Development of an emissions processing system for global and regional scale emission inventories A&WMA International Specialty Conference May 1-14, 21 Xi an, Shaanxi Province, China Jihyun Seo 1), Bujeon Jung 1), Ki-Chul Choi 1), Jung-Hun Woo 1), Young Sunwoo 1) 1) ATF, Konkuk university, Seoul, S. Korea
Contents 1 Background & objective 2 Framework 3 System structure 4 Emission results 5 Computational performance 6 Summary & on-going work
Background Background 1 Anthropogenic contributions to the chemical composition of the atmosphere affect the balance of both visible and infrared radiation in the Earth-atmosphere system(bond et. al., JGR, 24). Anthropogenic emissions are increasing rapidly and will continue to increase in the future. Therefore, the development of emission i inventories i for the past, present, and future is very important(ohara et. al., ACP, 27). FORCING Emissions Climate Atmospheric chemistry Air pollution meteorology Source : Jacob et. al., AE, 29 Source : Ohara et. al., ACP, 27
Background Background 2 But, emission inventory processing has many complicated steps. Steps Emission projection Chemical speciation Temporal allocation Spatial allocation Components Category Numbers Description Sectors > 3 Fuel combustion, fuel production, biofuel combustion, industrial process, landfills etc. Fuels > 4 Coal, heavy liquid fuel, light liquid fuel, gas Pollutants > 1 SO 2, NOx, CO, BC, OC, NMVOC, NH 3,CO 2,N 2 O, CH 4 Scenario >2 A2, B1 Number of cases > 24 To improve efficiency of the process, it is necessary to integrate & automate the process
Objective Objective Integrate and automate emissions processing Estimate a global emission inventory of pollutants and GHGs for the future. Pollutants : NOx, CO, NMVOC, NH 3,SO 2, BC, OC, GHGs (CO 2,N 2 O, CH 4 ) Period : year 21 ~21 (base-year: year 2, for 1 years) Temporal & spatial resolution : 1 X 1 grid, monthly (global) l) Scenario: A2 and B1 scenario of IMAGE model output Anthropogenic emissions
Framework Global emission processing system Base year inventory (global + regional) NIER (29) EDGAR 3.2 FT 2 Bond et al. INTEX 26 TRACE-P EI projection using RIVM IMAGE model (A2, B1) Chemical Speciation for GEOS-Chem Temporal lallocation Using EDGAR, SMOKE, and etc. Gl b l E i i I t Global Emissions Inventory (Monthly, 1 1, and speciated)
Framework Base year inventory Global emission Inventories Inventory EDGAR 3.2 FT2 spatial resolution temporal resolution 1x1 annual 2 period chemicals sector CO 2, CH 4, N 2 O, SF 6, HFC, PFC, CO, NOx, NMVOC, NH 3, SO 2 1 ~ Bond et al. 1x1 annual 26 BC, OC 3 Regional emission Inventories Inventory Inventory Domain Spatial resolution Temporal resolution years Sector / chemical Trace/ACE 2* Latitude : -1 ~ 5 Longitude : 6 ~ 15 1x1 annual 2 Anthropogenic : SO 2, NOx, CO, NMVOC, NH 3, OC, BC CO 2, CH 4 Biomass burning : SO 2, NOx, CO, NMVOC, NH 3, OC, BC CO 2, CH 4 INTEX 26** Latitude : -1 ~ 5 Longitude : 6 ~ 15.5x.5 annual 26 Anthropogenic : SO 2, NOx, CO, NMVOC, OC, BC, PM 1, PM 2.5 (Source : * Streets D. G., et al., 23.; **Zhang, Q., et al., 27)
System structure Structure Emission projection Make projection factor Q.A. Q.A. Compare input & output Summary yearly emission Import IMAGE 2.2 model output Reformat model output Extract projection factor Yearly emission projection Import base year inventory Import projection factor Matching projection factor by region number Estimate yearly emission Q.A. Monthly emission allocation Compare input & output Import monthly allocation factor Import projected emission i Matching monthly allocation factor by region Allocate monthly emission
