New Approach to Calculation of Atmospheric Model Physics: Accurate and Fast Neural Network Emulation of Longwave Radiation in a Climate Model

2005 ◽  
Vol 133 (5) ◽  
pp. 1370-1383 ◽  
Author(s):  
Vladimir M. Krasnopolsky ◽  
Michael S. Fox-Rabinovitz ◽  
Dmitry V. Chalikov
Keyword(s):  
Neural Network ◽  
Machine Learning ◽  
Climate Model ◽  
Longwave Radiation ◽  
Forecast Model ◽  
Atmospheric Model ◽  
Statistical Machine Learning ◽  
Network Emulation ◽  
New Approach ◽  
Climate Simulations

Abstract A new approach based on a synergetic combination of statistical/machine learning and deterministic modeling within atmospheric models is presented. The approach uses neural networks as a statistical or machine learning technique for an accurate and fast emulation or statistical approximation of model physics parameterizations. It is applied to development of an accurate and fast approximation of an atmospheric longwave radiation parameterization for the NCAR Community Atmospheric Model, which is the most time consuming component of model physics. The developed neural network emulation is two orders of magnitude, 50–80 times, faster than the original parameterization. A comparison of the parallel 10-yr climate simulations performed with the original parameterization and its neural network emulations confirmed that these simulations produce almost identical results. The obtained results show the conceptual and practical possibility of an efficient synergetic combination of deterministic and statistical learning components within an atmospheric climate or forecast model. A developmental framework and practical validation criteria for neural network emulations of model physics components are outlined.

Decadal Climate Simulations Using Accurate and Fast Neural Network Emulation of Full, Longwave and Shortwave, Radiation

2008 ◽  
Vol 136 (10) ◽  
pp. 3683-3695 ◽  
Author(s):  
Vladimir M. Krasnopolsky ◽  
Michael S. Fox-Rabinovitz ◽  
Alexei A. Belochitski
Keyword(s):  
Neural Network ◽  
Climate Model ◽  
Model Simulation ◽  
Weather Prediction ◽  
Detailed Comparison ◽  
Shortwave Radiation ◽  
Full Model ◽  
Climate Simulations ◽  
Radiation Parameterizations ◽  
Decadal Climate

An approach to calculating model physics using neural network emulations, previously proposed and developed by the authors, has been implemented in this study for both longwave and shortwave radiation parameterizations, or to the full model radiation, the most time-consuming component of model physics. The developed highly accurate neural network emulations of the NCAR Community Atmospheric Model (CAM) longwave and shortwave radiation parameterizations are 150 and 20 times as fast as the original/control longwave and shortwave radiation parameterizations, respectively. The full neural network model radiation was used for a decadal climate model simulation with the NCAR CAM. A detailed comparison of parallel decadal climate simulations performed with the original NCAR model radiation parameterizations and with their neural network emulations is presented. Almost identical results have been obtained for the parallel decadal simulations. This opens the opportunity of using efficient neural network emulations for the full model radiation for decadal and longer climate simulations as well as for weather prediction.

AUTOMATIC KEYWORD EXTRACTION USING ARTIFICIAL NEURAL NETWORK AND FEATURE EXTRACTION

2020 ◽  
pp. 63-74
Author(s):  
Son
Keyword(s):  
Neural Network ◽  
Machine Learning ◽  
Artificial Neural Network ◽  
Language Processing ◽  
Extraction Methods ◽  
Keyword Extraction ◽  
Learning Models ◽  
New Approach ◽  
Word Level ◽  
Machine Learning Models

Extracting keywords from documents is an essential task in natural language processing. A challenge of this task is to define a reasonable set of keywords from which we can find all relevant documents. This paper proposes a new approach that exploits word-level handcrafted features and machine learning models to select a single document's most important keywords. To evaluate the proposed solution, we compare our results with the latest supervised and unsupervised automatic keyword extraction methods. Experiment results show that our model achieves the best results on the 9/20 data corpus. It points out that our proposed approach is promising.

scholarly journals Stable climate simulations using a realistic GCM with neural network parameterizations for atmospheric moist physics and radiation processes

2021 ◽  
Author(s):  
Xin Wang ◽  
Yilun Han ◽  
Wei Xue ◽  
Guangwen Yang ◽  
Guang J. Zhang
Keyword(s):  
Neural Network ◽  
Machine Learning ◽  
Boundary Conditions ◽  
General Circulation ◽  
Data Science ◽  
Rapid Development ◽  
Parameterization Scheme ◽  
Multilayer Perceptrons ◽  
Climate Simulations ◽  
Cloud Resolving Model

