I would like to recommend some modeling software related to atmospheric science to everyone. Today I have mainly compiled some of the ones that are in high demand. You can learn more about them. If you have zero foundation, you can click this link >> Zero foundation to learn air pollution models (WRF, WRF-chem, smoke, camx, etc.)
Table of contents
3. Practical application of PMF source analysis technology for atmospheric particulate matter
4. Practical technical application of air pollution diffusion model Calpuff
10. Python practices technology in automated operation and pre- and post-processing of WRF models
11. Practical technical applications of common earth science data processing
13. Air quality simulation and pollution source analysis technology and case analysis based on CAMx
14. Systematic learning of CMAQ air quality model
1. (WRF-UCM) High-precision urbanization meteorological dynamic simulation technology and case applications
Climate change and response are the focus of government, scientific and business circles. Climate is the main driving factor in many fields (ecology, water resources, wind resources, carbon neutrality, etc.). A reasonable understanding of climate change is conducive to explaining the mechanisms and processes of ecological environment changes, and understanding current and future climate change is crucial. It is a prerequisite for ecological, environmental and energy assessment and carbon policy planning, and climate simulation is the most important means to obtain high-precision climate information. Modern ecology, hydrology, new energy and carbon neutral fields require sub-kilometre and higher resolution meteorology. For simulation, the WRF model is the most widely used meteorological model at home and abroad. There are more and more applications using this model for high-precision simulations of even hundreds of meters. This urban canopy model (WRF-UCM) can realize fine dynamic simulation of small- and medium-scale meteorological processes in cities, and its application scope and practical business and scientific research applications are becoming more and more extensive.
[Brief description of content]:
Basic theory of the model
1. Introduction to WRF-UCM model
2. What kind of computing platform does WRF-UCM use? Computing system?
Introduction to model platform installation from scratch
1. How to install the platform required for WRF-UCM mode? (windows platform + Vmware16)
2. How to build the system required for WRF-UCM compilation from scratch? (RockyLinux9)
3. Install the compiler (OneApi) required to convert WRF-UCM from code to program
4. Explanation of WRF-UCM mode input and output file format (NetCDF)
City module offline simulation case explanation
1. Drive data GLDAS data acquisition technology
2. Drive data ERA5 data acquisition technology
3. NoahMP-hrldas pre-processing case
4. NoahMP-UCM offline simulation case
City module online coupling (WRF+WRF-UCM) simulation case explanation
1. WRF+WRF-UCM pre-processing technology explanation
1.1 WRF+WRF-UCM simulation area setting technology
1.2 WRF+WRF-UCM elevation, land use, vegetation and other geographical data Explanation of meteorological data processing techniques such as temperature, pressure, humidity, and wind
2. How WRF+WRF-UCM simulates meteorological fields
2.1 WRF+WRF-UCM non-coupled simulation case
2.2 WRF+WRF-UCM coupled simulation case explanation
Practical application and case analysis
1. How to analyze the WRF-UCM simulation results? (NCL)
1.1 Example 1 (line graph)
1.2 Example (coloring picture)
1.3 Example (layer overlay)
1.4 Example (layer arrangement)
2. WRF DA data assimilation system theory, operation and technical application of new methods of variation and hybrid assimilation
Numerical prediction has become an important means to improve the quality of forecasts, and the quality of model initial values is an important link in determining the quality of numerical forecasts. As an effective method to improve the quality of model initial values, data assimilation has become a key scientific method in current scientific research and operational forecasting in many fields such as meteorology, ocean and atmospheric environment, and hydrology. The rapid development of new data assimilation methods and the increasing number of conventional meteorological data, satellite remote sensing observations, atmospheric environment and other data have laid a solid scientific foundation for the effective application of data assimilation, and have also led to many new and complex scientific problems, increasing practical Difficulty of application.
In order to effectively enhance the theoretical foundation and application capabilities of data assimilation and research level of the majority of scientific research and business personnel, and improve the quality and actual forecasting level of numerical forecasting services in the fields of meteorology, oceanography and atmospheric environment in my country. In response to the requirements of workers in the fields of meteorology and atmospheric environment, this content takes the WRF DA variational data assimilation system as the core, focuses on key and difficult issues in practical applications, and explains the data assimilation theory and methods, combined with actual weather conditions. For example, we are committed to cultivating and improving everyone's theoretical level and practical application ability.
【Target】:
1. Master and understand the theory of data assimilation;
2. Familiar with the configuration and application in the Linux environment;
3. Master the tuning and updating of background error covariance and physical constraint relationships;
4. Master the key technologies of new observation data assimilation, such as The construction of observation operators, the development of accompanying programs, etc.;
5. Be familiar with the analysis methods of assimilation results;
6. Master post-processing tools such as Bash, GRADS, and NCL for result processing and analysis;
7. Master the knowledge in meteorology, environment, ecology, and hydrology Expansion and methods of application technology in other fields;
8. Proficient in the application of WRF ETKF-3DVAR DA hybrid assimilation system;
[Brief description of content]:
1. Basic theories and methods of data assimilation
The basic concepts and development of data assimilation.
The main scientific issues of
data assimilation. The theoretical basis and early methods of data assimilation.
Various data assimilation schemes and their characteristics.
2. Environmental requirements, system installation, debugging and operation (practical operation) of WRF DA
Application of Linux basic commands
Setting of Linux environment variables
Compiler installation and setting
WRF DA environment variables and dynamic library configuration, etc.;
WRF DA installation, compilation and operation, etc.
