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, Journal Year: 2024, Volume and Issue: 4(1), P. 127 - 149

Published: Feb. 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.

Language: Английский

---

Accurate Wheat Yield Prediction Using Machine Learning and Climate-NDVI Data Fusion

IEEE Access, Journal Year: 2024, Volume and Issue: 12, P. 40947 - 40961

Published: Jan. 1, 2024

Due to exponential population growth, climate change, and an increasing demand for food, there is unprecedented need a timely, precise, dependable assessment of crop yield on large scale. Wheat, staple worldwide, requires accurate prompt prediction its output global food security. Traditionally, the development empirical models forecasting has relied data, satellite or combination both. Despite enhanced performance achieved by integrating contributions from various sources (Climate, Soil, Socioeconomic, Remote sensing) remain unclear. The lack well-defined comparisons between regression-based approaches different Machine Learning (ML) methods in necessitates further investigation. This study addresses gaps combining data multiple forecast wheat Multan region Punjab province Pakistan. findings are compared benchmark provided Crop Report Services (CRS) Punjab, with three widely used ML techniques (support vector machine (SVM), Random Forest (RF), Least Absolute Shrinkage Selection Operator (LASSO)) publicly available within GEE (Google Earth Engine) platform, including climate, satellite, soil properties, spatial information develop alternative using 2017 2022, selecting best attribute subset related output. set district-level simulated yields was analyzed Machin (SVM, RF, LASSO) as function seasonal weather, soil. results indicate that all datasets algorithms achieves better ( R ² : 0.74-0.88). Incorporating other properties into can improve 0.08 0.12. forest outperformed competitor Root Mean Square Error (RMSE) 0.05 q/ha 0.88. Comparative analysis shows random 97% SVM 93% yielded area.

Language: Английский

---

Deep Learning Methods of Satellite Image Processing for Monitoring of Flood Dynamics in the Ganges Delta, Bangladesh

Water, Journal Year: 2024, Volume and Issue: 16(8), P. 1141 - 1141

Published: April 17, 2024

Mapping spatial data is essential for the monitoring of flooded areas, prognosis hazards and prevention flood risks. The Ganges River Delta, Bangladesh, world’s largest river delta prone to floods that impact social–natural systems through losses lives damage infrastructure landscapes. Millions people living in this region are vulnerable repetitive due exposure, high susceptibility low resilience. Cumulative effects monsoon climate, rainfall, tropical cyclones hydrogeologic setting Delta increase probability floods. While engineering methods mitigation include practical solutions (technical construction dams, bridges hydraulic drains), regulation traffic land planning support systems, geoinformation rely on modelling remote sensing (RS) evaluate dynamics hazards. Geoinformation indispensable mapping catchments areas visualization affected regions real-time monitoring, addition implementing developing emergency plans vulnerability assessment warning supported by RS data. In regard, study used monitor southern segment Delta. Multispectral Landsat 8-9 OLI/TIRS satellite images were evaluated (March) post-flood (November) periods analysis extent landscape changes. Deep Learning (DL) algorithms GRASS GIS modules qualitative quantitative as advanced image processing. results constitute a series maps based classified

Language: Английский

---

Landscape Fragmentation and Deforestation in Sierra Leone, West Africa, Analysed Using Satellite Images

Transylvanian Review of Systematical and Ecological Research, Journal Year: 2024, Volume and Issue: 26(1), P. 13 - 26

Published: April 1, 2024

Abstract Monitoring rainforests in West Africa is necessary for natural resource management. Remote sensing valuable mapping tropical ecosystems and evaluation of landscape heterogeneity. This study presents analysis Sierra Leone which affects wildlife habitats biodiversity. Methods include modules “r.mapcalc”, “r.li.mps”, “r.li.edgedensity”, “r.forestfrag” GRASS GIS satellite image processing by computation mean patch size, edge density index fragmentation with six levels: exterior, patch, transitional, edge, perforated, interior. The results demonstrate increased deforestation over a 10-year period (2013 to 2023).

Language: Английский

---

Support Vector Machine Algorithm for Mapping Land Cover Dynamics in Senegal, West Africa, Using Earth Observation Data

Earth, Journal Year: 2024, Volume and Issue: 5(3), P. 420 - 462

Published: Sept. 6, 2024

This paper addresses the problem of mapping land cover types in Senegal and recognition vegetation systems Saloum River Delta on satellite images. Multi-seasonal landscape dynamics were analyzed using Landsat 8-9 OLI/TIRS images from 2015 to 2023. Two image classification methods compared, their performance was evaluated GRASS GIS software (version 8.4.0, creator: Development Team, original location: Champaign, Illinois, USA, currently multinational project) by means unsupervised k-means clustering algorithm supervised Support Vector Machine (SVM) algorithm. The identified machine learning (ML)-based analysis spectral reflectance multispectral results based processed indicated a decrease savannas, an increase croplands agricultural lands, decline forests, changes coastal wetlands, including mangroves with high biodiversity. practical aim is describe novel method creating maps RS data for each class improve accuracy. We accomplish this calculating areas occupied 10 classes within target area six consecutive years. Our indicate that, comparing algorithms, SVM approach increased accuracy, 98% pixels being stable, which shows qualitative improvements classification. contributes natural resource management environmental monitoring Senegal, West Africa, through advanced cartographic applied remote sensing Earth observation data.

