Development of neural network convection parameterizations for numerical climate and weather prediction models using cloud resolving model simulations

2010 ◽  
Author(s):  
Vladimir M. Krasnopolsky ◽  
Michael S. Fox-Rabinovitz ◽  
Alexei A. Belochitski
Keyword(s):  
Neural Network ◽  
Prediction Models ◽  
Weather Prediction ◽  
Model Simulations ◽  
Cloud Resolving Model

scholarly journals Updraft Constraints on Entrainment: Insights from Amazonian Deep Convection

2019 ◽  
Vol 76 (8) ◽  
pp. 2429-2442 ◽  
Author(s):  
Usama M. Anber ◽  
Scott E. Giangrande ◽  
Leo J. Donner ◽  
Michael P. Jensen
Keyword(s):  
Ad Hoc ◽  
Prediction Models ◽  
Mixing Time ◽  
Deep Convection ◽  
Weather Prediction ◽  
Limited Area ◽  
Grid Cells ◽  
Model Simulations ◽  
Cloud Resolving Model ◽  
Environmental Air

AbstractMixing of environmental air into clouds, or entrainment, has been identified as a major contributor to erroneous climate predictions made by modern comprehensive climate and numerical weather prediction models. Despite receiving extensive attention, the ad hoc treatment of this convective-scale process in global models remains poor. On the other hand, while limited-area high-resolution nonhydrostatic models can directly resolve entrainment, their sensitivity to model resolution, especially with the lack of benchmark mass flux observations, limits their applicability. Here, the dataset from the Observations and Modeling of the Green Ocean Amazon (GoAmazon2014/5) campaign focusing on radar retrievals of convective updraft vertical velocities is used with the aid of cloud-resolving model simulations of four deep convective events over the Amazon to provide insights into entrainment. Entrainment and detrainment are diagnosed from the model simulations by applying the mass continuity equation over cloud volumes, in which grid cells are identified by some thresholds of updraft vertical velocity and cloud condensates, and accounting for the sources and sinks of the air mass. Entrainment is then defined as the environmental air intruding into convective cores causing cloud volume to shrink, while detrainment is defined as cloudy grid cells departing the convective core and causing cloud volume to expand. It is found that the diagnosed entrainment from the simulated convective events is strongly correlated to the inverse of the updraft vertical velocities in convective cores, which enables a more robust estimation of the mixing time scale. This highlights the need for improved observational capabilities for sampling updraft velocities across diverse geographic and cloud conditions. Evaluation of a number of assumptions used to represent entrainment in parameterization schemes is also presented, as contrasted against the diagnosed one.

scholarly journals Multirule Based Diagnostic Approach for the Fog Predictions Using WRF Modelling Tool

2014 ◽  
Vol 2014 ◽  
pp. 1-11 ◽  
Author(s):  
Swagata Payra ◽  
Manju Mohan
Keyword(s):  
Prediction Models ◽  
Weather Prediction ◽  
Winter Season ◽  
Wrf Model ◽  
Onset Time ◽  
Northwestern Part ◽  
Model Simulations ◽  
Western Disturbances ◽  
Promising Tool ◽  
Local Perturbations

The prediction of fog onset remains difficult despite the progress in numerical weather prediction. It is a complex process and requires adequate representation of the local perturbations in weather prediction models. It mainly depends upon microphysical and mesoscale processes that act within the boundary layer. This study utilizes a multirule based diagnostic (MRD) approach using postprocessing of the model simulations for fog predictions. The empiricism involved in this approach is mainly to bridge the gap between mesoscale and microscale variables, which are related to mechanism of the fog formation. Fog occurrence is a common phenomenon during winter season over Delhi, India, with the passage of the western disturbances across northwestern part of the country accompanied with significant amount of moisture. This study implements the above cited approach for the prediction of occurrences of fog and its onset time over Delhi. For this purpose, a high resolution weather research and forecasting (WRF) model is used for fog simulations. The study involves depiction of model validation and postprocessing of the model simulations for MRD approach and its subsequent application to fog predictions. Through this approach model identified foggy and nonfoggy days successfully 94% of the time. Further, the onset of fog events is well captured within an accuracy of 30–90 minutes. This study demonstrates that the multirule based postprocessing approach is a useful and highly promising tool in improving the fog predictions.