Compilation, installation and use of GrADS and NCL drawings
3. Observation data and quality control
Basic requirements for observation data,
conventional observation data and unconventional observations,
satellite data and their assimilation,
new observation systems,
quality control and general methods of observation data.
4. Configuration of WRF DA assimilation system and structure of background error
The setting of horizontal grid;
the configuration of vertical levels;
the selection of physical scheme and parameterization scheme; the selection of
observation elements,
the role of background error covariance and commonly used construction methods, such as the innovation vector method (taking the GRAPES model as an example), NMC Methods, etc. (WRF DA's three-dimensional variational program is an example.
The construction method of background error.
The construction of medium and small scale variational assimilation B.
The construction and characteristics of background error covariance in hybrid assimilation.
The proportional relationship of B in hybrid assimilation.
5. New methods of WRF DA variational data assimilation and WRF DA hybrid assimilation
The basic framework of WRF DA variational assimilation,
the construction of observation operators, the
realization of dynamic balance and physical constraints,
the adjoint model and its programming,
the assimilation application method for solving new data.
Characteristics of the ETKF-WRF DA hybrid assimilation scheme;
Description of the ETKF ensemble perturbation scheme;
Extended control variable method and its settings;
Construction of flow-dependent background error covariance
; Example analysis;
6. Single-point test of variational assimilation and single-point test of mixed assimilation
Single-point test of mass field
Single-point test of wind field Single
-point test of humidity field
Experimental analysis of flow-dependent properties (combining trough and typhoon)
Analysis of incremental structural and physical constraints
7. Analysis of incremental assimilation analysis
The impact of different observation data on analysis. The
impact of different observation elements on analysis. Determination of
effective observation data.
Understanding of analysis increments.
8. Practical application of WRF DA and ETKF-3DVAR hybrid assimilation system
Combined with actual weather cases, focus on mastering the assimilation of WRF DA and WRF ETKF-3DVAR hybrid assimilation systems, parameter settings for forecasts, new data assimilation methods and system operation, result analysis, and coupling with other modules.
Assimilation analysis of heavy rain weather process
Assimilation analysis of typhoon weather process
3. Practical application of PMF source analysis technology for atmospheric particulate matter
At present, atmospheric particulate matter pollution has become an urgent environmental problem that needs to be solved in our country. Particulate matter pollution not only has an important impact on the climate and environment, but also seriously damages human health, especially in some heavily polluted weather, such as haze and sandstorms. In order to effectively and accurately control regional atmospheric particulate matter pollution, we first need to understand the source of particulate matter. Therefore, particulate matter source analysis has become a key technology to solve the problem of atmospheric particulate matter pollution. In order to help scientific researchers learn the basic theoretical knowledge and corresponding technologies of atmospheric particle source analysis more systematically, the "Advanced Training Course on Practical Technology Application of Atmospheric Particle Source Analysis" is specially held to help students master the basic knowledge of the physical and chemical properties of atmospheric particles and pass PMF method is a technology for source analysis. This training adopts a combination of "theoretical explanation + case practice + hands-on practice + discussion and interaction" to explain the basic knowledge of atmospheric particulate matter PMF source analysis technology in a simple and easy-to-understand manner and the experience and skills that need to be mastered in its application.
[Brief description of content]:
Chapter 1. Brief introduction to PMF source analysis technology and its input file preparation
1. What are the methods for analyzing air pollution sources?
2. What are the application conditions of each of these methods and their advantages and disadvantages?
3. Basic knowledge of atmospheric particulate matter and the main sources of each component Sources
of atmospheric particulate matter:
The composition of atmospheric particulate matter:
4. Introduction to PMF source analysis technology
5. Download and installation of PMF source analysis software
6. Obtaining the concentration of particulate matter components input by PMF
7. Preparation of PMF source analysis input files
Chapter 2: Principles of PMF source analysis technology, practical operation and application examples of PMF software
1. Basic principles of PMF source analysis
2. Basic operation of PMF source analysis software
3. Selection of PMF source analysis factors
4. PMF source analysis results and significance
Chapter 3: Optimization and error assessment of PMF source analysis results
1. Fpeak mode operation
2. Fpeak mode operation results
3. Introduction to error assessment method
4. Results of error assessment method
4. Practical technical application of air pollution diffusion model Calpuff
At present, air pollution is still an environmental problem that needs to be solved urgently in our country. In order to understand the impact of atmospheric pollutant emissions on the surrounding environment, it is necessary to understand the diffusion rules of pollutants. The Calpuff model is a three-dimensional unsteady Lagrangian diffusion model that can effectively handle the long-distance transport of pollutants in unsteady states (such as smoke, circulation, terrain, and coasts, etc.) and simulate pollutant concentrations. Prediction to better determine the source of pollutants at receptor sites. The model mainly includes: terrain and meteorological data preprocessing module, Calmet module, Calpuff module and Calpost module. At present, this model has been widely used in atmospheric environmental quality impact assessment and scientific research.