Abstract. In climate models, subgrid parameterizations of convection and cloud are one of the main reasons for the biases in precipitation and atmospheric circulation simulations. In recent years, due to the rapid development of data science, Machine learning (ML) parameterizations for convection and clouds have been proven the potential to perform better than conventional parameterizations. At present, most of the existing studies are on aqua-planet and idealized models, and the problems of simulated instability and climate drift still exist. In realistic configurated models, developing a machine learning parameterization scheme remains a challenging task. In this study, a group of deep residual multilayer perceptrons with strong nonlinear fitting ability is designed to learn a parameterization scheme from cloud-resolving model outputs. Multi-target training is achieved to best balance the fits across diverse neural network outputs. The optimal machine learning parameterization, named NN-Parameterization, is further chosen among feasible candidates for both high performance and long-term simulation. The results show that NN-Parameterization performs well in multi-year climate simulations and reproduces reasonable climatology and climate variability in a general circulation model (GCM), with a running speed of about 30 times faster than the cloud-resolving model embedded Superparameterizated GCM. Under real geographical boundary conditions, the hybrid ML-physical GCM well simulates the spatial distribution of precipitation and significantly improves the frequency of precipitation extremes, which is largely underestimated in the Community Atmospheric Model version 5 (CAM5) with the horizontal resolution of 1.9° × 2.5°. Furthermore, the hybrid ML-physical GCM simulates a stronger signal of the Madden-Julian oscillation with a more reasonable propagation speed, which is too weak and propagates too fast in CAM5. This study is a pioneer to achieve multi-year stable climate simulations using a hybrid ML-physical GCM in actual land-ocean boundary conditions. It demonstrates the emerging potential for using machine learning parameterizations in climate simulations.

scholarly journals Machine learning methods analysis in the document classification problem

2020 ◽  
pp. 081-087
Author(s):  
A.P. Zhyrkova ◽  
◽  
O.P. Ignatenko ◽  
◽  
Keyword(s):  
Neural Network ◽  
Machine Learning ◽  
Convolutional Neural Network ◽  
Input Data ◽  
Classification Problem ◽  
Model Accuracy ◽  
New Approach ◽  
Machine Learning Methods ◽  
The World ◽  
Current Article

Current situation with official documentary in the world, and especially in Ukraine, requires tools for electronical processing. One of the main tasks at this field is seal (or stamp) detection, which leads to documents classification based on mentioned criterion. Current article analyzes some of existed methods to resolve the problem, describes a new approach to classify documentary and reflects dependence of model accuracy to input data amount. As a result of this work is a convolutional neural network that classify 708 out of 804 images of official documents correctly. A corresponded percentage of model accuracy is 88.03, despite the fact of bias presence in input data.

scholarly journals Comparison of the Diurnal Cycle of Outgoing Longwave Radiation from a Climate Model with Results from ERBE

2008 ◽  
Vol 47 (12) ◽  
pp. 3188-3201 ◽  
Author(s):  
G. Louis Smith ◽  
Pamela E. Mlynczak ◽  
David A. Rutan ◽  
Takmeng Wong
Keyword(s):  
Diurnal Cycle ◽  
Outgoing Longwave Radiation ◽  
Climate Model ◽  
Longwave Radiation ◽  
Principal Component ◽  
Atmospheric Model ◽  
Radiation Budget ◽  
Boreal Summer ◽  
Diurnal Cycles ◽  
Hadley Centre

Abstract The diurnal cycle of outgoing longwave radiation (OLR) computed by a climate model provides a powerful test of the numerical description of various physical processes. Diurnal cycles of OLR computed by version 3 of the Hadley Centre Atmospheric Model (HadAM3) are compared with those observed by the Earth Radiation Budget Satellite (ERBS) for the boreal summer season (June–August). The ERBS observations cover the domain from 55°S to 55°N. To compare the observed and modeled diurnal cycles, the principal component (PC) analysis method is used over this domain. The analysis is performed separately for the land and ocean regions. For land over this domain, the diurnal cycle computed by the model has a root-mean-square (RMS) of 11.4 W m−2, as compared with 13.3 W m−2 for ERBS. PC-1 for ERBS observations and for the model are similar, but the ERBS result has a peak near 1230 LST and decreases very slightly during night, whereas the peak of the model result is an hour later and at night the OLR decreases by 7 W m−2 between 2000 and 0600 LST. Some of the difference between the ERBS and model results is due to the computation of convection too early in the afternoon by the model. PC-2 describes effects of morning/afternoon cloudiness on OLR, depending on the sign. Over ocean in the ERBS domain, the model RMS of the OLR diurnal cycle is 2.8 W m−2, as compared with 5.9 W m−2 for ERBS. Also, for the model, PC-1 accounts for 66% of the variance, while for ERBS, PC-1 accounts for only 16% of the variance. Thus, over ocean, the ERBS results show a greater variety of OLR diurnal cycles than the model does.

Development of Neural-Network Interatomic Potentials for Structural Materials

2016 ◽  
Vol 258 ◽  
pp. 69-72
Author(s):  
Ryo Kobayashi ◽  
Tomoyuki Tamura ◽  
Ichiro Takeuchi ◽  
Shuji Ogata
Keyword(s):  
Neural Network ◽  
Machine Learning ◽  
Md Simulation ◽  
Interatomic Potential ◽  
Basis Functions ◽  
Interatomic Potentials ◽  
Machine Learning Method ◽  
Learning Method ◽  
New Approach ◽  
The Neural Network

The validity of the molecular dynamics (MD) simulation is highly dependent on the accuracy or reproducibility of interatomic potentials used in the MD simulation. The neural-network (NN) interatomic potential is one of promising interatomic potentials based on machine-learning method. However, there are some parameters that should be determined heuristically before making the NN potential, such as the shape and number of basis functions. We have developed a new approach to select only relevant basis functions from a lot of candidates systematically and less heuristically without loosing the accuracy of the potential. The present NN potential for Si system shows very good agreements with the results obtained using ab-initio calculations.

Sign in / Sign up

Export Citation Format

Share Document