scholarly journals Using Ensemble of Neural Networks to Learn Stochastic Convection Parameterizations for Climate and Numerical Weather Prediction Models from Data Simulated by a Cloud Resolving Model

2013 ◽  
Vol 2013 ◽  
pp. 1-13 ◽  
Author(s):  
Vladimir M. Krasnopolsky ◽  
Michael S. Fox-Rabinovitz ◽  
Alexei A. Belochitski
Keyword(s):  
Numerical Weather Prediction ◽  
Prediction Models ◽  
Meteorological Data ◽  
Weather Prediction ◽  
Simulated Data ◽  
Boreal Winter ◽  
Ensemble Technique ◽  
Numerical Weather ◽  
Cloud Resolving Model ◽  
Convection Parameterization

A novel approach based on the neural network (NN) ensemble technique is formulated and used for development of a NN stochastic convection parameterization for climate and numerical weather prediction (NWP) models. This fast parameterization is built based on learning from data simulated by a cloud-resolving model (CRM) initialized with and forced by the observed meteorological data available for 4-month boreal winter from November 1992 to February 1993. CRM-simulated data were averaged and processed to implicitly define a stochastic convection parameterization. This parameterization is learned from the data using an ensemble of NNs. The NN ensemble members are trained and tested. The inherent uncertainty of the stochastic convection parameterization derived following this approach is estimated. The newly developed NN convection parameterization has been tested in National Center of Atmospheric Research (NCAR) Community Atmospheric Model (CAM). It produced reasonable and promising decadal climate simulations for a large tropical Pacific region. The extent of the adaptive ability of the developed NN parameterization to the changes in the model environment is briefly discussed. This paper is devoted to a proof of concept and discusses methodology, initial results, and the major challenges of using the NN technique for developing convection parameterizations for climate and NWP models.

Evaluating Convective Initiation in High-Resolution Numerical Weather Prediction Models Using GOES-16 Infrared Brightness Temperatures

2021 ◽  
Vol 149 (4) ◽  
pp. 1153-1172
Author(s):  
David S. Henderson ◽  
Jason A. Otkin ◽  
John R. Mecikalski
Keyword(s):  
High Resolution ◽  
Prediction Models ◽  
Weather Prediction ◽  
Cloud Microphysics ◽  
Cloud Water ◽  
Convective Initiation ◽  
Model Simulations ◽  
Brightness Temperatures ◽  
Cloud Characteristics ◽  
Verification Methods

AbstractThe evolution of model-based cloud-top brightness temperatures (BT) associated with convective initiation (CI) is assessed for three bulk cloud microphysics schemes in the Weather Research and Forecasting Model. Using a composite-based analysis, cloud objects derived from high-resolution (500 m) model simulations are compared to 5-min GOES-16 imagery for a case study day located near the Alabama–Mississippi border. Observed and simulated cloud characteristics for clouds reaching CI are examined by utilizing infrared BTs commonly used in satellite-based CI nowcasting methods. The results demonstrate the ability of object-based verification methods with satellite observations to evaluate the evolution of model cloud characteristics, and the BT comparison provides insight into a known issue of model simulations producing too many convective cells reaching CI. The timing of CI from the different microphysical schemes is dependent on the production of ice in the upper levels of the cloud, which typically occurs near the time of maximum cloud growth. In particular, large differences in precipitation formation drive differences in the amount of cloud water able to reach upper layers of the cloud, which impacts cloud-top glaciation. Larger cloud mixing ratios are found in clouds with sustained growth leading to more cloud water lofted to the upper levels of the cloud and the formation of ice. Clouds unable to sustain growth lack the necessary cloud water needed to form ice and grow into cumulonimbus. Clouds with slower growth rates display similar BT trends as clouds exhibiting growth, which suggests that forecasting CI using geostationary satellites might require additional information beyond those derived at cloud top.