[Brief description of content]:
Chapter 1, Basic knowledge of Calpuff
1. Introduction to Calpuff model
2. Basic theory of Calpuff model
3. Calpuff model download and installation
1) Calpro system installation
2) Installation environment requirements
3) Auxiliary software that needs to be installed
Chapter 2, Data Preprocessing
1. Grid setting
2. Geographic data preprocessing
1) Terrain data preprocessing
2) Land use data preprocessing
3) Geographic data synthesis 3. Meteorological data
preprocessing 1
) Ground meteorological data preprocessing
2) High altitude meteorological data preprocessing
Chapter 3, Calmet Meteorological Module
1. File/information input
1) Input file
2) Input shared grid data
3) Fill in operation information
4) Mixing layer height
5) Temperature and relative humidity settings
6) Wind field input
7) Weather station data input
2. Operation
3. Output
case: input and operation of specific data of Calmet model in a certain place
Chapter 4, Calpuff module
1. File/information input
1) Input file
2) Input shared grid data
3) Fill in running information 4
) Grid settings 5
) Simulate pollutant species input 6
) Chemical transformation method selection
7) Dry/wet deposition selection
8) Model Parameter selection
9) Source data input
10) Receptor point position input
2, run
3, output
case: Calpuff model specific data input and run results
Chapter 5, Calpost post-processing module
1. File/information input
1) Processing option selection
2) Pollutant name and concentration field data input
3) Output option selection
2. Run
3. Data analysis
case: Calpost post-processing module specific data input and operation result analysis
Chapter 6, Post Tools post-processing tools and drawing tools
1. Post Tools post-processing tools
1) Prtmet meteorological post-processing module
2) Append post-processing module
3) Postutil post-processing module
4) Calsum post-processing module
2. Drawing tools
Case: concentration data drawing example
5. Application of the latest CAMx-Python fusion technology and application of air pollution source analysis methods
With the rapid development of our country's economy, our country is facing increasingly serious air pollution problems. Air pollution is the comprehensive result of human activities in industrial and agricultural production, life, transportation, urbanization and other aspects. At the same time, meteorological factors are the key natural factors to control air pollution. Air pollution problems are local, local, regional, and even global. In addition to having a serious impact on the local area, local pollutant emissions will also greatly affect the atmospheric environment in downwind areas due to power transmission. Numerical model simulation is an important tool for analyzing the spatial and temporal distribution and compositional contributions of air pollutants. The simulation results can be used to analyze the sources, causes, pollution levels, duration, main components, relative contributions and other issues of air pollution, which is helpful for analysis and reasonable control. Pollution source emissions provide reference for industrial adjustment. Currently, a variety of air quality models have been developed at home and abroad based on different theories, uses, and design concepts. These models are widely used in the establishment of air quality forecast and warning systems, air pollution prevention and control, environmental impact assessment and other work.
The CAMx model is an atmospheric pollutant calculation model based on atmospheric chemistry for ozone, particulate matter and haze weather processes. This model is continuously developed and improved by the Ramboll technical team with the support of the US Environmental Protection Agency and many state environmental protection departments. The U.S. Environmental Protection Agency uses CAMx to evaluate the ozone and PM concentration reduction effects brought about by national emission reduction plans, and many states use CAMx to develop local emission reduction plans. In the past 20 years, this model has also been gradually applied to many countries and regions in Asia (including China), Europe, Africa, Australia and the Americas.
[Brief description of content]:
CAMx pattern framework, application case analysis and local case configuration instructions
1. CAMx model framework
2. CAMx application case analysis
3. Multiple nested simulation area configuration instructions and methods
4. CAMx pollution source input file production technology based on SMOKE model
Explanation of basic Linux operations, CAMx mode compilation technology, and air quality simulation case analysis and operation
1. Explanation of basic Linux operating commands and installation of dependent libraries
2. CAMx mode compilation, installation and test case running
3. CAMx input preprocessing tool compilation technology
4. CAMx input file preparation
5. Air quality simulation case operation explanation and analysis
CAMx extension and detection tool capabilities and usage in pattern debugging and case analysis
Each tool function is based on its usage in pattern debugging and case analysis.
1. CAMx Expansion and Probing Tools (Probing Tools)
2. Ozone/Particulate Matter Source Apportionment Tool (SA)
3. Sensitivity analysis tools: DDM/HDDM
4. Process analysis tools (PA: IPR/IRR and CPA)
5. Reaction tracers (RTRAC)
Air pollution source analysis case operation (Ozone/Particulate Matter Source Analysis Tool (SA))
Ozone/Particulate Matter Source Analysis Tool (SA)
1. CAMx-SA tool compilation
2. CAMx-SA tool input file preparation (1) area map
3. CAMx-SA tool input file preparation (2) emission groups
4. CAMx-SA case configuration and operation
5. Post-processing and interpretation of simulation results
Sensitivity analysis and tool operation and simulation result post-processing technology and result interpretation
Sensitivity analysis tools (DDM/HDDM)
1. CAMx-DDM tool compilation method
2. CAMx-DDM tool input file preparation
3. CAMx-DDM case configuration and operation method
4. Simulation result post-processing technology and result interpretation
Running process analysis tools and interpreting results
Sensitivity Analysis Tool (PA)
1. CAMx-PA tool compilation method
2. CAMx-PA tool input file preparation
3. CAMx-PA case configuration (IPR and CPA) process and operation method
4. Simulation result post-processing technology and result interpretation
CAMx data post-processing based on Python (1)
1. Introduction to Python
2. Basics of Python programming
3. Python environment installation
4. Data extraction methods and scripts
5. Evaluation of model simulation results
6. Time series analysis of pollutant concentration monitoring values and simulated values
CAMx data post-processing based on Python (2)
1. Concentration distribution diagram drawing
2. Graphical drawing of sensitivity analysis results
3. Graphical drawing of process analysis results
6. MCM box model practical technology application and O3 formation pathway, generation potential, and sensitivity analysis
At present, atmospheric ozone pollution has become an urgent environmental problem to be solved during my country's "14th Five-Year Plan" period. Ozone pollution not only has an important impact on climate, but also seriously damages human health and plant growth. In order to effectively and accurately control regional atmospheric ozone pollution, it is necessary to understand the main pathways of ozone generation and its precursors. The OBM box model can be used to simulate the occurrence and evolution process of photochemical pollution, study the generation mechanism of ozone and conduct sensitivity analysis, and explore the impact of precursor emissions on photochemical pollution. Box models usually consist of chemical mechanism, physical process, initial conditions, input and output modules, with chemical mechanism being its core part. MCM (Master Chemical Mechanism) contains about 6,700 organic compounds and about 17,000 reactions. It can describe the chemical processes of atmospheric gas phase organic compounds in detail and is widely used in the field of atmospheric science research.
