scholarly journals An automatic and effective parameter optimization method for model tuning

2015 ◽  
Vol 8 (5) ◽  
pp. 3791-3822
Author(s):  
T. Zhang ◽  
L. Li ◽  
Y. Lin ◽  
W. Xue ◽  
F. Xie ◽  
...  
Keyword(s):  
General Circulation ◽  
Model Development ◽  
Computational Cost ◽  
Model Performance ◽  
Objective Evaluation ◽  
Optimization Methods ◽  
General Circulation Models ◽  
Optimization Method ◽  
Development Stage ◽  
Optimum Combination

Abstract. Physical parameterizations in General Circulation Models (GCMs), having various uncertain parameters, greatly impact model performance and model climate sensitivity. Traditional manual and empirical tuning of these parameters is time consuming and ineffective. In this study, a "three-step" methodology is proposed to automatically and effectively obtain the optimum combination of some key parameters in cloud and convective parameterizations according to a comprehensive objective evaluation metrics. Different from the traditional optimization methods, two extra steps, one determines parameter sensitivity and the other chooses the optimum initial value of sensitive parameters, are introduced before the downhill simplex method to reduce the computational cost and improve the tuning performance. Atmospheric GCM simulation results show that the optimum combination of these parameters determined using this method is able to improve the model's overall performance by 9%. The proposed methodology and software framework can be easily applied to other GCMs to speed up the model development process, especially regarding unavoidable comprehensive parameters tuning during the model development stage.

scholarly journals An automatic and effective parameter optimization method for model tuning

2015 ◽  
Vol 8 (11) ◽  
pp. 3579-3591 ◽  
Author(s):  
T. Zhang ◽  
L. Li ◽  
Y. Lin ◽  
W. Xue ◽  
F. Xie ◽  
...  
Keyword(s):  
Simplex Method ◽  
General Circulation ◽  
Model Development ◽  
Model Performance ◽  
Parameter Tuning ◽  
General Circulation Models ◽  
Development Stage ◽  
Optimum Combination ◽  
Downhill Simplex Method ◽  
Downhill Simplex

Abstract. Physical parameterizations in general circulation models (GCMs), having various uncertain parameters, greatly impact model performance and model climate sensitivity. Traditional manual and empirical tuning of these parameters is time-consuming and ineffective. In this study, a "three-step" methodology is proposed to automatically and effectively obtain the optimum combination of some key parameters in cloud and convective parameterizations according to a comprehensive objective evaluation metrics. Different from the traditional optimization methods, two extra steps, one determining the model's sensitivity to the parameters and the other choosing the optimum initial value for those sensitive parameters, are introduced before the downhill simplex method. This new method reduces the number of parameters to be tuned and accelerates the convergence of the downhill simplex method. Atmospheric GCM simulation results show that the optimum combination of these parameters determined using this method is able to improve the model's overall performance by 9 %. The proposed methodology and software framework can be easily applied to other GCMs to speed up the model development process, especially regarding unavoidable comprehensive parameter tuning during the model development stage.

scholarly journals An effective parameter optimization with radiation balance constraint in CAM5 (version 5.3)

2019 ◽  
Author(s):  
Li Wu ◽  
Tao Zhang ◽  
Yi Qin ◽  
Wei Xue
Keyword(s):  
Parameter Optimization ◽  
General Circulation ◽  
Performance Indicator ◽  
Optimization Problems ◽  
Model Performance ◽  
Optimization Methods ◽  
General Circulation Models ◽  
Atmospheric Model ◽  
Radiation Balance ◽  
Performance Metric

Abstract. Uncertain parameters in physical parameterizations of General Circulation Models (GCMs) greatly impact model performance. In recent years, automatic parameter optimization has been introduced for tuning model performance of GCMs but most of the optimization methods are unconstrained optimization methods under a given performance indicator, so that the calibrated model may break through essential constraints that models have to keep, such as the radiation balance at top of model, which is known for its importance to the conservation of model energy. In this study, an automated and efficient parameter optimization with the radiation balance constraint is presented and applied in Community Atmospheric Model (CAM5) in terms of a synthesized performance metric using global means of radiation, precipitation, relative humidity, and temperature. The tuned parameters are from the parameterization schemes of convection and cloud. And the radiation constraint is defined as the deviation of the net longwave flux at top of model (FLNT) and net solar flux at top of model (FSNT) less than 1 W m−2. Results show that the synthesized performance under the optimal parameters is 6.3 % better than the control run (CNTL) as well as the radiation imbalance is as low as 0.1 W m−2. The proposed method provides the insight for physics-guided optimization under the premise of a profound understanding of models and it can be easily applied to optimization problems with other prerequisite constraints in GCMs.