Calibration of solar radiation ensemble forecasts using convolutional neural network

2021 ◽  
Author(s):  
Florian Dupuy ◽  
Yen-Sen Lu ◽  
Garrett Good ◽  
Michaël Zamo
Keyword(s):  
Neural Network ◽  
Solar Radiation ◽  
Convolutional Neural Network ◽  
Prediction Models ◽  
Weather Prediction ◽  
Special Kind ◽  
Ground Truth ◽  
Weather Forecast ◽  
Ensemble Forecasts ◽  
Processing Techniques

<p><span>E</span><span>nsemble </span><span>forecast </span><span>approaches have become state-of-the-art for the quantification of weather forecast uncertainty. </span><span>However</span><span>, ensemble forecasts </span><span>from</span><span> numerical weather prediction models (NWPs) still tend to be biased and underdispersed, </span>hence justifying the use of statistical post-processing techniques <span>to improve forecast skill. </span></p><p>In this study, ensemble forecasts are post-processed using a convolutional neural network (CNN). CNNs are the most popular machine learning tool to deal with images. In our case, CNNs allow to integrate information from spatial patterns contained in NWP outputs.</p><p>We focus on solar radiation forecasts for 48 hours ahead over Europe from the 35-members ARPEGE (Météo-France global NWP) and a 512-members WRF (Weather Research and Forecasting) ensembles. We used a U-Net (a special kind of CNN) designed to produce a probabilistic forecast (quantiles) using as ground truth the CAMS (Copernicus Atmosphere Monitoring System) radiation service dataset with a spatial resolution of 0.2°.</p>

A Novel Soil Moisture Predicting Method Based on Artificial Neural Network and Xinanjiang Model

2010 ◽  
Vol 121-122 ◽  
pp. 1028-1032 ◽  
Author(s):  
Jing Wen Xu ◽  
Jun Fang Zhao ◽  
Wan Chang Zhang ◽  
Xiao Xun Xu
Keyword(s):  
Neural Network ◽  
Artificial Neural Network ◽  
Soil Moisture ◽  
Prediction Models ◽  
Weather Prediction ◽  
Eastern China ◽  
Agricultural Drought ◽  
Xinanjiang Model ◽  
Predicting Model ◽  
Artificial Neural

Soil moisture plays an important role in agricultural drought predicting, therefore there is an increasing demand for detailed predictions of soil moisture, especially at basin scales. However, so far soil moisture predictions are usually obtained as a by-product of climate and weather prediction models coupled with a land surface parameterization scheme, and there has been little dedicated work to meet this urgent need at basin scales. In order to improve the basin hydrological models’ performance in the soil moisture forecasting, an integrated soil moisture predicting model based on Artificial Neural Network (ANN) and Xinanjiang model is proposed and presented in this paper. The performance of the new integrated soil moisture predicting model was tested in the Linyi watershed with a drainage area of 10040 km2, located in the semi-arid area of the eastern China. The results suggest that the soil moisture simulated by the integrated ANN-Xinanjiang model is more agree with the observed ones than that simulated by Xinanjiang, and that the simulated soil moisture by both the models has the similar trend and temporal change pattern with the observed one.

scholarly journals Neural Network Based Quality Control of CYGNSS Wind Retrieval

2020 ◽  
Vol 12 (17) ◽  
pp. 2859
Author(s):  
Rajeswari Balasubramaniam ◽  
Christopher Ruf
Keyword(s):  
Neural Network ◽  
Quality Control ◽  
Wind Speed ◽  
Outlier Detection ◽  
Prediction Models ◽  
Weather Prediction ◽  
Filter Performance ◽  
Ocean Science ◽  
Trade Offs ◽  
Speed Range

