Spatial Downscaling of GRACE Data Based on XGBoost Model for Improved Understanding of Hydrological Droughts in the Indus Basin Irrigation System (IBIS)

Remote Sensing, Journal Year: 2023, Volume and Issue: 15(4), P. 873 - 873

Published: Feb. 4, 2023

Climate change may cause severe hydrological droughts, leading to water shortages which will require be assessed using high-resolution data. Gravity Recovery and Experiment (GRACE) satellite Terrestrial Water Storage (TWSA) estimates offer a promising solution monitor drought, but its coarse resolution (1°) limits applications small regions of the Indus Basin Irrigation System (IBIS). Here we employed machine learning models such as Extreme Gradient Boosting (XGBoost) Artificial Neural Network (ANN) downscale GRACE TWSA from 1° 0.25°. The findings revealed that XGBoost model outperformed ANN with Nash Sutcliff Efficiency (NSE) (0.99), Pearson correlation (R) Root Mean Square Error (RMSE) (5.22 mm), Absolute (MAE) (2.75 mm) between predicted GRACE-derived TWSA. Further, Deficit Index (WSDI) WSD (Water Deficit) were used determine severity episodes respectively. results WSDI exhibited strong agreement when compared Standardized Precipitation Evapotranspiration (SPEI) at different time scales (1-, 3-, 6-months) self-calibrated Palmer Drought Severity (sc-PDSI). Moreover, IBIS had experienced increasing drought episodes, e.g., eight detected within years 2010 2016 −1.20 −1.28 total −496.99 mm −734.01 mm, Partial Least Regression (PLSR) climatic variables indicated potential evaporation largest influence on after precipitation. this study helpful for drought-related decision-making in IBIS.

Language: Английский

---

Prediction of flash flood susceptibility using integrating analytic hierarchy process (AHP) and frequency ratio (FR) algorithms

Frontiers in Environmental Science, Journal Year: 2023, Volume and Issue: 10

Published: Jan. 5, 2023

The landscape of Pakistan is vulnerable to flood and periodically affected by floods different magnitudes. aim this study was aimed assess the flash susceptibility district Jhelum, Punjab, using geospatial model Frequency Ratio Analytical Hierarchy Process. Also, considered eight most influential flood-causing parameters are Digital Elevation Model, slop, distance from river, drainage density, Land use/Land cover, geology, soil resistivity (soil consisting rocks formation) rainfall deviation. data collected weather stations in vicinity area. Estimated weight allotted each flood-inducing factors with help AHP FR. Through use overlay analysis, were brought together, value density awarded maximum possible score. According several areas region based on have been classified zones viz, very high risk, moderate low risk. In light results obtained, 4% area that accounts for 86.25 km 2 at risk flood. like Bagham, Sohawa, Domeli, Turkai, Jogi Tillas, Chang Wala, Dandot Khewra located elevation. Whereas Potha, Samothi, Chaklana, Bagrian, Tilla Jogian, Nandna, Rawal high-risk damaged badly history This first its kind conducted Jhelum District provides guidelines disaster management authorities response agencies, infrastructure planners, watershed management, climatologists.

Language: Английский

---

Modelling, mapping and monitoring of forest cover changes, using support vector machine, kernel logistic regression and naive bayes tree models with optical remote sensing data

Heliyon, Journal Year: 2023, Volume and Issue: 9(2), P. e13212 - e13212

Published: Jan. 26, 2023

The present study is designed to monitor the spatio-temporal changes in forest cover using Remote Sensing (RS) and Geographic Information system (GIS) techniques from 1990 2017. Landsat data (Thematic mapper [TM]), 2000 2010 (Enhanced Thematic Mapper [ETM+]), 2013 2017 (Operational Land Imager/Thermal Infrared Sensor [OLI/TIRS]) were classified into classes termed snow, water, barren land, built-up area, forest, vegetation. method was built multitemporal images machine learning Support Vector Machine (SVM), Naive Bayes Tree (NBT) Kernel Logistic Regression (KLR). According results, area decreased 19,360 km2 (26.0%) 18,784 (25.2%) 2010, while increased 18,640 (25.0%) 26,765 (35.9%) due "One billion tree Project". our findings, SVM performed better than KLR NBT on all three accuracy metrics (recall, precision, accuracy) F1 score >0.89. demonstrated that concurrent reforestation land areas improved methods of sustaining RS GIS everyday forestry organization practices Khyber Pakhtun Khwa (KPK), Pakistan. results beneficial, especially at decision-making level for local or provincial government KPK understanding global scenario regional planning.