Topic 1: Knowledge base of O3 formation in the atmosphere, principles of MCM and Atchem 2, and Linux system installation
1. Explanation of the principle knowledge of O3 formation in the atmosphere
2. Explanation of MCM principles and basic processes
3. Atchem 2 explanation and download and installation
4. Linux system installation
5. Other tools required to run Atchem 2
A. Fortran; B. Python; C. make, cmake
Topic 2: MCM modeling, data input, model operation and result output [explanation + case operation]
1. Establishment of MCM box model
1) Chemical mechanism
A. Facsimile format; B. RO2; C. Extraction of MCM
2) Setting of model parameters
3) Environmental variables
A. Temperature; B. Atmospheric pressure; C. Relative humidity; D. Water; E, solar altitude angle; F, boundary layer height; G, aerosol surface area;
H, diffusion rate; I, JFAC; J, Roof
4) Photolysis rate
A, constant photolysis rate; B, limited photolysis rate ;C Calculation of photolysis rate; D. JFAC calculation
5) Various config. files
2. MCM box model operation
3. MCM model operation result analysis
Case: Analysis of MCM box model operation results
Topic 3: O3 formation pathway, generation potential and sensitivity analysis [explanation + case operation]
1. O3 formation pathway
Case: Contribution of different reaction pathways to O3 formation
2. O3 sensitivity analysis I: Relative incremental reactivity method (RIR)
Case: Determine the main source of O3 through the calculation of RIR
3. O3 sensitivity analysis Ⅱ: EKMA curve drawing
1) Obtaining O3 contour data
2) EKMA curve drawing
Case: Determining the main source of O3 through drawing the EKMA curve
4. O3 generation potential
Case: Calculation of VOCS O3 generation potential
7. Improve practical application capabilities in the fields of meteorology, oceanography, and hydrology based on Python machine learning and deep learning technology
Python is a powerful, free, open source, object-oriented programming language that can be used on different operating systems and platforms. Its concise syntax and interpreted language make it an ideal scripting language. In addition to the standard library, there are also a wealth of third-party libraries. Python has excellent performance in data processing, scientific computing, mathematical modeling, data mining and data visualization. The above advantages make Python widely used in scientific research and engineering projects in the fields of meteorology, oceanography, geography, climate, hydrology and ecology. It is foreseeable that Python will become one of the mainstream programming languages in the fields of meteorology, oceanography, hydrology and other geosciences in the future.
Artificial intelligence and big data technology have achieved disruptive results in many industries. The meteorological and oceanographic fields have massive models and observation data, which are natural scenarios for the application of big data and artificial intelligence. Python is also currently the most popular language for machine learning and deep learning applications. For professionals in the field of meteorology and oceanography, Python is the first choice for machine learning and deep learning work.
Topic 1. Installation and Getting Started with Python Software
1.1 Python background and its application in meteorology
1.2 Anaconda interpretation and installation and Jupyter configuration
1.3 Python basic syntax
Topic 2. Commonly used scientific computing libraries for meteorology
2.1 Numpy library
2.2 Pandas library
2.4 Xarray library
Topic 3. Commonly used visualization libraries for meteorology and oceanography
3.1 Introduction to visualization libraries Matplotlib, Cartopy, etc.
3.2 Basic drawing
(1) Line chart drawing
(2) Scatter chart drawing
(3) Color filling/contour line
(4) Flow field vector diagram
Topic 4. Reptiles and Meteorological Ocean Data
(1) Introduction to the Request library
(2) Crawling the weather map of the Central Meteorological Observatory
(3) Crawling FNL data
(4) ERA5 download
Topic 5. Commonly used interpolation methods in meteorology and oceanography
(1) Regular grid data interpolation to sites
(2) Radial basis function RBF interpolation
(3) Inverse distance weight IDW interpolation
(4) Kriging interpolation
Topic 6: Basic Theory and Practical Operation of Machine Learning
6.1 Basic principles of machine learning
(1) Introduction to machine learning
(2) Integrated learning (Bagging and Boosting)
(3) Common model principles (random forest, Adaboost, GBDT, Xgboost, lightGBM)
6.2 Machine learning library scikit-learn
(1) Introduction to sklearn
(2) sklearn completes the classification task
(3) sklearn completes the regression task
Topic 7. Application examples of machine learning
This topic, on the basis of a detailed explanation of two types of integrated learning algorithms commonly used in machine learning, Bagging and Boosting, and an in-depth explanation of the two types of algorithms and their commonly used representative models, combines three learning examples and explains some common machine learning techniques in series. Integrate theory with practice.
