Predicting and Reconstructing Aerosol-Cloud-Precipitation Interactions with Physics-Informed Neural Networks

被引:1
作者
Hu, Alice V. [1 ]
Kabala, Zbigniew J. [2 ]
机构
[1] Pioneer Acad, 101 Greenwood Ave,Ste 170, Jenkintown, PA 19046 USA
[2] Duke Univ, Dept Civil & Environm Engn, Durham, NC 27708 USA
关键词
Koren-Feingold model; aerosol-cloud-precipitation interactions; physics-informed neural networks; artificial neural networks; prediction; reconstruction; RELEASE HISTORY; DIURNAL-VARIATIONS; MODEL; VARIABILITY; EQUATION; SYSTEM; CYCLE;
D O I
10.3390/atmos14121798
中图分类号
X [环境科学、安全科学];
学科分类号
08 ; 0830 ;
摘要
Interactions between clouds, aerosol, and precipitation are crucial aspects of weather and climate. The simple Koren-Feingold conceptual model is important for providing deeper insight into the complex aerosol-cloud-precipitation system. Recently, artificial neural networks (ANNs) and physics-informed neural networks (PINNs) have been used to study multiple dynamic systems. However, the Koren-Feingold model for aerosol-cloud-precipitation interactions has not yet been studied with either ANNs or PINNs. It is challenging for pure data-driven models, such as ANNs, to accurately predict and reconstruct time series in a small data regime. The pure data-driven approach results in the ANN becoming a "black box" that limits physical interpretability. We demonstrate how these challenges can be overcome by combining a simple ANN with physical laws into a PINN model (not purely data-driven, good for the small data regime, and interpretable). This paper is the first to use PINNs to learn about the original and modified Koren-Feingold models in a small data regime, including external forcings such as wildfire-induced aerosols or the diurnal cycle of clouds. By adding external forcing, we investigate the effects of environmental phenomena on the aerosol-cloud-precipitation system. In addition to predicting the system's future, we also use PINN to reconstruct the system's past: a nontrivial task because of time delay. So far, most research has focused on using PINNs to predict the future of dynamic systems. We demonstrate the PINN's ability to reconstruct the past with limited data for a dynamic system with nonlinear delayed differential equations, such as the Koren-Feingold model, which remains underexplored in the literature. The main reason that this is possible is that the model is non-diffusive. We also demonstrate for the first time that PINNs have significant advantages over traditional ANNs in predicting the future and reconstructing the past of the original and modified Koren-Feingold models containing external forcings in the small data regime. We also show that the accuracy of the PINN is not sensitive to the value of the regularization factor (lambda), a key parameter for the PINN that controls the weight for the physics loss relative to the data loss, for a broad range (from lambda=1x103 to lambda=1x105).
引用
收藏
页数:25
相关论文
共 66 条
[1]   Pan evaporation estimation by relevance vector machine tuned with new metaheuristic algorithms using limited climatic data [J].
Adnan, Rana Muhammad ;
Mostafa, Reham R. ;
Dai, Hong-Liang ;
Heddam, Salim ;
Kuriqi, Alban ;
Kisi, Ozgur .
ENGINEERING APPLICATIONS OF COMPUTATIONAL FLUID MECHANICS, 2023, 17 (01)
[2]   Application of Advanced Optimized Soft Computing Models for Atmospheric Variable Forecasting [J].
Adnan, Rana Muhammad ;
Meshram, Sarita Gajbhiye ;
Mostafa, Reham R. ;
Islam, Abu Reza Md. Towfiqul ;
Abba, S. I. ;
Andorful, Francis ;
Chen, Zhihuan .
MATHEMATICS, 2023, 11 (05)
[3]   Modelling groundwater level fluctuations by ELM merged advanced metaheuristic algorithms using hydroclimatic data [J].
Adnan, Rana Muhammad ;
Dai, Hong-Liang ;
Mostafa, Reham R. ;
Islam, Abu Reza Md. Towfiqul ;
Kisi, Ozgur ;
Heddam, Salim ;
Zounemat-Kermani, Mohammad .
GEOCARTO INTERNATIONAL, 2023, 38 (01)
[4]   Modeling Multistep Ahead Dissolved Oxygen Concentration Using Improved Support Vector Machines by a Hybrid Metaheuristic Algorithm [J].
Adnan, Rana Muhammad ;
Dai, Hong-Liang ;
Mostafa, Reham R. ;
Parmar, Kulwinder Singh ;
Heddam, Salim ;
Kisi, Ozgur .
SUSTAINABILITY, 2022, 14 (06)
[5]   Estimating reference evapotranspiration using hybrid adaptive fuzzy inferencing coupled with heuristic algorithms [J].
Adnan, Rana Muhammad ;
Mostafa, Reham R. ;
Islam, Abu Reza Md. Towfiqul ;
Kisi, Ozgur ;
Kuriqi, Alban ;
Heddam, Salim .
COMPUTERS AND ELECTRONICS IN AGRICULTURE, 2021, 191
[6]  
Alapati S, 2000, HYDROL PROCESS, V14, P1003, DOI 10.1002/(SICI)1099-1085(20000430)14:6<1003::AID-HYP981>3.0.CO
[7]  
2-W
[8]   Coagulation in a diffusing Gaussian aerosol puff: Comparison of analytical approximations with numerical solutions [J].
Anand, S. ;
Mayya, Y. S. .
JOURNAL OF AEROSOL SCIENCE, 2009, 40 (04) :348-361
[9]   Low Reynolds number capture of small particles on a cylinder by diffusion, interception, and inertia at subcritical Stokes numbers: Numerical calculations, correlations, and small diffusivity asymptote [J].
Arias-Zugasti, M. ;
Rosner, D. E. ;
de la Mora, J. Fernandez .
AEROSOL SCIENCE AND TECHNOLOGY, 2019, 53 (12) :1367-1380
[10]   The interactions between wildfires, bamboo and aerosols in the south-western Amazon: A conceptual model [J].
Bae, Jiyong ;
Kim, Daehyun .
PROGRESS IN PHYSICAL GEOGRAPHY-EARTH AND ENVIRONMENT, 2021, 45 (04) :621-631