Spatial modeling of radon potential mapping using deep learning algorithms

Panahi, Mahdi; Yariyan, Peyman; Rezaie, Fatemeh; Kim, Sung Won; Sharifi, Alireza; Alesheikh, Ali Asghar; Lee, Jongchun; Kim, Seonhong; Yoo, Juhee; Lee, Saro

doi:10.1080/10106049.2021.2022011

Cited by 15 publications

(3 citation statements)

References 103 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Those model performance were statistical metrics commonly used in regression analysis and did not have any units due to its measurement to test the prediction result between the continuous regression data that valued to be the index of land subsidence susceptibility maps. The lower MSE, RMSE, MAE, and StD values indicate higher model performance [83], [84].…”

Section: ) Model Performance Evaluationmentioning

confidence: 99%

Land Subsidence and Groundwater Storage Assessment Using ICOPS, GRACE, and Susceptibility Mapping in Pekalongan, Indonesia

Hakim,

Fadhillah,

Lee

et al. 2023

IEEE Trans. Geosci. Remote Sensing

View full text Add to dashboard Cite

Floods in Pekalongan, Indonesia, often occur due to river water overflowing during heavy monsoon rain. Simultaneously, the northern coastal area of Pekalongan, located adjacent to the Java Sea, has been affected by coastal floods due to sea level rise. The flood conditions in this area were exacerbated by land subsidence, leading to coastal inundation. Monitoring land subsidence in Pekalongan has become essential in predicting other possible land subsidence occurrence areas and mitigating the possible hazards caused by land subsidence. The analysis of land subsidence has been much easier since the introduction of radar satellites. In this study, 124 synthetic aperture radar (SAR) datasets from the Sentinel-1 radar satellite between 2017 and 2022 in descending tracks were used. The data were processed through a time-series interferometry SAR (InSAR) method based on the improved combined scatterers interferometry with optimized point scatterers (ICOPS) algorithm to provide accurate measurements over large areas by improving the selection of measurement points from persistent scatterer (PS) and distributed scatterer (DS) points using a deep learning algorithm based on a convolutional neural network (CNN), and the resulting optimized measurement points were then spatially clustered using optimized hot spot analysis (OHSA) to estimate significant points statistically and define them as hot spot points. The results of time-series deformation in Pekalongan were compared with the GPS station measurements. From the comparison, a good correlation in terms of deformation patterns between time-series InSAR and GPS measurements was observed. Our study revealed that land subsidence in Pekalongan has occurred mostly in settlement areas under the young Manuscript

show abstract

Section: ) Model Performance Evaluationmentioning

confidence: 99%

Land Subsidence and Groundwater Storage Assessment Using ICOPS, GRACE, and Susceptibility Mapping in Pekalongan, Indonesia

Hakim,

Fadhillah,

Lee

et al. 2023

IEEE Trans. Geosci. Remote Sensing

View full text Add to dashboard Cite

show abstract

“…In the field of remote sensing, deep learning technology has attracted extensive attention from scholars, and many experts have utilized deep learning methods for tree species classification, achieving good classification results [10][11][12]. Particularly, convolutional neural networks (CNNs) have achieved significant success in computer vision tasks such as image classification, object detection, and semantic segmentation [13][14][15]. Due to their powerful feature extraction capabilities, CNNs have become the most commonly used neural networks in hyperspectral tree species classification [16][17][18].…”

Section: Introductionmentioning

confidence: 99%

A Hybrid convolution neural network for the classification of tree species using hyperspectral imagery