7.1 Application of machine learning and deep learning in meteorology
Application of AI in meteorological model correction, short-term forecast, climate prediction and other scenarios
7.2 Correction of wind speed forecast by GFS numerical model
(1) Random forest selection of important features
(2) K nearest neighbor and decision tree model correction of wind speed
(3) Gradient boosting decision tree GBDT correction of wind speed
(4) Model evaluation and comparison
7.3 Intelligent correction of typhoon forecast data
(1) Introduction and preprocessing of CMA typhoon forecast data set
(2) Random forest model correction of typhoon forecast
(3) XGBoost model correction of typhoon forecast
(4) Typhoon "fireworks" forecast effect verification
7.4 Machine learning to predict wind power of wind farms
(1) lightGBM model to predict wind power
(2) Parameter adjustment tool—GridSearch verified on K-fold
Topic 8. Basic theory and practical operation of deep learning
8.1 Basic theory of deep learning
Explain the basic theoretical knowledge of deep learning, gain an in-depth understanding of the basic theory and working principles of machine learning, master how to construct and optimize neural network models (such as artificial neural network ANN, convolutional neural network CNN, recurrent neural network RNN, etc.), and improve the understanding of existing The ability to understand and apply deep learning algorithms and technologies can better cope with subsequent practical problems and applications in marine meteorology-related fields.
8.2 Pytorch library
(1) Introduction to sklearn, common functions and machine learning methods.
Learn the common functions of the classic machine learning library sklearn, such as obtaining public data sets such as iris and handwritten fonts, dividing training sets and test sets, model building and model verification, etc.
(2) Introduction to pytorch and building models
Learn the currently popular deep learning framework pytorch, understand tensor, automatic derivation, gradient boosting, etc., take the BP neural network to learn the sin function as an example, and master how to build single-layer and multi-layer neural networks, and how to use GPU for model operations. .
Topic 9. Application examples of deep learning
In this topic, on the basis of learning to use ANN to predict shallow water equations, we will further learn how to use the PINN method to add dynamic equations to the model to alleviate the problem of poor physical interpretability of deep learning. In addition, meteorological data is a typical spatiotemporal data. Learn the classic time series prediction method LSTM, and the spatial convolution algorithm UNET.
9.1 Deep learning to predict shallow water equation patterns
(1) Introduction to shallow water models and data acquisition
(2) Traditional neural network ANN learns shallow water equations
(3) Physical constraint network PINN learns shallow water equations
9.2 LSTM method to predict ENSO
(1) Introduction to ENSO and data introduction
(2) Introduction to the principle of LSTM method
(3) LSTM method to predict meteorological sequence data
9.3 Deep learning - convolutional network
(1) Introduction to convolutional neural network
(2) Prediction of radar echo by Unet
Topic 10. EOF statistical analysis
10.1 Introduction to EOF basics and eofs library
10.2 EOF analysis of sea surface temperature data
(1) SST data to calculate anomalies and remove trends
(2) SST performs EOF analysis and visualization
Topic 11. Pattern post-processing
11.1 WRF mode post-processing
(1) Introduction to wrf-python library
(2) Extracting site data
(3) 500hPa form field drawing
(4) Vertical profile - radar reflectivity as an example
11.2 ROMS mode post-processing
(1) xarray as an example to operate ROMS output data
(2) Vertical coordinate conversion, S coordinate to depth coordinate
(3) Vertical section drawing
(4) Horizontal coloring map drawing
8. FLEXPART Lagrangian particle diffusion model modeling technology and practical experience and skills in studying the source-sink relationship of atmospheric pollutants
At present, air pollution is one of the important environmental problems in our country. In order to effectively and accurately control regional air pollution, the sources of pollutants need to be clarified. The Lagrangian particle diffusion model FLEXPART describes the long-distance, mesoscale transport, diffusion, dry and wet deposition, and radiation attenuation of tracers in the atmosphere by calculating the trajectories of a large number of particles released from point, line, surface, or volume sources. process. This model can simulate the diffusion of tracers from the source area to the surroundings through time forward operation, and can also determine the potential source area distribution that has an impact on a fixed site through backward operation.
1. Explanation of Lagrangian particle diffusion model and FLEXPART model
1. Introduction and application characteristics of Lagrangian particle diffusion model
2. Introduction to FLEXPART model and its download and installation
2. Linux system and FLEXPART mode installation
1. Linux system installation
2. Basic Linux command exercises
3. FLEXPART dependent library installation
4. FLEXPART mode compilation
5. FLEXPART-WRF mode compilation
3. FLEXPART mode input and model parameter description
1. Meteorological field data acquisition
2. FLEXPART-WRF mode parameter description
3. FLEXPART mode parameter description
4. FLEXPART model result post-processing technical method
4. FLEXPART mode operation exercises
1. Start from scratch and establish a FLEXPARF-WRF operation case
2. Start from scratch and establish a FLEXPARF operation case
3. FLEXPART output result post-processing tool installation
4. FLEXPART post-processing tool usage exercises
5. FLEXPART application case configuration, operation and post-processing
1. Output area adjustment case (comparative analysis)
2. Pollutant diffusion and concentration analysis
3. Pollution source area analysis
4. Air mass trajectory analysis
5. Greenhouse gas simulation case
9. Based on SMOKE multi-mode emission inventory processing technology and EDGAR/MEIC inventory production and VOCs emission accounting
With the rapid development of our country's economy, our country is facing increasingly serious air pollution problems. In recent years, serious air pollution problems have significantly affected the national economy and people's livelihood, attracting increasing attention from the government, academic circles and people. Air pollution is the comprehensive result of human activities in industrial and agricultural production, life, transportation, urbanization and other aspects. At the same time, meteorological factors are the key natural factors to control air pollution. Air pollution problems are local, local, regional, and even global. In addition to having a serious impact on the local area, local pollutant emissions will also greatly affect the atmospheric environment in downwind areas due to power transmission. Numerical model simulation is an important tool for analyzing the spatial and temporal distribution and compositional contributions of air pollutants. The simulation results can be used to analyze the sources, causes, pollution levels, duration, main components, relative contributions and other issues of air pollution, which is helpful for analysis and reasonable control. Pollution source emissions provide reference for industrial adjustment. Currently, a variety of air quality models have been developed at home and abroad based on different theories, uses, and design concepts. These models are widely used in the establishment of air quality forecast and warning systems, air pollution prevention and control, environmental impact assessment and other work.