scholarly journals An effective parameter optimization with radiation balance constraint in CAM5 (version 5.3)

2020 ◽  
Vol 13 (1) ◽  
pp. 41-53
Author(s):  
Li Wu ◽  
Tao Zhang ◽  
Yi Qin ◽  
Wei Xue
Keyword(s):  
Parameter Optimization ◽  
General Circulation ◽  
Performance Indicator ◽  
Optimization Problems ◽  
Model Performance ◽  
Optimization Methods ◽  
General Circulation Models ◽  
Atmospheric Model ◽  
Absolute Difference ◽  
Radiation Balance

Abstract. Uncertain parameters in physical parameterizations of general circulation models (GCMs) greatly impact model performance. In recent years, automatic parameter optimization has been introduced for tuning model performance of GCMs, but most of the optimization methods are unconstrained optimization methods under a given performance indicator. Therefore, the calibrated model may break through essential constraints that models have to keep, such as the radiation balance at the top of the model. The radiation balance is known for its importance in the conservation of model energy. In this study, an automated and efficient parameter optimization with the radiation balance constraint is presented and applied in the Community Atmospheric Model (CAM5) in terms of a synthesized performance metric using normalized mean square error of radiation, precipitation, relative humidity, and temperature. The tuned parameters are from the parameterization schemes of convection and cloud. The radiation constraint is defined as the absolute difference of the net longwave flux at the top of the model (FLNT) and the net solar flux at the top of the model (FSNT) of less than 1 W m−2. Results show that the synthesized performance under the optimal parameters is 6.3 % better than the control run (CNTL) and the radiation imbalance is as low as 0.1 W m−2. The proposed method provides an insight for physics-guided optimization, and it can be easily applied to optimization problems with other prerequisite constraints in GCMs.

scholarly journals Automatic tuning of the Community Atmospheric Model CAM5.3 by using short-term hindcasts with an improved downhill simplex optimization method

2018 ◽  
Author(s):  
Tao Zhang ◽  
Minghua Zhang ◽  
Yanluan Lin ◽  
Wei Xue ◽  
Wuyin Lin ◽  
...  
Keyword(s):  
General Circulation ◽  
Model Simulation ◽  
Computational Cost ◽  
Model Performance ◽  
Optimization Procedure ◽  
General Circulation Models ◽  
Atmospheric Model ◽  
Automatic Tuning ◽  
Short Term ◽  
Cloud Forcing

Abstract. Traditional trial-and-error tuning of uncertain parameters in global atmospheric General Circulation Models (GCM) is time consuming and subjective. This study explores the feasibility of automatic optimization of GCM parameters for fast physics by using short-term hindcasts. An automatic workflow is described and applied to the Community Atmospheric Model (CAM5) to optimize several parameters in its cloud and convective parameterizations. We show that the auto-optimization leads to 10 % reduction of the overall bias in CAM5, which is already a well calibrated model, based on a pre-defined metric that includes precipitation, temperature, humidity, and longwave/shortwave cloud forcing. The computational cost of the entire optimization procedure is about equivalent to about a single 12-year atmospheric model simulation. The tuning reduces the large underestimation in the CAM5 longwave cloud forcing by decreasing the threshold relative humidity and the sedimentation velocity of ice crystals in the cloud schemes; it reduces the overestimation of precipitation by increasing the adjustment time in the convection scheme. The physical processes behind the tuned model performance for each targeted field are discussed. Limitations of the automatic tuning are described, including the slight deterioration in some targeted fields that reflect the structural errors of the model. It is pointed out that automatic tuning can be a viable supplement to process-oriented model evaluations and improvement.

scholarly journals Automatic tuning of the Community Atmospheric Model (CAM5) by using short-term hindcasts with an improved downhill simplex optimization method

2018 ◽  
Vol 11 (12) ◽  
pp. 5189-5201 ◽  
Author(s):  
Tao Zhang ◽  
Minghua Zhang ◽  
Wuyin Lin ◽  
Yanluan Lin ◽  
Wei Xue ◽  
...  
Keyword(s):  
General Circulation ◽  
Model Simulation ◽  
Computational Cost ◽  
Model Performance ◽  
Optimization Procedure ◽  
General Circulation Models ◽  
Atmospheric Model ◽  
Automatic Tuning ◽  
Short Term ◽  
Cloud Forcing

