Predicting of dust storm source by combining remote sensing, statistic-based predictive models and game theory in the Sistan watershed, southwestern Asia

Boroughani, Mahdi; Pourhashemi, Sima; Gholami, Hamid; Kaskaoutis, Dimitris G.

doi:10.1007/s40333-021-0023-3

Cited by 22 publications

(11 citation statements)

References 79 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For building the predictive DL models of the dust sources, an inventory map 10 —a map showing current sources of dust—is needed. To obtain it, we consulted the inventory of dust sources provided by Boroughani et al 15 and the World Bank 27 . Predictive DL models were then constructed by dividing datasets randomly into train (70% or 222 samples) and test procedures (30% or 99 samples) (Fig.…”

Section: Methodsmentioning

confidence: 99%

“…Therefore, among variables suggested to represent dust source controls, identifying the influential or important variables is the key to accurate spatial modeling. Unfortunately, there is a general dearth of literature on the use of feature selection algorithms in the aeolian realm 14 , 15 . One study applied the leave-one-feature-out (LOFO) algorithm for identifying the influential variables controlling dust sources in central Asia 10 .…”

Section: Introductionmentioning

confidence: 99%

“…In another work 16 , three variables comprising precipitation, digital elevation model, and soil organic carbon were identified as the key variables controlling dust emissions in southeastern Iran by the Boruta algorithm. Soil, geomorphology, and slope were identified as the three critical variables controlling dust sources in the Sistan watershed (Iran-Afghanistan borders) by two statistical-based predictive models, including frequency ratios (FR) and weights of evidence (WOE) 15 . In the present study, we applied the Dragonfly optimization (FDO) algorithm (a technique less applied to feature selection) to identify influential variables controlling dust provenance in the study area.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Novel deep learning hybrid models (CNN-GRU and DLDL-RF) for the susceptibility classification of dust sources in the Middle East: a global source

Gholami

Mohammadifar

2022

Sci Rep

Self Cite

View full text Add to dashboard Cite

Dust storms have many negative consequences, and affect all kinds of ecosystems, as well as climate and weather conditions. Therefore, classification of dust storm sources into different susceptibility categories can help us mitigate its negative effects. This study aimed to classify the susceptibility of dust sources in the Middle East (ME) by developing two novel deep learning (DL) hybrid models based on the convolutional neural network–gated recurrent unit (CNN-GRU) model, and the dense layer deep learning–random forest (DLDL-RF) model. The Dragonfly algorithm (DA) was used to identify the critical features controlling dust sources. Game theory was used for the interpretability of the DL model’s output. Predictive DL models were constructed by dividing datasets randomly into train (70%) and test (30%) groups, six statistical indicators being then applied to assess the DL hybrid model performance for both datasets (train and test). Among 13 potential features (or variables) controlling dust sources, seven variables were selected as important and six as non-important by DA, respectively. Based on the DLDL-RF hybrid model – a model with higher accuracy in comparison with CNN-GRU–23.1, 22.8, and 22.2% of the study area were classified as being of very low, low and moderate susceptibility, whereas 20.2 and 11.7% of the area were classified as representing high and very high susceptibility classes, respectively. Among seven important features selected by DA, clay content, silt content, and precipitation were identified as the three most important by game theory through permutation values. Overall, DL hybrid models were found to be efficient methods for prediction purposes on large spatial scales with no or incomplete datasets from ground-based measurements.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Novel deep learning hybrid models (CNN-GRU and DLDL-RF) for the susceptibility classification of dust sources in the Middle East: a global source

Gholami

Mohammadifar

2022

Sci Rep

Self Cite

View full text Add to dashboard Cite

show abstract

“…We used MODIS images from the Terra (morning) and Aqua (afternoon) satellites (Vickery and Eckardt, 2013) to identify dust aerosols. We define dusty days, when the horizontal visibility is less than 2000 m for at least one hour during the day based on available weather stations in Iran (Vickery and Eckardt, 2013;Boroughani et al, 2021). According to the mentioned condition, more than 500 dusty days were identified during 2010-2021 distributed over the stations in Birjand, Zahedan, Kerman, Bam, Doostabad, Bisheh, Rafsanjan and Mighan.…”

