Simulating large emitters using CMAQ and a local
- Slides: 30
Simulating large emitters using CMAQ and a local scale finite element method. Analysis in the surroundings of Barcelona Albert Oliver, Raúl Arasa, Agustí Pérez-Foguet, Mª Ángeles González HARMO 17 Budapest May 2016
Motivation To Improve the prediction at fine scales: • Large emitters • Near source transport and chemistry Current approaches: • Nested grid modelling • Adaptive grid modelling • Hybrid modelling • Plume-in-grid modelling • Statistical models • CFD HARMO 17 · May 2016 · Budapest · 2
Proposed approach In this work we propose to compare two different approaches • WRF-ARW/AEMM/CMAQ • Nested grid modelling • 1 km – 300 m • Adaptive Finite Element method • Plume-in-grid modelling HARMO 17 · May 2016 · Budapest · 3
Proposed approach WRF-ARW/AEMM/CMAQ Meteorology Transport reaction Emissions Initial and Boundary conditions Local FEM Mass consistent model (Wind) Plume rise (Briggs) Transport and reaction HARMO 17 · May 2016 · Budapest · 4
Outline WRF-ARW/AEMM/CMAQ model Local scale Finite Element model Application to Barcelona surroundings Conclusions HARMO 17 · May 2016 · Budapest · 5
WRF-ARW The mesoscale meteorological model used is Weather Research and Forecasting—Advanced Research (WRF-ARW) version 3. 6. 1 Specially suited to the subscale grid modelling Time-splitting methods, and high order (both time and space) HARMO 17 · May 2016 · Budapest · 6
AEMM Air Emission Model of Meteosim (AEMM v 3. 0) developed by Meteosim S. L. Numerical, deterministic, Eulerian, local-scale model It allows to obtain the intensity of emissions in different areas, either anthropogenic (traffic, industry, residential, etc. ) or natural (emissions caused by vegetation or erosion dust) for the area of interest HARMO 17 · May 2016 · Budapest · 7
CMAQ CMAQ v 5. 0. 1 CB-5 chemical mechanism AERO 5 aerosol module EBI solver Discretization 1 km The WRF-ARW/AEMM/CMAQ approach has been used successfully in urban areas (Catalonia, Madrid), industrial areas (Tarragona, Ponferrada), and arid areas (Perú, Chile) HARMO 17 · May 2016 · Budapest · 8
Adaptive Finite Element Method Convection-Diffusion-Reaction equation Wind field Plume rise FEM discretization Adaptivity HARMO 17 · May 2016 · Budapest · 9
Adaptive finite element method Convection – diffusion – reaction equation A. Oliver et al. Adaptive Finite Element Simulation of Stack Pollutant Emissions over Complex Terrains. Energy 2013. HARMO 17 · May 2016 · Budapest · 10
Wind field Interpolate Wind field from WRF-ARW Mass-consistent model HARMO 17 · May 2016 · Budapest · 11
Plume rise Briggs equations • Buoyancy Momentum HARMO 17 · May 2016 · Budapest · 12
FEM discretization • Temporal discretization: Crank-Nicolson • Spatial discretization: Least Squares FEM • System solver: Conjugate gradient preconditioned with an Incomplete Cholesky Factorization HARMO 17 · May 2016 · Budapest · 13
Adaptivity Mesh adaptation • Mesh is only adapted to topography and plume rise • Necessity to adapt to the solution • Error indicator using log (wide range of solutions) Mesh refinement fixing a minimum size L. Monforte and A. Pérez-Foguet. A multimesh adaptive scheme for air quality modeling with the finite element method. Int. J. Numer. Meth. Fluids 2014 HARMO 17 · May 2016 · Budapest · 14
Application Barcelona surroundings HARMO 17 · May 2016 · Budapest · 15
Results Barcelona surroundings CMAQ nested domains HARMO 17 · May 2016 · Budapest · 16
Results Zoom of the nested domain HARMO 17 · May 2016 · Budapest · 17
Results Day: 2/12/2013 • High concentration levels Simulation 48 h (24 h spin up) CMAQ grid resolution 1 km, 32 layers FEM domain 20 x 20 km, resolution from 1 km to ~1 m HARMO 17 · May 2016 · Budapest · 18
FEM Mesh HARMO 17 · May 2016 · Budapest · 19
FEM Mesh HARMO 17 · May 2016 · Budapest · 20
Max. 1 h levels FEM CMAQ HARMO 17 · May 2016 · Budapest · 21
Measurement stations HARMO 17 · May 2016 · Budapest · 22
Near emitter station HARMO 17 · May 2016 · Budapest · 23
Near emitter station Streamlines from the emitter HARMO 17 · May 2016 · Budapest · 24
Near emitter station Streamlines from the emitter Hour description of wind not enough HARMO 17 · May 2016 · Budapest · 25
Near emitter station HARMO 17 · May 2016 · Budapest · 26
Station distant to the emitter HARMO 17 · May 2016 · Budapest · 27
Station distant to the emitter HARMO 17 · May 2016 · Budapest · 28
Conclusions Combination of a nested grid and a local scale finite element model is a promising approach The WRF-ARW/AEMM/CMAQ system using a 1 km grid captures the observed data far from the emitter Near the emitter, the finite element model is closer to the measured data, while far from the emitter the CMAQ-ARW model is better. HARMO 17 · May 2016 · Budapest · 29
Future work Use an smaller resolution for the wind field in the local scale model. Explore how to combine both models, operationally, in a hybrid model. HARMO 17 · May 2016 · Budapest · 30
Late post mortem changes
Conditions simulating rigor mortis
Local vertical local horizontal frame
A think local act local multicountry type of strategy
A "think local, act local" multidomestic type of strategy
Two drawbacks of a think local act local
Large margin classification using the perceptron algorithm
Using system.collections.generic
Defrost using internal heat is accomplished using
Local maximum
State and local taxes and spending
A large heavy truck and a small baby carriage
How to know if an effect size is small medium or large
Empty
Much comparative
Superlative form warm
Large comparative and superlative
By and large what is the cause of intentional rating errors
Large and fast: exploiting memory hierarchy
Kitchen matching game
A large truck and a volkswagen have a head-on collision
In 479 bc a large unified greek army
Journey to digestion
How to remember reactivity series of metals
Once upon a time short story moral
Littauer suture removal
Scott bainbridge md
Dna substitution
Global alignment
Oracle grants management
Learning with local and global consistency