Atmospheric pollutant emissions are the source of air pollution, and meteorological factors are important factors affecting the degree of pollution. Therefore, the air quality model requires meteorological data and pollutant emission inventories as inputs. Due to the complexity of air pollution sources, data lag, dynamic changes, and regularity The characteristics of the air pollution source emission inventory input are not obvious, making the preparation of the air pollution source emission inventory input a key and difficult point.
【Target】:
1. Master the sources of uncertainty and quantitative analysis methods for air pollution source emission inventories;
2. Take VOCs emissions as an example to master emission source accounting and component inventory establishment methods;
3. Master the technical methods for processing air pollution source emission inventories based on the SMOKE model;
4. Master the multi-mode emission inventory input preparation method based on SMOKE;
5. Master the application method in CMAx, CMAQ, WRF-chem and other modes through case analysis and practice;
6. Master the EDGAR/MEIC inventory processing technology method through practical examples;
[Brief description of content]:
| Air pollution source emission inventory preparation, processing and uncertainty analysis technical methods; | 1. Explanation of the preparation method of urban air emission inventory; 2. Emission inventory uncertainty analysis technical methods; 3. Emission inventory uncertainty and model debugging; 4. Emission inventory processing technology explanation and process; 5. SMOKE/MEGAN model explanation; |
| Emissions accounting (VOCs emissions as an example) and uncertainty analysis technology; | Operation: 1. Pollutant emission accounting (VOCs emissions from solvent use sources as an example); 2. Component emission list establishment (VOCs component list as an example); 3. Basic command operations of LINUX; 4. Emission coefficient uncertainty analysis operation; 5. Emission inventory uncertainty analysis operations; |
| SMOKE local case establishment method; | 1. Case method based on SMOKE local emission inventory processing; 2. Pollution source classification and document preparation methods; 3. Area code compilation and document preparation methods; 4. Investigation of time characteristics of air pollution sources and preparation methods of time spectrum files; 5. Pollution source composition spectrum and file preparation method based on model atmospheric chemical mechanism; 6. SMOKE input preparation using grid emission inventory as input |
| SMOKE/MEGAN installation and testing | Operation: 1. SMOKE environment configuration and precautions; 2. SMOKE output drawing tool installation 3. SMOKE model installation and compilation 4. SMOKE test case operation; 5. MEGAN model local case; |
| SMOKE input file preparation exercise | Operation: 1. Spatial Allocator installation and configuration 2. Establishment of area source spatial distribution coefficient 3. Time spectrum file preparation operation |
| SMOKE local case configuration and debugging operations | Operation: 1. Preparation of point source and area source emission inventory input files; 2. SMOKE local case configuration and debugging; 3. Emission inventory integration and CMAQ-ready emission inventory output; |
| CMAQ/CAMx case execution | Operation: 1. CAMx mode data conversion interface installation 2. Multi-mode simulation area setting instructions 3. CMAQ case running 4. CAMx case running |
| WRF-Chem mode case run | Operation: 1. Install WRH-Chem mode data conversion interface 2. Run WRF-Chem case 3. Ask questions |
| Global Emissions Inventory (EDGAR) processing case based on SMOKE | Operation: 1. Case configuration instructions 2. Input file preparation 3. SMOKE-EDGAR case running 4. Result check |
| National Emissions Inventory (MEIC) processing case based on SMOKE | Operation: 1. Case configuration instructions 2. Input file preparation 3. SMOKE-MEIC case running 4. Result check |
10. Python technology practice in automated operation and pre- and post-processing of WRF models
Today's personnel engaged in meteorology and related fields often involve meteorological numerical models and their data processing. Whether as a means of operational forecasting or as a scientific research tool, mastering meteorological numerical models and efficient pre- and post-processing languages is a very important skill. . As a leader in mesoscale meteorological numerical models, WRF has complete model functions and is the first choice for most people. Mastering the model is only the first step. It is also an important task to further process the results of the numerical model into products needed for our business or scientific research. In terms of current trends, the Python language has become the first choice.
For most people, especially new users, the installation of WRF mode is cumbersome and unnecessary. It can be used as a skill to be mastered later. This course skips the tedious installation steps and focuses directly on the running part of the mode. Through short and quick steps, Teaching, quick master mode operation. Furthermore, the python language is combined with WRF mode operation to automate the mode operation and improve the efficiency of scientific research and business. At the same time, master the common scenarios of python in WRF pre- and post-processing, including data processing, visual drawing, etc.
Mastering the combined application of WRF model + Python language can be effectively used in business, scientific research and engineering projects in the fields of meteorology, oceanography, geography, climate, hydrology and ecology.
