Random Forest Classifier Algorithm of Geographic Resources Analysis Support System Geographic Information System for Satellite Image Processing: Case Study of Bight of Sofala, Mozambique

Coasts, Год журнала: 2024, Номер 4(1), С. 127 - 149

Опубликована: Фев. 26, 2024

Mapping coastal regions is important for environmental assessment and monitoring spatio-temporal changes. Although traditional cartographic methods using a geographic information system (GIS) are applicable in image classification, machine learning (ML) present more advantageous solutions pattern-finding tasks such as the automated detection of landscape patches heterogeneous landscapes. This study aimed to discriminate patterns along eastern coasts Mozambique ML modules Geographic Resources Analysis Support System (GRASS) GIS. The random forest (RF) algorithm module ‘r.learn.train’ was used map landscapes shoreline Bight Sofala, remote sensing (RS) data at multiple temporal scales. dataset included Landsat 8-9 OLI/TIRS imagery collected dry period during 2015, 2018, 2023, which enabled evaluation dynamics. supervised classification RS rasters supported by Scikit-Learn package Python embedded GRASS Sofala characterized diverse marine ecosystems dominated swamp wetlands mangrove forests located mixed saline–fresh waters coast Mozambique. paper demonstrates advantages areas. integration Earth Observation data, processed decision tree classifier land cover characteristics recent changes ecosystem Mozambique, East Africa.

Язык: Английский

---

Development of a spatial framework for flash flood damage assessment and mitigation by coupling analytics of machine learning and household level survey data – A case study of rapid collaborative assessments and disbursement of public funds to the affectees of floods 2022, Punjab Pakistan

International Journal of Disaster Risk Reduction, Год журнала: 2024, Номер 108, С. 104463 - 104463

Опубликована: Апрель 17, 2024

Язык: Английский

---

Deep Autoencoders for Unsupervised Anomaly Detection in Wildfire Prediction

Earth and Space Science, Год журнала: 2024, Номер 11(11)

Опубликована: Ноя. 1, 2024

Abstract Wildfires pose a significantly increasing hazard to global ecosystems due the climate crisis. Due its complex nature, there is an urgent need for innovative approaches wildfire prediction, such as machine learning. This research took unique approach, differentiating from classical supervised learning, and addressed gap in unsupervised prediction using autoencoders clustering techniques anomaly detection. Historical weather normalized difference vegetation index data sets of Australia 2005–2021 were utilized. Two main analyzed. The first used deep autoencoder obtain latent features, which then fed into models, isolation forest, local outlier factor one‐class support vector machines second approach reconstruct input use reconstruction errors identify anomalies. Long Short‐Term Memory fully connected (FC) employed this part, both way learning only nominal data. FC outperformed counterparts, achieving accuracy 0.71, F1‐score 0.74, MCC 0.42. These findings highlight practicality method, it effectively predicts wildfires absence ground truth, utilizing technique.

Язык: Английский

---

Deep learning for rapid crop damage assessment after cyclones

Natural Hazards, Год журнала: 2025, Номер unknown

Опубликована: Фев. 7, 2025

Язык: Английский

---

Assessing Machine Learning Models on Temporal and Multi‐Sensor Data for Mapping Flooded Areas

Transactions in GIS, Год журнала: 2025, Номер 29(2)

Опубликована: Март 17, 2025

ABSTRACT Natural disasters, particularly floods, are escalating in frequency and intensity, disproportionately impacting economically disadvantaged populations leading to substantial economic losses. This study leverages temporal multi‐sensor data from Synthetic Aperture Radar (SAR) multispectral sensors on Sentinel satellites evaluate a range of supervised semi‐supervised machine learning (ML) models. These models, combined with feature extraction selection techniques, effectively process large datasets map flood‐affected areas. Case studies Brazil Mozambique demonstrate the efficacy methods. The Support Vector Machine (SVM) an RBF kernel, despite achieving high kappa values, tended overestimate flood extents. In contrast, Classification Regression Trees (CART) Cluster Labeling (CL) methods exhibited superior performance both qualitatively quantitatively. Gaussian Mixture Model (GMM), however, showed sensitivity input was least effective among tested. analysis highlights critical need for careful ML models preprocessing techniques mapping, facilitating rapid, data‐driven decision‐making processes.

Язык: Английский

---

Mapping groundwater-related flooding in urban coastal regions

Journal of Hydrology, Год журнала: 2025, Номер unknown, С. 132907 - 132907

Опубликована: Март 1, 2025

Язык: Английский

---

High-Resolution Flood Susceptibility Mapping and Exposure Assessment in Pakistan: An Integrated Artificial Intelligence, Machine Learning and Geospatial Framework

International Journal of Disaster Risk Reduction, Год журнала: 2025, Номер unknown, С. 105442 - 105442

Опубликована: Март 1, 2025

Язык: Английский

---

A state-of-the-art review on the quantitative and qualitative assessment of water resources using google earth engine

Environmental Monitoring and Assessment, Год журнала: 2024, Номер 196(12)

Опубликована: Ноя. 28, 2024

Язык: Английский

---

Leveraging Sentinel-2 and Geographical Information Systems in Mapping Flooded Regions around the Sesia River, Piedmont, Italy

GeoHazards, Год журнала: 2024, Номер 5(2), С. 485 - 503

Опубликована: Май 28, 2024

Sentinel-2 data are crucial in mapping flooded areas as they provide high spatial and spectral resolution but under cloud-free weather conditions. In the present study, we aimed to devise a method for area using multispectral from optical sensors Geographical Information Systems (GISs). As case selected site located Northern Italy that was heavily affected by flooding events on 3 October 2020, when Sesia River Piedmont region hit severe disturbance, heavy rainfall, strong winds. The developed thresholding technique through water indices. More specifically, Normalized Difference Water Index (NDWI) Modified (MNDWI) were chosen among most widely used methods with applications across various environments, including urban, agricultural, natural landscapes. corresponding product Copernicus Emergency Management Service (EMS) evaluate predicted our method. results showed both indices captured satisfactory level of detail. NDWI demonstrated slightly higher accuracy, where it also appeared be more sensitive separation soil vegetation cover. study findings may useful disaster management linked flooded-area rehabilitation following flood event, can valuably assist decision policy making towards sustainable environment.

Язык: Английский

---

UAV-based DEM augmentation using ConSinGAN for efficient flood parameter prediction with machine learning and 1D hydrodynamic models

Physics and Chemistry of the Earth Parts A/B/C, Год журнала: 2024, Номер 135, С. 103675 - 103675

Опубликована: Июль 14, 2024

Язык: Английский

---