Artificial neural networks as a tool in urban storm drainage

Loke, E.; Warnaars, Eric; Jacobsen, Peter; Nelen, Fons; Almeida, M. C.

doi:10.1016/s0273-1223(97)00612-4

Cited by 15 publications

(1 citation statement)

References 3 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Input, hidden, and output layers make up ANNs' standard three-layer structure. Neurons acquire nonlinearity through activation functions when they receive data from various inputs (Loke et al, 1997). Multilayer Perceptron (MLP) is the most frequently used architecture in ANN (Chen et al, 2020).…”

Section: Artificial Neural Network (Ann)mentioning

confidence: 99%

A New Perspective on Estimation of Gas Flaring Volume From Space: OLI/TIRS, VIIRS, and TROPOMI

Asadi‐Fard,

Falahatkar,

Tanha Ziarati

et al. 2024

JGR Atmospheres

View full text Add to dashboard Cite

Gas flaring (GF) has the negative impact on the environment, climate, and human health. So, regular monitoring of flares and quantification of their volume is necessary. Iran has many natural oil/gas processing plants and petrochemical companies which are concentrated in the southern region. Pars Special Economic Energy Zone (PSEEZ) is an industry part with different kinds of active flares, thus a significant potential source of environmental impacts due to gas flaring. Remotely sensed data are used in gas‐flaring detection, volume estimation, and pollution emission. In this study, we applied day/nighttime radiation and air pollutant data to estimate gas flaring volumes. We developed artificial neural network models (ANN) for finding the relationship between the field measurement of GF volume as the dependent variable and shortwave infrared and thermal infrared bands of Landsat 8, M10 band of Visible Infrared Imaging Radiometer Suite, and air pollutant (NO2, CO, O3, and SO2) of TROPOMI as independent variables. Results showed that R2 values were 0.73 for the ANN model from 2018 to 2019. The sensitivity analysis demonstrated that the thermal infrared bands of B10 and B11 of Landsat 8 had the most important role in the estimation of gas flaring volume. In contrast, the SWIR bands of Landsat 8 and all TROPOMI products were insignificant. The findings of this research help to shed light on the use of remotely sensed data in estimating the volume of gas flaring at the regional/global scale by integration of the ANN model.

show abstract

Section: Artificial Neural Network (Ann)mentioning

confidence: 99%