Abstract. Traditional trial-and-error tuning of uncertain parameters in global atmospheric general circulation models (GCMs) is time consuming and subjective. This study explores the feasibility of automatic optimization of GCM parameters for fast physics by using short-term hindcasts. An automatic workflow is described and applied to the Community Atmospheric Model (CAM5) to optimize several parameters in its cloud and convective parameterizations. We show that the auto-optimization leads to 10 % reduction of the overall bias in CAM5, which is already a well-calibrated model, based on a predefined metric that includes precipitation, temperature, humidity, and longwave/shortwave cloud forcing. The computational cost of the entire optimization procedure is about equivalent to a single 12-year atmospheric model simulation. The tuning reduces the large underestimation in the CAM5 longwave cloud forcing by decreasing the threshold relative humidity and the sedimentation velocity of ice crystals in the cloud schemes; it reduces the overestimation of precipitation by increasing the adjustment time in the convection scheme. The physical processes behind the tuned model performance for each targeted field are discussed. Limitations of the automatic tuning are described, including the slight deterioration in some targeted fields that reflect the structural errors of the model. It is pointed out that automatic tuning can be a viable supplement to process-oriented model evaluations and improvement.

scholarly journals A multi-diagnostic approach to cloud evaluation

2017 ◽  
Vol 10 (7) ◽  
pp. 2547-2566 ◽  
Author(s):  
Keith D. Williams ◽  
Alejandro Bodas-Salcedo
Keyword(s):  
General Circulation ◽  
Model Development ◽  
Model Performance ◽  
Coupled Model ◽  
Cloud Amount ◽  
General Circulation Models ◽  
Diagnostic Techniques ◽  
Model Errors ◽  
Intercomparison Project ◽  
Model Development Process

Abstract. Most studies evaluating cloud in general circulation models present new diagnostic techniques or observational datasets, or apply a limited set of existing diagnostics to a number of models. In this study, we use a range of diagnostic techniques and observational datasets to provide a thorough evaluation of cloud, such as might be carried out during a model development process. The methodology is illustrated by analysing two configurations of the Met Office Unified Model – the currently operational configuration at the time of undertaking the study (Global Atmosphere 6, GA6), and the configuration which will underpin the United Kingdom's Earth System Model for CMIP6 (Coupled Model Intercomparison Project 6; GA7). By undertaking a more comprehensive analysis which includes compositing techniques, comparing against a set of quite different observational instruments and evaluating the model across a range of timescales, the risks of drawing the wrong conclusions due to compensating model errors are minimized and a more accurate overall picture of model performance can be drawn. Overall the two configurations analysed perform well, especially in terms of cloud amount. GA6 has excessive thin cirrus which is removed in GA7. The primary remaining errors in both configurations are the in-cloud albedos which are too high in most Northern Hemisphere cloud types and sub-tropical stratocumulus, whilst the stratocumulus on the cold-air side of Southern Hemisphere cyclones has in-cloud albedos which are too low.

scholarly journals A multi-diagnostic approach to cloud evaluation

2017 ◽  
Author(s):  
Keith D. Williams ◽  
Alejandro Bodas-Salcedo
Keyword(s):  
General Circulation ◽  
Model Development ◽  
Model Performance ◽  
Coupled Model ◽  
Cloud Amount ◽  
General Circulation Models ◽  
Diagnostic Techniques ◽  
Model Errors ◽  
Intercomparison Project ◽  
Model Development Process

Abstract. Most studies evaluating cloud in general circulation models present new diagnostic techniques or observational datasets, or apply a limited set of existing diagnostics to a number of models. In this study, we use a range of diagnostic techniques and observational datasets to provide a thorough evaluation of cloud, such as might be carried out during a model development process. The methodology is illustrated by analysing two configurations of the Met Office Unified Model – the currently operational configuration at the time of undertaking the study (Global Atmosphere 6, GA6), and the configuration which will underpin the United Kingdom's Earth System Model for CMIP6 (Coupled Model Intercomparison Project 6) (GA7). By undertaking a more comprehensive analysis which includes compositing techniques, comparing against a set of quite different observational instruments and evaluating the model across a range of timescales, the risks of drawing the wrong conclusions due to compensating model errors are minimised and a more accurate overall picture of model performance can be drawn. Overall the two configurations analysed perform well, especially in terms of cloud amount. GA6 has excessive thin cirrus which is removed in GA7. The primary remaining errors in both configurations are the in-cloud albedos which are too high in most northern hemisphere cloud types and sub-tropical stratocumulus, whilst the stratocumulus on the cold air side of southern hemisphere cyclones has in-cloud albedo's which are too low.