[Brief description of content]:
Topic 1. WRF Basics and
Linux Basics
1 WRF Basics and Linux Basics
1.1 Theoretical knowledge and application prospects of WRF mode
Introduction to WRF dynamic theory
WRF mode code framework
WRF mode application prospects
1.2 Linux knowledge and basic operations
Linux basic commands
Set up scheduled tasks
and supercompute submission Task
1.3 WRF mode installation
Basic library installation (GNU compiler, zlib, libpng, jasper, hdf5, netcdf4, mpich2)
WRF/WPS installation
Topic 2. WRF mode operation
2.1 WRFDomainWizard setting simulation area
2.2 WPS (geogrid, ungrib, metgrid) pre-processing
GFS/FNL/ERA5 data-driven, nested simulation
2.3 WRF (real.exe, wrf.exe) integration operation
parameterization scheme setting
Experience in using common parameters
2.4 Restart
2.5 How to get better WRF simulation results (WRF running experience and precautions)
Topic 3. Python basics
3.1 Python installation and basic syntax
3.2 Python commonly used meteorological data processing libraries
numpy, datetime, pandas, scipy, netcdf-python
3.3 Python meteorological drawing basics
1) Line chart drawing
2) Color filling and isoline + map
3) Flow field vector + map
4) Map drawing (cartopy)
Topic 4. Introduction to WRF application cases
4 WRF cases
4.1 Typhoon simulation in the northwest Pacific
4.2 Severe weather process-hail case simulation
4.3 WRF-solar simulation of solar radiation
Topic 5. python helps WRF automated operation
5 python helps WRF automated operation
5.1 python automatically runs WRF historical examples
5.2 Build an automated WRF operational forecast system
1) python automatically downloads GFS real-time forecast data
2) python automatically runs the WRF forecast system
3) Python automatically sends emails prompting WRF running results
Topic 6. WRF model pre- and post-processing
6 WRF model pre- and post-processing
6.1 Introduction to WRF post-processing 6.2
Introduction to wrf-python library
6.3 python draws WRF simulation area and terrain
6.4 python extracts site forecast elements and draws
6.5 python draws 500hPa high-altitude form field
6.6 python draws radar Reflectance DBZ vertical profile
6.7 python draws 700hPa water vapor field
6.8 python draws surface rainfall field
11. Practical technical applications of common earth science data processing
In earth sciences, different data are stored in a variety of data formats based on the characteristics of the specific discipline. In scientific research work, it is necessary to comprehensively use and analyze a variety of data. Therefore, it is necessary to find and learn common data format solutions and focus research energy on specific scientific issues.
In response to the above problems, this content selects common data in atmospheric science, hydrology and ecology for explanation.
The main formats explained:
l Global atmospheric reanalysis data netCDF
l Snow depth ASCII
l Snow cover ASCII/TIFF
l Sea temperature data netCDF
l Vegetation index data NDVI
l Land use data HDF
Required processing tools:
Anaconda 5.0+ (python 3.6), xarray==0.13, netcdf4==1.5.3, rasterio, pandas, pyhdf, fiona, shapely, gdal
12. Application of regional meteorology-atmospheric chemistry online coupling model (WRF/Chem) in the atmospheric environment
With the rapid development of our country's economy, our country is facing increasingly serious air pollution problems. In recent years, serious air pollution problems have significantly affected the national economy and people's livelihood, attracting increasing attention from the government, academic circles and people. Air pollution is the comprehensive result of human activities in industrial and agricultural production, life, transportation, urbanization and other aspects. At the same time, meteorological factors are the key natural factors to control air pollution. Air pollution problems are local, local, regional, and even global. In addition to having a serious impact on the local area, local pollutant emissions will also greatly affect the atmospheric environment in downwind areas due to power transmission. Numerical model simulation is an important tool for analyzing the spatial and temporal distribution and compositional contributions of air pollutants. The simulation results can be used to analyze the sources, causes, pollution levels, duration, main components, relative contributions and other issues of air pollution, which is helpful for analysis and reasonable control. Pollution source emissions provide reference for industrial adjustment. The simulation results can analyze cross-regional pollutant transport issues and calculate the dry and wet deposition fluxes of carbon, nitrogen and other components, thereby estimating the potential impact of atmospheric pollutants on ecosystems such as vegetation and soil. The simulation results can also be further analyzed and applied in the fields of environmental and energy assessment, environmental assessment and planning, industrial structure, environmental carrying capacity changes, ecosystem stability and changes, etc.
[Learning Objectives]:
1. Master the principles, debugging, and operation methods of WRF-Chem mode.
2. Master WRF-Chem mode data preparation, pre-processing and related parameter setting methods through case operations.
3. Master the methods of post-processing and drawing of simulation results (software operations such as ARWPOST and NCL).
4. Master the application of WRF-Chem in atmospheric environment (PM2.5, ozone), visibility, and urbanization through case analysis operations.
5. This course provides guidance on the problems encountered by students in actual projects.
[Expert]:
The lecturer expert is a teacher from a university. He has been engaged in research on atmospheric environment, aerosol simulation and radiation effects for a long time. He has rich practical experience and has presided over and participated in the completion of a number of scientific research projects.
[Brief description of content]:
13. Air quality simulation and pollution source analysis technology and case analysis based on CAMx
With the rapid development of our country's economy, our country is facing increasingly serious air pollution problems. Air pollution is the comprehensive result of human activities in industrial and agricultural production, life, transportation, urbanization and other aspects. At the same time, meteorological factors are the key natural factors to control air pollution. Air pollution problems are local, local, regional, and even global. In addition to having a serious impact on the local area, local pollutant emissions will also greatly affect the atmospheric environment in downwind areas due to power transmission. Numerical model simulation is an important tool for analyzing the spatial and temporal distribution and compositional contributions of air pollutants. The simulation results can be used to analyze the sources, causes, pollution levels, duration, main components, relative contributions and other issues of air pollution, which is helpful for analysis and reasonable control. Pollution source emissions provide reference for industrial adjustment. Currently, a variety of air quality models have been developed at home and abroad based on different theories, uses, and design concepts. These models are widely used in the establishment of air quality forecast and warning systems, air pollution prevention and control, environmental impact assessment and other work.
