Local Air Pollution Modeling Local Air Pollution Modeling AIM/Air - PowerPoint PPT Presentation

Local Air Pollution Modeling Local Air Pollution Modeling AIM/Air AIM/Air Takeshi Fujiwara and Yuzuru Matsuoka Takeshi Fujiwara and Yuzuru Matsuoka 2003/03/13 2003/03/13 th AIM International Workshop The 8 th AIM International Workshop The 8

AIM/Air AIM/Air • Introduction of an air quality modeling in the framework of AIM AIM family • One of supplementary models of AIM/Enduse AIM/Enduse • Consolidated with the emission inventory and energy bottom-up model, multi-scale multi-species model

Framework of AIM/Air Source models AIM/Local and In door receptor model Inventory Receptor models Human health Residential impacts Source recepter Emission Inventory matrix Water bodies Air quality acidification Commercial /eutrophication Air Model Soils Long range Industry acidification transport (LRT) /eutrophication model Forests Transport LPS model acidification /eutrophication Area Source model Energy

Characteristics • Sources : Large point sources (tall stack >200m) and area sources. Line sources are included in area sources. • Target pollutants : SO 2 , NO 2 , SPM • Coupling of models : Point source model (LPS), area source model (AS) and long-range transport model(LRT) • Multi-scale description : x km(LPS) + m x km(AS) + mn x km(LRT) grids (integer m , n )

Modeling concept(1) Long range transportation model Height 3 Height 2 Height 1 AS(Euler) model LPS (plume and puff) model

Modeling concept(2) 50km 50km 3 Layered Euler model 1 Layered 3 rd layer AS model (~5000m) (Euler) 2 nd layer (~3000m) 1 st layer LPS model (~1000m) (plume and puff) 10km Vertical Relation between AS model and LRT model

Modeling concept(3) LPS cell 30km 10km 50km AS 10km cell 50km LRT 30km cell Relation among LPS grid, AS grid and LRT grid

General Flow of AIM/Air LPS data LPS model Large point Plume and Puff model for sources data pollutants emitted from large point sources MASCON AS model GIS ECMWF Wind velocity is A diffusion model in the Geographic Meteorology data interpolated lower mixing layer for Information of the world. based on air mass pollutants emitted from System balance area sources AS data LRT model Area sources data A model for Long-Range Transportation of air pollutants.

AIM/Air Data & Process ECMWF GRIBEX emit.as emit.lps ReadGRIB u.d v.d w.d prep.d cloud.d geo.d sg.d rad.d temp.d interpAS interpLRT interpLPS stabLPS balanceAS balanceLRT wind.lps wind.as wind.lrt diffusionAS diffusionLRT diffusionLPS

Data & Process (Weather) ECMWF GRIBEX ReadGRIB u.d v.d w.d prep.d cloud.d geo.d sg.d rad.d temp.d Precipitation Wind velocity(u,v,w) Geo potential Total cloud cover Surface Geo potential Solar radiation Temperature

Data & Process (LPS) emit.lps LPS location Stack height Emission intensity u.d v.d w.d cloud.d Gas flow rate geo.d sg.d rad.d temp.d interpLPS stabLPS Interpolation of wind velocity at LPS site and stack height wind.lps Calculation of atmospheric stability diffusionLPS

Data & Process (AS) emit.as Emission intensity at each cell u.d v.d w.d geo.d sg.d interpAS Primary interpolation of wind velocity balanceAS Secondary interpolation of wind velocity using mass balance equation (MASCON) wind.as diffusionAS

emit.as emit.lps Data & Process (LRT) u.d v.d w.d prep.d geo.d sg.d Primary interpolation of wind velocity interpLRT Secondary balanceLRT interpolation of wind velocity using mass balance equation wind.lrt (MASCON) diffusionLRT

Primary interpolation { ( ) } ( ) ( ) 1 = × − × + − × u a E E u E E u α β β α − G , a G , hPa hPa G , hPa G , a hPa E E β α G , hPa G , hPa γ hPa γ hPa P r e s s u r e l e v e l Interpolation β hPa β hPa Geopotential P r e s s u r e l e v e l α hPa α hPa 3000m 3000m P r e s s u r e l e v e l Surface 1000m 1000m Geopotential N ( ) ∑ λ u x , y , z i i i i ( ) ( ) ( ) ( ) = = λ = − + − + − 2 2 2 i 1 u x , y , z , 1 x x y y z z i i i i N ∑ λ i = i 1

LPS model (1) • Corresponding to tall stack emissions (larger than 100 ～ 200m). Emission input to this LPS model is not included in the AS model. • A plume model (windy) or puff model (no wind). • Complex terrain modification is similar to EPA’s ISC3. • Calculate concentration at each cell (1km x 1km), within 10 km to every directions near the emission sites, hour by hour through a year.