Global Navigation Satellite System – Reflectometry (GNSS-R) is a relatively new field in remote sensing that uses reflected GPS signals from the Earth’s surface to study the state of the surface geophysical parameters under observation. The CYGNSS is a first of its kind GNSS-R constellation mission launched in December 2016. It aims at providing high quality global scale GNSS-R measurements that can reliably be used for ocean science applications such as the study of ocean wind speed dynamics, tropical cyclone genesis, coupled ocean wave modelling, and assimilation into Numerical Weather Prediction models. To achieve this goal, strong quality control filters are needed to detect and remove outlier measurements. Currently, quality control of CYGNSS data products are based on fixed thresholds on various engineering, instrument, and measurement conditions. In this work we develop a Neural Network based quality control filter for automated outlier detection of CYGNSS retrieved winds. The primary merit of the proposed ML filter is its ability to better account for interactions between the individual engineering, instrument and measurement conditions than can separate thresholded flags for each one. Use of Machine Learning capabilities to capture inherent patterns in the data can create an efficient and effective mechanism to detect and remove outlier measurements. The resulting filter has a probability of outlier detection (PD) >75% and False Alarm Rate (FAR) < 20% for a wind speed range of 5 to 18 m/s. At least 75% of the outliers with wind speed errors of at least 5 m/s are removed while ~100% of the outliers with wind speed errors of at least 10 m/s are removed. This filter significantly improves data quality. The standard deviation of wind speed retrieval error is reduced from 2.6 m/s without the filter to 1.7 m/s with it over a wind speed range of 0 to 25 m/s. The design space for this filter is also analyzed in this work to characterize trade-offs between PD and FAR. Currently the filter performance is applicable only up to moderate wind speeds, as sufficient data is available only in this range to train the filter, as a way forward, more data over time can help expand the usability of this filter to higher wind speed ranges as well.

scholarly journals Gridded Visibility Products over Marine Environments Based on Artificial Neural Network Analysis

2019 ◽  
Vol 9 (21) ◽  
pp. 4487 ◽  
Author(s):  
Yulong Shan ◽  
Ren Zhang ◽  
Ismail Gultepe ◽  
Yaojia Zhang ◽  
Ming Li ◽  
...  
Keyword(s):  
Neural Network ◽  
Artificial Neural Network ◽  
Network Analysis ◽  
Prediction Models ◽  
Weather Prediction ◽  
Machine Learning Algorithms ◽  
Marine Environments ◽  
Neural Network Analysis ◽  
Artificial Neural Network Analysis ◽  
Artificial Neural

The reconstruction and monitoring of visibility over marine environments is critically important because of a lack of observations. To travel safely in marine environments, a high quality of visibility data is needed to evaluate navigation risk. Currently, although visibility is available through numerical weather prediction models as well as ground and spaceborne remote sensing platforms and ship measurements, issues still exist over the remote marine environments and northern latitudes. To improve visibility prediction and reduce navigational risks, gridded visibility data based on artificial neural network analysis can be used over marine environments, and the problem can be regarded as an air quality prediction problem based on machine learning algorithms. This new method based on artificial intelligence techniques developed here is tested over the Indian Ocean. The mean error of the inferred visibility from the artificial neural network analysis is found to be less than 8.0%. The results suggested that satellite-based optical thickness and numerical model-based reanalysis data can be used to infer gridded visibility values based on artificial neural network analysis, and that could help us reconstruct and monitor surface gridded visibility values over marine and remote environments.

Towards an Adaptive Artificial Neural Network based Postprocessor

2021 ◽  
Author(s):  
Paul J. Roebber
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Linear Regression ◽  
Multiple Linear Regression ◽  
Network Architecture ◽  
Prediction Models ◽  
Weather Prediction ◽  
Training Data ◽  
Dynamical Model ◽  
Artificial Neural

AbstractWe introduce an adaptive form of postprocessor where algorithm structures are neural networks where the number of hidden nodes and the network training features evolve. Key potential advantages of this system are the flexible, nonlinear mapping capabilities of neural networks and, through backpropagation, the ability to rapidly establish capable predictors in an algorithm population. The system can be implemented after one initial training process and future changes to postprocessor inputs (new observations, new inputs or model upgrades) are incorporated as they become available. As in prior work, the implementation in the form of a predator-prey ecosystem allows for the ready construction of ensembles. Computational requirements are minimal, and the use of a moving data window means that data storage requirements are constrained.The system adds predictive skill to a demonstration dynamical model representing the hemispheric circulation, with skill competitive with or exceeding that obtainable from multiple linear regression and standard artificial neural networks constructed under typical operational limitations. The system incorporates new information rapidly and the dependence of the approach on the training data size is similar to multiple linear regression. A loss of performance occurs relative to a fixed neural network architecture in which only the weights are adjusted after training, but this loss is compensated for by gains from the ensemble predictions. While the demonstration dynamical model is complex, current numerical weather prediction models are considerably more so, and thus a future step will be to apply this technique to operational weather forecast data.

Sign in / Sign up

Export Citation Format

Share Document