The CAMx model is an atmospheric pollutant calculation model based on atmospheric chemistry for ozone, particulate matter and haze weather processes. This model is continuously developed and improved by the Ramboll technical team with the support of the US Environmental Protection Agency and many state environmental protection departments. The U.S. Environmental Protection Agency uses CAMx to evaluate the ozone and PM concentration reduction effects brought about by national emission reduction plans, and many states use CAMx to develop local emission reduction plans. In the past 20 years, this model has also been gradually applied to many countries and regions in Asia (including China), Europe, Africa, Australia and the Americas.
[Learning Objectives]:
1. Master the technical method of regional air quality simulation case configuration based on the CAMx model;
2. Master the input preparation method of the CAMx model atmospheric emission inventory based on the SMOKE model;
3. Master the pollution source analysis tool (SA) based on the CAMx model Case configuration technical methods.
[Expert]: A senior expert from the field of air pollution source research. He has long been engaged in the preparation of air pollution source emission inventories, pollution source analysis research, air emission inventory processing technology methods, and air quality model application research. He has rich experience in air emission source inventory processing and pollution sources. Analytical and air quality modeling experience.
[Brief description of content]:
14. Systematic learning of CMAQ air quality model
The issue of air pollution has increasingly attracted the attention of governments at all levels and the public. From the release of real-time data monitoring to the release of air quality numerical forecasts and forecast products, my country has made certain progress in air quality monitoring and forecasting. With the rapid development of computer technology, the improvement of air pollution monitoring methods, and people's in-depth understanding of atmospheric physical and chemical processes, the development and use of advanced atmospheric chemistry models to predict China's air quality is of great significance for reducing air pollution disasters and improving people's lives. Quality has a positive meaning. The air quality forecasting model system (WRF-CMAQ) and pollution source treatment technology are currently important means of early warning and comprehensive management of large-scale haze weather. Its promotion and application across the country will help improve the actual business forecasting level and enhance disaster prevention and reduction capabilities. , achieve significant social and economic benefits.
[Expert] :
Teacher Cheng [Researcher]: He has been engaged in scientific research on air quality simulation and improvement for a long time. He has independently developed a variety of pollution source inversion methods and aerosol lidar data assimilation technology based on three-dimensional variation. He has more than ten years of experience in using the CMAQ model. Practical operating experience, taught CMAQ model in many universities. Familiar with all aspects from model installation, operation, interpretation to model result display, with special attention to the application and promotion of CMAQ models in the fields of environment, meteorology and other fields.
[Brief description of content]:
15. High-precision weather simulation software WRF (Weather Research Forecasting) technology and case applications
Climate is the main driving factor in many fields (ecology, water resources, wind resources, carbon neutrality, etc.). A reasonable understanding of climate change is conducive to explaining the mechanisms and processes of ecological environment changes, and understanding current and future climate change is crucial. It is a prerequisite for ecological, environmental and energy assessment and carbon policy planning, and climate simulation is the most important means to obtain high-precision climate information. Modern ecology, hydrology, new energy and carbon neutral fields require sub-kilometre and higher resolution meteorology. For simulation, the WRF model is the most widely used meteorological model at home and abroad. There are more and more applications using this model for high-precision simulations of even hundreds of meters. On the other hand, the model continues to expand the model module, and now it can realize aerosol and chemical processes (WRF-CHEM), data assimilation (WRF-DA), hydrological processes (WRF-HYDRO), urbanization (URBAN) and other processes The precise simulation has more and more application scope and practical business and scientific research applications.
Note: Please prepare your own computer and install the required software in advance.
16. CMIP6 data processing
The International Coupled Model Intercomparison Program has entered a new phase, Phase 6 (CMIP6), which will provide richer global climate model data for climate change research. Compared with CMIP5, the CMIP6 model has two main characteristics: first, the processes considered by CMIP6 are more complex, and many models realize two-way coupling of atmospheric chemical processes; second, the resolution of the atmosphere and ocean models has been significantly improved, among which the atmosphere The maximum horizontal resolution of the model can reach 25km around the world. In addition, the RCP scenario of CMIP5 only considers the goal of achieving stable CO2 concentration and corresponding radiative forcing in the next 100 years, and does not target a specific social development path. The new shared socioeconomic path in CMIP6 fully considers this and provides With the addition of more diverse emission scenarios, more reasonable simulation results can be provided for mitigation and adaptation research and regional climate projections, thus making up for the shortcomings of the RCP scenarios in CMIP5 to a large extent.
In the International Coupled Model Comparison Program, GCMs provide global large-scale information for constructing climate change. However, when conducting climate research at the regional scale, relatively low-resolution information produces large deviations in regional climate change projections. Scaling methods play an important role in transforming large-scale information into regional scales, including dynamic downscaling, statistical downscaling, and methods combining the two.
Supporting data:
Complimentary CMIP6 monthly data (500G+)
Complimentary CMIP6 daily data (1.8T+
Free global VIPPHEN phenology data (40G+)
Free ERA5-LAND land surface reanalysis data (about 5T)
Contains 11 variables: temperature, air pressure, radiation, evaporation, precipitation, humidity
Brief description of content:
Note: Please prepare the required learning environment by yourself.