Language: Английский

---

Identification of time-varying wetlands neglected in Pakistan through remote sensing techniques

Environmental Science and Pollution Research, Journal Year: 2023, Volume and Issue: 30(29), P. 74031 - 74044

Published: May 18, 2023

Language: Английский

---

Wetland identification through remote sensing: Insights into wetness, greenness, turbidity, temperature, and changing landscapes

Big Data Research, Journal Year: 2023, Volume and Issue: 35, P. 100416 - 100416

Published: Nov. 9, 2023

Language: Английский

---

A hybrid of ensemble machine learning models with RFE and Boruta wrapper-based algorithms for flash flood susceptibility assessment

International Journal of Applied Earth Observation and Geoinformation, Journal Year: 2023, Volume and Issue: 122, P. 103401 - 103401

Published: July 14, 2023

Flash floods are among the world most destructive natural disasters, and developing optimum hybrid Machine Learning (ML) models for flash flood susceptibility (FFS) modeling remains a challenge. This study proposed novel intelligence algorithms based on of several ensemble ML (i.e., Bagged Flexible Discriminant Analysis (BAFDA), Extreme Gradient Boosting (XBG), Rotation Forest (ROF) Boosted Generalized Additive Model (BGAM)) wrapper-based factor optimization Recursive Feature Elimination (RFE) Boruta) to improve accuracy FFS mapping at Neka-Haraz watershed in Iran. In addition, Random Search (RS) method is meta-optimization developed hyper-parameters. considers 20 conditioning factors (CgFs) 380 non-flood locations create geospatial database. The performance each model was evaluated by area under receiver operating characteristic (ROC) curve (AUC) validation methods, such as efficiency. demonstrated good performance, with BGAM-Boruta achieving highest (AUC = 0.953, Efficiency 0.910), followed ROF-Boruta 0.952), ROF-RFE 0.951), BAFDA-Boruta 0.950), BGAM-RFE ROF 0.949), BGAM 0.948), BAFDA-RFE 0.943), XGB-Boruta BAFDA 0.939), XGB-RFE 0.938) XGB 0.911). model, regional coverage about 46% high very areas. Moreover, revealed that distance river, slope, rainfall, altitude, road CgFs significant this region.

Language: Английский

---

Remote sensing based forest cover classification using machine learning

Scientific Reports, Journal Year: 2024, Volume and Issue: 14(1)

Published: Jan. 2, 2024

Abstract Pakistan falls significantly below the recommended forest coverage level of 20 to 30 percent total area, with less than 6 its land under cover. This deficiency is primarily attributed illicit deforestation for wood and charcoal, coupled a failure embrace advanced techniques estimation, monitoring, supervision. Remote sensing leveraging Sentinel-2 satellite images were employed. Both single-layer stacked temporal layer from various dates utilized classification. The application an artificial neural network (ANN) supervised classification algorithm yielded notable results. Using image Sentinel-2, impressive 91.37% training overall accuracy 0.865 kappa coefficient achieved, along 93.77% testing 0.902 coefficient. Furthermore, approach demonstrated even better method 98.07% accuracy, 97.75% coefficients 0.970 0.965, respectively. random (RF) algorithm, when applied, achieved 99.12% 92.90% 0.986 0.882. Notably, satellite, RF reached exceptional performance 99.79% 96.98% validation 0.996 0.954. In terms cover ANN identified 31.07% in District Abbottabad region. comparison, recorded slightly higher 31.17% forested area. research highlights potential remote machine learning algorithms improving assessment monitoring strategies.

