Applying Data Analysis and Machine Learning Methods to Predict Permafrost Coast Erosion DOI Creative Commons
Daria Bogatova, Stanislav Ogorodov

Geosciences, Год журнала: 2024, Номер 15(1), С. 2 - 2

Опубликована: Дек. 26, 2024

This study aims to establish a scientific and methodological basis for predicting shoreline positions using modern data analysis machine learning techniques. The focus area is 5 km section of the Ural coast along Baydaratskaya Bay in Kara Sea. region was selected due its diverse geomorphological features, varied lithological composition, significant presence permafrost processes, all contributing complex patterns change. Applying advanced methods, including correlation factor analysis, enables identification natural signs that highlight areas active coastal retreat. These insights are valuable arctic development planning, as they help recognize zones at highest risk transformation. erosion process can be conceptualized comprising two primary components construct predictive model first random variable encapsulates effects local structural changes coastline alongside fluctuations climatic conditions. component statistically characterized define confidence interval variability. second represents systematic shift, which reflects regular average over time. more suited modeling. Thus, information processing methods allow us move from descriptive numerical assessments dynamics processes. goal ultimately support responsible sustainable highly sensitive region.

Язык: Английский

Formalization for Subsequent Computer Processing of Kara Sea Coastline Data DOI Creative Commons
Daria Bogatova, Stanislav Ogorodov

Data, Год журнала: 2024, Номер 9(12), С. 145 - 145

Опубликована: Дек. 9, 2024

This study aimed to develop a methodological framework for predicting shoreline dynamics using machine learning techniques, focusing on analyzing generalized data without distinguishing areas with higher or lower retreat rates. Three sites along the southwestern Kara Sea coast were selected this investigation. The analyzed key coastal features, including lithology, permafrost, and geomorphology, combination of field studies remote sensing data. Essential datasets compiled formatted computer-based analysis. These included information permafrost geomorphological characteristics zone, climatic factors influencing shoreline, measurements bluff top positions rates over defined time periods. tops determined through imagery varying resolutions measurements. A novel aspect involved employing geostatistical methods analyze erosion rates, providing new insights into dynamics. analysis allowed us identify experiencing most significant changes. By continually refining neural network models these datasets, we can improve our understanding complex interactions between natural evolution, ultimately aiding in developing effective management strategies.

Язык: Английский

Процитировано

1

A Novel Method to Detect Short-Term Erosion and Deposition Within the Intertidal Zones and its Application to the Large Radial Sand Ridges in China DOI
Zhipeng Sun, Xiaojing Niu

IEEE Geoscience and Remote Sensing Letters, Год журнала: 2023, Номер 21, С. 1 - 5

Опубликована: Ноя. 21, 2023

This study proposes a novel method to detect short-term erosion and deposition within the intertidal zones based solely on waterlines extracted from high-resolution satellite images. The judgment of or is intersection waterlines, which can reflect topographic changes. procedures have been demonstrated by its application large Radial Sand Ridges in China, where are drastic due strong tidal currents. were divided into four elevation zones, Sentinel-2 images used obtain spatial distribution at each sub-zone. results show that detected greatly linked with creeks. had further for preliminary risk assessment wind power structures area. It found southern more possible risk. an inexpensive way preliminarily monitor zones.

Язык: Английский

Процитировано

1

Applying Data Analysis and Machine Learning Methods to Predict Permafrost Coast Erosion DOI Creative Commons
Daria Bogatova, Stanislav Ogorodov

Geosciences, Год журнала: 2024, Номер 15(1), С. 2 - 2

Опубликована: Дек. 26, 2024

This study aims to establish a scientific and methodological basis for predicting shoreline positions using modern data analysis machine learning techniques. The focus area is 5 km section of the Ural coast along Baydaratskaya Bay in Kara Sea. region was selected due its diverse geomorphological features, varied lithological composition, significant presence permafrost processes, all contributing complex patterns change. Applying advanced methods, including correlation factor analysis, enables identification natural signs that highlight areas active coastal retreat. These insights are valuable arctic development planning, as they help recognize zones at highest risk transformation. erosion process can be conceptualized comprising two primary components construct predictive model first random variable encapsulates effects local structural changes coastline alongside fluctuations climatic conditions. component statistically characterized define confidence interval variability. second represents systematic shift, which reflects regular average over time. more suited modeling. Thus, information processing methods allow us move from descriptive numerical assessments dynamics processes. goal ultimately support responsible sustainable highly sensitive region.

Язык: Английский

Процитировано

0