LPS model (2) Plume model ( ) ( )         − +  2 2  2 ( ) Q y z H z H       = − − + −   P e e C x , y , z exp exp       πσ σ σ σ σ   2 2 2 2 u     2  2 2    y z y z z x : downstream coordinate, y : horizontally transverse coordinate, z : vertical coordinate (representative height=1.5m, σ y , σ z ： diffusion coefficients and calculated with the next equations, Q p : emission from LPS, u : wind velocity, H e : effective height of stacks α α σ = γ ⋅ σ = γ ⋅ x y , x z y y z z γ y, γ z , α y, α z : parameters given with Pasquill-Gifford diagram. He is given by the next CONCAWE equation. Q H is a heat emission (cal/s), = + ∆ H H H e and u h is a wind velocity at the outlet − 3 1 ∆ = 0 . 175 H Q u 2 4 H h ( ) = ρ − Q C q T T H p g g 0

LPS model (3) Puff model in case of no wind/weak wind     −  π ⋅     2 2 2 ux 1 u x ux      +    1 exp erfc     η αη α η αη  2  2 2     −  2   2   2 ( ) Q u − − − −   =   P C x , y , z exp   ( ) α   3 2 π γ   π ⋅ −  2       2 2 2 2 2 ux 1 u x ux     + +     1 exp erfc     αη η α η  αη  2 2 2    2     2    + + − + α 2 ( ) η = + + − 2 2 2 2 x y z H − γ e 2 α 2 ( ) η = + + + 2 2 2 2 x y z H + γ e 2 ( ) 1 ∞ ∫ = − 2 t erfc W e dt π W where α and γ is given with Pasquill stability classification. In this case, H e is given by Briggs equation. − 3 θ   8 d 1 ∆ =   H 1 . 4 Q 4 H   dz d θ dz is potential temperatu re gradient (K/m)

Euler Model(AS & LRT) Equation ∂ ∂ ∂ ∂ ∂ ∂ ∂  ∂  ∂ ∂     c ( uc ) ( vc ) ( wc ) c c c = − − − + +   + + +     i i i i i i i k k k R Q   ∂ ∂ ∂ ∂ ∂ ∂ ∂ ∂ ∂ ∂ h h v i i       t x y z x x y y z z c Concentration ⋅ i C N v k k Diffusion parameters N ∆ x h v N ∆ Reaction terms R Q y ･ ･ ･ ⋅ ⋅ E ⋅ i i C C C W u P E ∆ u W y E W v S ∆ x S ･ ⋅ C S Input and output through a cell

Deposition of Sulfur { ( ) } = − ∆ − ⋅ ⋅ ∆ ⋅ ∆ ⋅ ∆ − ⋅ ∆ ⋅ ∆ ⋅ ∆ R W z PR C x y z kC x y z P , SO P , SO P , SO P , SO P , SO 2 2 2 2 2 { ( ) } = ⋅ ⋅ ∆ ⋅ ∆ ⋅ ∆ + − ∆ − ⋅ ⋅ ∆ ⋅ ∆ ⋅ ∆ R k C x y z W z PR C x y z P , SO P , SO P , SO P , SO P , SO 2 2 4 4 4 ∆ C ⋅ ∆ ⋅ ∆ ⋅ ∆ = − + + P , SO 2 x y z E E Q R ∆ in , SO out , SO P , SO SO Emission t 2 2 2 2 ∆ C ⋅ ∆ ⋅ ∆ ⋅ ∆ = − + + P , SO 4 x y z E E Q R ∆ , , , in SO out SO P SO SO t 4 4 4 4 Oxidation 2- SO 4 SO 2 k: reaction rate of SO2 → SO4 2 – Dry Wet W: dry deposition ratio deposition deposition PR: wet deposition rate

AIM/Air Program Package • Programs in C language: – 20 programs including utility programs – 6410 lines including comment lines • Graphics: X-window. – figcont: concentration – figarrow: wind direction and strength • Convenient batch programs: – aircalc: menu of calculation – airview: menu of visualization

Implementation Principles • Platform : Linux and MS Windows • Emission projection and inventory : Based on AIM/Database. With MS Access. • Weather database : ECMWF data 0.5 o grid, every 6 hours + local weather station information • Air Models : C program worked on Linux and MS Windows. Complete programs and communicate with other modules by files. • Complicate calculation with Linux, transport the aggregated output to MS Windows in order to use by end users.

Conclusion • AIM/Air software package was developed. • AIM/Air includes LPS, AS, LRT models. Multi-scale diffusion simulation is supported. • This tool becomes a powerful tool for studies on local air pollution of each country and global air pollution in the Asia.