Language: Английский

---

Flood susceptibility mapping contributes to disaster risk reduction: A case study in Sindh, Pakistan

International Journal of Disaster Risk Reduction, Journal Year: 2024, Volume and Issue: 108, P. 104503 - 104503

Published: April 23, 2024

Floods are a widespread and damaging natural phenomenon that causes harm to human lives, resources, property has agricultural, eco-environmental, economic impacts. Therefore, it is crucial perform flood susceptibility mapping (FSM) identify susceptible zones mitigate reduce damage. This study assessed the damage caused by 2022 flash in Sindh identified flood-susceptible based on frequency ratio (FR) analytical hierarchy process (AHP) models. Flood inventory maps were generated, containing 150 sampling points, which manually selected from Landsat imagery. The points split into 70% for training 30% validating results. Furthermore, fourteen conditioning factors considered analysis developing FSM. final FSM categorized five zones, representing levels very low high. results areas under high Ghotki (FR 4.42% AHP 5.66%), Dadu 21.40% 21.29%), Sanghar 6.81% 6.78%). Ultimately, accuracy was evaluated using receiver operating characteristics area curve method, resulting 82%, 83%), 91%, 90%), 96%, 95%). enhances scientific understanding of impacts across diverse regions emphasizes importance accurate informed decision-making. findings provide valuable insights supportive policymakers, agricultural planners, stakeholders engaged risk management adverse consequences floods.

Language: Английский

---

Optimizing flood susceptibility assessment in semi-arid regions using ensemble algorithms: a case study of Moroccan High Atlas

Natural Hazards, Journal Year: 2024, Volume and Issue: 120(8), P. 7787 - 7816

Published: March 21, 2024

Abstract This study explores and compares the predictive capabilities of various ensemble algorithms, including SVM, KNN, RF, XGBoost, ANN, DT, LR, for assessing flood susceptibility (FS) in Houz plain Moroccan High Atlas. The inventory map past flooding was prepared using binary data from 2012 events, where “1” indicates a flood-prone area “0” non-flood-prone or extremely low area, with 762 indicating areas. 15 different categorical factors were determined selected based on importance multicollinearity tests, slope, elevation, Normalized Difference Vegetation Index, Terrain Ruggedness Stream Power Land Use Cover, curvature plane, profile, aspect, flow accumulation, Topographic Position soil type, Hydrologic Soil Group, distance river rainfall. Predicted FS maps Tensift watershed show that, only 10.75% mean surface predicted as very high risk, 19% 38% estimated respectively. Similarly, Haouz plain, exhibited an average 21.76% very-high-risk zones, 18.88% 18.18% low- very-low-risk zones applied algorithms met validation standards, under curve 0.93 0.91 learning stages, Model performance analysis identified XGBoost model best algorithm zone mapping. provides effective decision-support tools land-use planning risk reduction, across globe at semi-arid regions.

Language: Английский

---

Mapping and monitoring of spatio-temporal land use and land cover changes and relationship with normalized satellite indices and driving factors

Geology Ecology and Landscapes, Journal Year: 2023, Volume and Issue: unknown, P. 1 - 17

Published: March 6, 2023

Climate change has become a severe threat all around the world. Pakistan is also affected by climate change. It problem in any part of country. land degradation caused climatic disturbance and anthropogenic activities. In this study, we used Landsat TM for years 1980 1990, ETM+ 2000 2010, OLI/TIRS year 2020 were classified using Maximum Likelihood Classification (MLC) into built-up area, barren land, vegetation, water area. This study been conducted to assess spatio-temporal analysis Land Use/Land Cover (LULC) with driving factors from 1980–2020 Cholistan Desert, Punjab, Pakistan, relationship between different normalized satellite indices LULC. Post-classification detection methods then variation over period. The area increased 6.25% 1980–2020. Kappa coefficient was estimated at 0.83, 0.82, 0.85, 0.88 1980, 2000, 2020, respectively. At annual scale (the 2020), significant positive observed among LULC, NDVI, MNDWI regions located western region, while negative trends 2010 year.

Language: Английский

---