On the use of idealised test cases for ocean model development

2021 ◽  
Author(s):  
Julie Deshayes
Keyword(s):  
Boundary Conditions ◽  
General Circulation ◽  
Model Development ◽  
General Circulation Models ◽  
Ocean Model ◽  
Ocean Dynamics ◽  
Test Cases ◽  
Training Tool ◽  
Research Activities ◽  
Historical Progress

<p>When comparing realistic simulations produced by two ocean general circulation models, differences may emerge from alternative choices in boundary conditions and forcings, which alters our capacity to identify the actual differences between the two models (in the equations solved, the discretization schemes employed and/or the parameterizations introduced). The use of idealised test cases (idealized configurations with analytical boundary conditions and forcings, resolving a given set of equations) has proven efficient to reveal numerical bugs, determine advantages and pitfalls of certain numerical choices, and highlight remaining challenges. I propose to review historical progress enabled by the use of idealised test cases, and promote their utilization when assessing ocean dynamics as represented by an ocean model. For the latter, I would illustrate my talk using illustrations from my own research activities using NEMO in various contexts. I also see idealised test cases as a promising training tool for inexperienced ocean modellers, and an efficient solution to enlarge collaboration with experts in adjacent disciplines, such as mathematics, fluid dynamics and computer sciences.</p>

Application of Cluster Analysis to Climate Model Performance Metrics

2011 ◽  
Vol 50 (8) ◽  
pp. 1666-1675 ◽  
Author(s):  
Satoru Yokoi ◽  
Yukari N. Takayabu ◽  
Kazuaki Nishii ◽  
Hisashi Nakamura ◽  
Hirokazu Endo ◽  
...  
Keyword(s):  
Cluster Analysis ◽  
General Circulation ◽  
Performance Metrics ◽  
Climate Model ◽  
Model Performance ◽  
General Circulation Models ◽  
Clustering Methods ◽  
Dynamic Constraints ◽  
Overall Performance

AbstractThe overall performance of general circulation models is often investigated on the basis of the synthesis of a number of scalar performance metrics of individual models that measure the reproducibility of diverse aspects of the climate. Because of physical and dynamic constraints governing the climate, a model’s performance in simulating a certain aspect of the climate is sometimes related closely to that in simulating another aspect, which results in significant intermodel correlation between performance metrics. Numerous metrics and intermodel correlations may cause a problem in understanding the evaluation and synthesizing the metrics. One possible way to alleviate this problem is to group the correlated metrics beforehand. This study attempts to use simple cluster analysis to group 43 performance metrics. Two clustering methods, the K-means and the Ward methods, yield considerably similar clustering results, and several aspects of the results are found to be physically and dynamically reasonable. Furthermore, the intermodel correlation between the cluster averages is considerably lower than that between the metrics. These results suggest that the cluster analysis is helpful in obtaining the appropriate grouping. Applications of the clustering results are also discussed.

scholarly journals Thermal-based early breast cancer detection using inception V3, inception V4 and modified inception MV4

2021 ◽  
Author(s):  
Mohammed Abdulla Salim Al Husaini ◽  
Mohamed Hadi Habaebi ◽  
Teddy Surya Gunawan ◽  
Md Rafiqul Islam ◽  
Elfatih A. A. Elsheikh ◽  
...  
Keyword(s):  
Breast Cancer ◽  
Neural Networks ◽  
Deep Learning ◽  
Convolutional Neural Networks ◽  
Early Stage ◽  
Computational Cost ◽  
Optimization Methods ◽  
Optimization Method ◽  
Deep Convolutional Neural Networks ◽  
Breast Thermography

AbstractBreast cancer is one of the most significant causes of death for women around the world. Breast thermography supported by deep convolutional neural networks is expected to contribute significantly to early detection and facilitate treatment at an early stage. The goal of this study is to investigate the behavior of different recent deep learning methods for identifying breast disorders. To evaluate our proposal, we built classifiers based on deep convolutional neural networks modelling inception V3, inception V4, and a modified version of the latter called inception MV4. MV4 was introduced to maintain the computational cost across all layers by making the resultant number of features and the number of pixel positions equal. DMR database was used for these deep learning models in classifying thermal images of healthy and sick patients. A set of epochs 3–30 were used in conjunction with learning rates 1 × 10–3, 1 × 10–4 and 1 × 10–5, Minibatch 10 and different optimization methods. The training results showed that inception V4 and MV4 with color images, a learning rate of 1 × 10–4, and SGDM optimization method, reached very high accuracy, verified through several experimental repetitions. With grayscale images, inception V3 outperforms V4 and MV4 by a considerable accuracy margin, for any optimization methods. In fact, the inception V3 (grayscale) performance is almost comparable to inception V4 and MV4 (color) performance but only after 20–30 epochs. inception MV4 achieved 7% faster classification response time compared to V4. The use of MV4 model is found to contribute to saving energy consumed and fluidity in arithmetic operations for the graphic processor. The results also indicate that increasing the number of layers may not necessarily be useful in improving the performance.

Sign in / Sign up

Export Citation Format

Share Document