 9 as Medium and 5 as Low hazard glacial lakes as per the hazard assessment. Change detection analysis of PDGLs from 1980 to 2020 revealed an increase in area from 5.92 km2 to 8.46 km2 thereby, indicating a growth of 2.51 ±0.9 km2(30%) at a rate of 0.063 km2/year. The formation and growth of glacial lakes in this area is attributed to continuous glacier recession under the warming trend of temperature and declining nature of precipitation. In this study, the findings showed that Tavg and Tmin are rising significantly at a rate of 0.004ºC/year and 0.013ºC/year respectively. This study provides an important database for future GLOF studies in the region.

Flood susceptibility mapping of Northeast coastal districts of Tamil Nadu India using Multi-source Geospatial data and Machine Learning techniques

Abstract: Abstract Flooding is one of the most challenging and important natural disasters to predict, it is becoming more frequent and more intense. The study area is badly damaged by devastating flood in 2015. We assessed the flood susceptibility to northern coastal area of Tamil Nadu using various machine learning algorithms such as Gradient Boosting Machine (GBM), XGBoost (XGB), Rotation Forest (RTF), Support Vector Machine (SVM), and Naive Bayes (NB). Google Earth Engine (GEE) is used to demarcate flooded areas using Sentinel-l and other multi-source geospatial data to generate influential factors. Recursive Feature Elimination (RFE) removes weak factors in this study. The flood susceptibility resultant map is classified into five classes: very low, low, moderate, high, and very high. The GBM algorithm attained high classification accuracy with an area under the curve (AUC) value of 92%. The study area is urbanized and vulnerable identifying flood inundation useful for effective planning and implementation.

Application of novel deep boosting framework-based earthquake induced landslide hazards prediction approach in Sikkim Himalaya

Abstract: Abstract A major earthquake (6.9 Moment magnitude) occurred in the Sikkim and Darjeeling areas of the Indian Himalaya as well as in the adjacent Nepal on 18th September 2011, triggering a large number of landslides. A total of 188 landslide locations were extracted in order to create the landslide inventory map (LIM). The earthquake-induced landslide susceptibility maps (LSMs) were created using an Artificial Neural Network (ANN) model and three novel deep learning approaches (DLAs), namely Deep Boosting (DB), Deep Learning Neural Network (DLNN), and Deep Learning Tree (DLT), as well as training points and 22 conditioning factors. The earthquake-induced LSMs validated using several statistical indices and the results showed optimal accuracy for all models, where DB yielding the highest prediction rate curve (PRC) of 98.5%. This is followed by DLT (97%), DLNN (96%), and ANN (91%). The results demonstrate maximum efficacy of the proposed earthquake-induced LSM.

Soil erosion spatio-temporal exploration and Geodetection of driving factors using InVEST-sediment delivery ratio and Geodetector models in Dongsheng, China

Abstract: Abstract Assessment of soil erosion processes and accurately identifying erosion drivers is crucial for effective control and conservation strategies. We introduced a new approach involving the integration of the SDR of the InVEST and Geodetector models to explore the spatio-temporal variability of erosion and investigate the key drivers contributing to soil erosion. The results revealed an average soil loss increase of 19 to 37 t ha−1 yr−1 in the watershed from 1990 to 2020, respectively. Cropland areas had the highest growth in soil loss from 26.53 in 1990 to 118.76 t ha−1 yr−1 in 2020. The Geodetector revealed that the key variables influencing erosion are vegetation, slope, precipitation, and aspect factors. The interaction of slope and aspect, slope and precipitation, had the strongest influence on erosion than other variables, with q values of 0.80 and 0.79, respectively. The approach of this study can be used in other regions characterized by severe soil erosion.

National-scale flood risk assessment using GIS and remote sensing-based hybridized deep neural network and fuzzy analytic hierarchy process models: a case of Bangladesh

Abstract: Abstract Assessing flood risk is challenging due to complex interactions among flood susceptibility, hazard, exposure, and vulnerability parameters. This study presents a novel flood risk assessment framework by utilizing a hybridized deep neural network (DNN) and fuzzy analytic hierarchy process (AHP) models. Bangladesh was selected as a case study region, where limited studies examined flood risk at a national scale. The results exhibited that hybridized DNN and fuzzy AHP models can produce the most accurate flood risk map while comparing among 15 different models. About 20.45% of Bangladesh are at flood risk zones of moderate, high, and very high severity. The northeastern region, as well as areas adjacent to the Ganges–Brahmaputra–Meghna rivers, have high flood damage potential, where a significant number of people were affected during the 2020 flood event. The risk assessment framework developed in this study would help policymakers formulate a comprehensive flood risk management system.

Climate change enhances the positive contribution of human activities to vegetation restoration in China

Abstract: Abstract Ecological projects have huge impacts on vegetation restoration (VR) in China. However, are all of the effects due to human activities (HA) without the contribution of climate change (CC)? It is unclear what role CC plays in VR. Here, we quantitatively evaluate the contributions of climate change and human activities (CC_con and HA_con) to the net primary productivity (NPP) by the method of partial derivative and six different scenarios. HA contributed 61.19% to the NPP trend and controlled 32.84% of the NPP growth in China. Among the climate factors, temperature made the largest contribution to the NPP (45.89%) and had the largest control area (36.36%). CC promoted the positive HA_con in 54% of China, and in 68.35% of the area, the enhancement effect reached more than 50%. Our results answer a longstanding question and emphasize the important role of CC in enhancing the positive HA_con to VR.

Simulation of land use/land cover change at a basin scale using satellite data and markov chain model

Abstract: Abstract The aim of the study is to analyze the past and future land use and land cover change (LULCC) in Betwa River Basin (BRB), central India. The LULC maps were derived from Landsat satellite images using Maximum Likelihood Classifier (MLC). The artificial neural network (ANN) embedded with Land Change Modeler (LCM) was trained with driver variables. The model prediction accuracy has been accessed by evaluating Receiver Operating Characteristic (ROC) values. The study reveals that during the period 1990-2020 agriculture land, open forest, and built-up land area have increased significantly. Further, the LULCC prediction for the period 2030-2050 suggests that expansion in open forest and the built-up land area will continue while agriculture land area will keep back in the future. This research provides up-to-date LULCC information of BRB, which would be useful for all the stakeholders governing river basin management and land resource planning in this region.

Forest fire susceptibility prediction using machine learning models with resampling algorithms, Northern part of Eastern Ghat Mountain range (India)

Abstract: Abstract Periodic forest fires destruct to biodiversity, ecosystem productivity and multiple ecosystem services. Forest fires are currently turning a leading cause of forest degradation. The principal objective of this research is to predict forest fire vulnerable zones over Similipal biosphere reserve (SBR; Odisha) using different machine learning (ML) models, such as support vector machine (SVM), random forest (RF) and multivariate adaptive regression splines (MARS). Different resampling methods (CV and bootstrap) have also been applied for optimizing the result and better accuracy. Results show that 10-fold cross validation (CV) technique performed best on SVM model (AUC = 0.83) whereas bootstrap performed best on RF (AUC = 0.80) and MARS model (AUC= 0.84). The main advantage of MARS model is that it only uses input variable and significantly increases the performance of the model. The novelty of this research is application of various ML algorithms through resampling techniques to reduce the biasness and improves the reliability of the models.