Language: Английский

---

Automation of image processing through ML algorithms of GRASS GIS using embedded Scikit-Learn library of Python

Examples and Counterexamples, Journal Year: 2025, Volume and Issue: 7, P. 100180 - 100180

Published: Feb. 3, 2025

Language: Английский

---

Time Series Analysis of Landsat Images for Monitoring Flooded Areas in the Inner Niger Delta, Mali

Artificial Satellites, Journal Year: 2023, Volume and Issue: 58(4), P. 278 - 313

Published: Dec. 1, 2023

Abstract This paper presents an R-based approach to mapping dynamics of the flooded areas in Inner Niger Delta (IND), Mali, using time series analysis Landsat 8–9 satellite images. As largest inland wetland West Africa, habitats IND offers high potential for biodiversity flood-dependent eco systems. is one most productive Africa. Mapping based on images enables provide strategies land management and rice planting modelling vegetation types IND. Our libraries R programming language processing six images, each image was taken November from 2013 2022. By capturing spatial temporal structures 2013, 2015, 2018, 2020, 2021 2022, remote sensing data are combined yield estimates landscape that temporally coherent, while helping analyse fluctuations extent fluvial wetlands caused by hydrological processes seasonal flooding. Further, allowing packages support processing, NDVI, SAVI EVI indices visualising changes distribution different cover classes over realised. In this context, Earth observation advanced scripting tools provides new insights into complex interlace climate-hydrological responses. study contributes sustainable natural resources improving knowledge functioning ecosystems

Language: Английский

---

Use of Optical and Radar Imagery for Crop Type Classification in Africa: A Review

Sensors, Journal Year: 2024, Volume and Issue: 24(11), P. 3618 - 3618

Published: June 3, 2024

Multi-source remote sensing-derived information on crops contributes significantly to agricultural monitoring, assessment, and management. In Africa, some challenges (i.e., small-scale farming practices associated with diverse crop types system complexity, cloud coverage during the growing season) can imped monitoring using multi-source sensing. The combination of optical sensing synthetic aperture radar (SAR) data has emerged as an opportune strategy for improving precision reliability type mapping monitoring. This work aims conduct extensive review in Africa great detail well current research progress based Radar satellites. this context may provide high spatial resolution detailed spectral information, which allows differentiation different their signatures. However, satellites important contributions given ability technology penetrate cover, particularly African tropical regions, opposed data. explores various techniques employed integrate SAR classification applicability limitations countries. Furthermore, are discussed combination, such availability, sensor compatibility, need accurate ground truth model training validation. study also highlights potential advanced modelling machine learning algorithms, support vector machines, random forests, convolutional neural networks) accuracy automation combined Finally, concludes future directions recommendations utilizing systems. it emphasizes importance developing robust scalable models that accommodate diversity types, practices, environmental conditions prevalent Africa. Through utilization technologies, informed decisions be made sustainable strengthen nutritional security, contribute socioeconomic development continent.

Language: Английский

---

Investigating land cover changes and their impact on land surface temperature in Tay Ninh province, Vietnam

Environmental Monitoring and Assessment, Journal Year: 2024, Volume and Issue: 197(1)

Published: Dec. 20, 2024

Language: Английский

---

Image Segmentation of the Sudd Wetlands in South Sudan for Environmental Analytics by GRASS GIS Scripts

Analytics, Journal Year: 2023, Volume and Issue: 2(3), P. 745 - 780

Published: Sept. 21, 2023

This paper presents the object detection algorithms GRASS GIS applied for Landsat 8-9 OLI/TIRS data. The study area includes Sudd wetlands located in South Sudan. describes a programming method automated processing of satellite images environmental analytics, applying scripting GIS. documents how land cover changed and developed over time Sudan with varying climate settings, indicating variations landscape patterns. A set modules was used to process by language. It streamlines geospatial tasks. functionality image is called within scripts as subprocesses which automate operations. cutting-edge tools present cost-effective solution remote sensing data modelling analysis. based on discrimination spectral reflectance pixels raster scenes. Scripting syntax are run from terminal, enabling pass commands module. ensures automation high speed processing. algorithm challenge that patterns differ substantially, there nonlinear dynamics types due factors effects. Time series analysis several multispectral demonstrated changes Sudd, affected degradation landscapes. map generated each 2015 2023 using 481 maximum-likelihood discriminant approaches classification. methodology segmentation ‘i.segment’ module, clustering classification ‘i.cluster’ ‘i.maxlike’ modules, accuracy assessment ‘r.kappa’ computing NDVI cartographic mapping implemented benefits techniques reported effects various threshold levels segmentation. performed 371 times 90% minsize = 5; converged 37 41 iterations. following segments defined images: 4515 2015, 4813 2016, 4114 2017, 5090 2018, 6021 2019, 3187 2020, 2445 2022, 5181 2023. percent convergence 98% processed images. Detecting possible spaceborne datasets advanced applications algorithms. implications approach discussed. wrapper functions

Language: Английский

---