Wang,

Jiang

2024

PLoS ONE

View full text Add to dashboard Cite

In recent years, the advancement of hyperspectral remote sensing technology has greatly enhanced the detailed mapping of tree species. Nevertheless, delving deep into the significance of hyperspectral remote sensing data features for tree species recognition remains a challenging endeavor. The method of Hybrid-CS was proposed to addresses this challenge by synergizing the strengths of both deep learning and traditional learning techniques. Initially, we extract comprehensive correlation structures and spectral features. Subsequently, a hybrid approach, combining correlation-based feature selection with an optimized recursive feature elimination algorithm, identifies the most valuable feature set. We leverage the Support Vector Machine algorithm to evaluate feature importance and perform classification. Through rigorous experimentation, we evaluate the robustness of hyperspectral image-derived features and compare our method with other state-of-the-art classification methods. The results demonstrate: (1) Superior classification accuracy compared to traditional machine learning methods (e.g., SVM, RF) and advanced deep learning approaches on the tree species dataset. (2) Enhanced classification accuracy achieved by incorporating SVM and CNN information, particularly with the integration of attention mechanisms into the network architecture. Additionally, the classification performance of a two-branch network surpasses that of a single-branch network. (3) Consistent high accuracy across different proportions of training samples, indicating the stability and robustness of the method. This study underscores the potential of hyperspectral images and our proposed methodology for achieving precise tree species classification, thus holding significant promise for applications in forest resource management and monitoring.

show abstract

“…Based on part of the data, Li et al developed an ensemble learning model to estimate monthly ZCTA-level radon concentrations for ZCTAs in Greater Boston, a densely populated fraction of this study region . Multiple studies used other classic statistical learning methods to model the spatial distributions of radon concentrations in different parts of the world. ,− In this study, we developed an innovative geographical machine learning method to model the complex relationships between ZCTA-level radon concentrations and various predictors. This approach was particularly useful for studying large and heterogeneous areas, where the relationships between radon and predictors may vary significantly across space.…”

Section: Introductionmentioning

confidence: 99%

Predicting Monthly Community-Level Radon Concentrations with Spatial Random Forest in the Northeastern and Midwestern United States

Li,

Stern,

Garshick

et al. 2023

Environ. Sci. Technol.

View full text Add to dashboard Cite

In 1987, the United States Environmental Protection Agency recommended installing a mitigation system when the indoor concentration of radon, a well-known carcinogenic radioactive gas, is at or above 148 Bq/m 3 . In response, tens of millions of short-term radon measurements have been conducted in residential buildings over the past three decades either for disclosure or to initially evaluate the need for mitigation. These measurements, however, are currently underutilized to assess population radon exposure in epidemiological studies. Based on two relatively small radon surveys, Lawrence Berkeley National Laboratory developed a state-of-the-art national radon model. However, this model only provides coarse and invariant radon estimations, which limits the ability of epidemiological studies to accurately investigate the health effects of radon, particularly the effects of acute exposure. This study involved obtaining over 2.8 million historical short-term radon measurements from independent laboratories. With the use of these measurements, an innovative spatial random forest (SRF) model was developed based on geological, architectural, socioeconomical, and meteorological predictors. The model was used to estimate monthly community-level radon concentrations for ZIP Code Tabulation Areas (ZCTAs) in the northeastern and midwestern regions of the United States from 2001 to 2020. Via cross-validation, we found that our ZCTA-level predictions were highly correlated with observations. The prediction errors declined quickly as the number of radon measurements in a ZCTA increased. When ≥15 measurements existed, the mean absolute error was 24.6 Bq/m 3 , or 26.5% of the observed concentrations (R 2 = 0.70). Our study demonstrates the potential of the large amount of historical short-term radon measurements that have been obtained to accurately estimate longitudinal ZCTA-level radon exposures at unprecedented levels of resolutions and accuracy.

show abstract

Spatial modeling of radon potential mapping using deep learning algorithms

Cited by 15 publications

References 103 publications

Land Subsidence and Groundwater Storage Assessment Using ICOPS, GRACE, and Susceptibility Mapping in Pekalongan, Indonesia

Land Subsidence and Groundwater Storage Assessment Using ICOPS, GRACE, and Susceptibility Mapping in Pekalongan, Indonesia

A Hybrid convolution neural network for the classification of tree species using hyperspectral imagery

Predicting Monthly Community-Level Radon Concentrations with Spatial Random Forest in the Northeastern and Midwestern United States

Contact Info

Product

Resources

About