Section: Dust Aerosol Mapmentioning

confidence: 99%

Mapping land degradation risk due to wind and water erosion

Boroughani

Mirchooli

Hadavifar

et al. 2023

Preprint

View full text Add to dashboard Cite

Abstract. Land degradation is a cause of many social, economic, and environmental problems. Therefore identification and monitoring of high-risk areas for land degradation are necessary. Despite the importance of land degradation, the topic receives often relatively little attention. The present study aims to create a land degradation map in terms of soil erosion caused by wind and water erosion of semi-dry land. We focus on the Lut watershed in Iran encompassing the Lut Desert that is influenced by both monsoon rainfalls and dust storms. Dust sources are identified using MODIS satellite images with the help of four different indices to quantify uncertainty. The dust source maps are assessed with three machine learning algorithms encompassing artificial neural network (ANN), random forest (RF), and flexible discriminant analysis (FDA) to map dust sources paired with soil erosion susceptibility due to water. We assess the accuracy of the maps from the machine learning results with the metric Area Under the Curve (AUC) of the Receiver Operating Characteristic (ROC). The maps for water and aeolian soil erosion are used to identify different classes of land degradation risks. The results show that 43 % of the watershed is prone to land degradation in terms of both aeolian and water erosion. Most regions (45 %) have a risk of water erosion and some regions (7 %) a risk of aeolian erosion. Only a small fraction (4 %) of the total area of the region had a low to very low susceptibility for land degradation. The results of this study underline the risk of land degradation for an inhabited region in Iran. Future work should focus on land degradation associated with soil erosion from water and storms in larger regions to evaluate the risks also elsewhere.

show abstract

“…Game theory can effectively increase the rationality of indicator weights. The combination of game theory and GIS can make the evaluation results more reasonable and intuitive, and has been well applied in the risk assessment of water erosion hazards [14] and dust storm hazards [15][16][17].…”

Section: Introductionmentioning

confidence: 99%

Windblown sand hazards risk assessment along the highways based on GIS-game theory combination weight

Li,

Lv,

Tao

et al. 2024

PLoS ONE

View full text Add to dashboard Cite

Windblown sand hazards seriously threaten the safe operation of highways in desert areas. Reasonable risk assessment can provide the basis for windblown sand hazards prevention and risk reduction. To facilitate the formulation of better windblown sand hazards prevention and reduction strategies, a new windblown sand hazards risk assessment model along the highways was proposed, in which seven evaluation indicators were selected from danger of the hazard-causing factors, vulnerability of the hazard-forming environment, and the vulnerability of the hazard-bearing body. The model was established based on the combination weighting method of game theory, and the risk map was generated based on the GIS platform. Finally, the model was applied to the windblown sand hazards risk assessment along the Wuhai-Maqin Highway. The result showed that the risk of the windblown sand hazards along the Wuhai-Maqin Highway is mainly medium, low, and very low. High and very high risk windblown sand hazards sections account for only 33% of the total length of the highway. The high and very high risk highway sections of the windblown sand hazards are mainly distributed in the hinterland of shifting dunes area and near the horizontal curve with a small radius in the flat sandy land area. By comparing with the real information of windblown sand hazards along the highway, correlation was up to 85.93%, which verified the accuracy of the model. The model can be applied to windblown sand hazards risk assessment along the highways.

show abstract

Predicting of dust storm source by combining remote sensing, statistic-based predictive models and game theory in the Sistan watershed, southwestern Asia

Cited by 22 publications

References 79 publications

Novel deep learning hybrid models (CNN-GRU and DLDL-RF) for the susceptibility classification of dust sources in the Middle East: a global source

Novel deep learning hybrid models (CNN-GRU and DLDL-RF) for the susceptibility classification of dust sources in the Middle East: a global source

Mapping land degradation risk due to wind and water erosion

Windblown sand hazards risk assessment along the highways based on GIS-game theory combination weight

Contact Info

Product

Resources

About