System structure Flow chart Input Output IMAGE 2.22 model emission data by scenario Base year emission Monthly allocation factor Yearly emission projection factor extraction Assumption - 2 scenarios - 1 pollutants - 3 sources - 1 years scenarios pollutants sectors fuels Yearly emission projection Monthly emission allocation Air quality modeling input (GEOS-Chem) Final output by by by scenarios pollutants sectors scenarios pollutants sectors fuels
System structure Yearly emission projection factor extraction year region source fuel emission year region1 region2 region3 region4 Main logic IMAGE 2.2 model emission output data Reformat emission data region f_197 f_1971 f_1972 f_1973 f_1974 f_1975 Yearly emission projection factor
System structure Yearly emission projection Main logic Yearly emission projection factor coal, heavy liquid, light liquid, gas lon lat region grid id province id coal heavy liquid light liquid gas Base year inventory Projected emission
System structure Monthly emission allocation lon lat country id region grid id province id coal heavy liquid light liquid gas Projected emission inventory country province id sector Jan. Feb. Mar. Apr. May Jun Jul. Aug. Sep. Oct. Nov. Dec. Monthly allocation factors Main logic * Global Industry, power generation, fuel production Latitude x>1, x<-1, else Energy transformation in a lump Residential province id Residential Latitude x>6, x<-6, 3<x 6, -6 x<-3, x 3, -3 x< Transportation in a lump * Asia Industry, power generation, energy transformation, transformation, fuel production county id
Global temporal allocation factor 2.5 2 Power plant (fuel use) * 15 1.5 1 Area source combustion (industrial combustion) * Small combustion sources(residential : fuel use) # * Refineries Industrial processes.5 solvent use Transport (traffic) Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Agriculture * for areas around the equator use 1/1/1/1/1/1/1/1/1/1/1/1 for Southern Hemisphere shift all factors 6 months # considering climatic zone (temperate, tropical, polar zone)
Emission results Yearly emission trend SO 2 NOx CO NH 3 CH 4 A2 Sc cenario Tg SO2 3 25 2 15 1 5 Tg NOx 35 14 3 12 Fuel production Fuel production 25 1 Waste incineration Waste incineration 2 Industry process Industry 8 process Transportation 15 Transportation 6 Residential 1 Residential 4 Power generation 5 Power 2 generation Industry Industry 2 21 22 23 24 25 26 27 28 29 21 2 21 22 23 24 25 26 27 28 29 21 2 21 22 23 24 25 26 27 28 29 21 Tg CO Tg NH3 18 16 14 Fuel production 12 Waste incineration 1 Industry process 8 Transportation 6 Residential 4 Power generation 2 Industry 2 21 22 23 24 25 26 27 28 29 21 Tg CH4 18 16 14 12 1 8 6 4 2 2 21 22 23 24 25 26 27 28 29 21 B1 Scenario Tg SO2 3 25 2 15 1 5 2 21 22 23 24 25 26 27 28 29 21 Tg NOx 35 3 Fuel production 25 Waste incineration 2 Industry process Transportation 15 Residential 1 Power 5 generation Industry 1 2 3 4 5 6 7 8 9 2 21 22 23 24 25 26 27 28 29 21 14 12 Fuel production 1 Waste incineration 8 Industry process Transportation 6 Tg CO Residential 4 Power 2 generation Industry 2 21 22 23 24 25 26 27 28 29 21 18 16 Fuel 14 production Waste 12 incineration 1 Industry process 8 Transportation 6 Residential 4 Power generation 2 Industry Tg NH3 2 21 22 23 24 25 26 27 28 29 21 Tg CH4 18 16 14 12 1 8 6 4 2 2 21 22 23 24 25 26 27 28 29 21 * growth rate=(21-2)/2*1 SO 2 NOx CO NH 3 CH 4 A2 growth rate(%) 59 26 96 2 47 B1 growth rate(%) -84-56 -79 56 35 Main contributor sector Power generation Power generation /transportation Residential Livestock Landfills
Emission results Monthly emission trend Red mark: A2 Blue mark: B1
Emission results Emission in 21 by region 1% 9% 8% 7% 6% 5% 4% 3% 2% 1% % Base-year SO2 NOx CO2 CO CH4 BC OC NH3 Northern America Southern America Africa Europe Former USSR Middle East Other Asia East Asia Oceania Other Main contributor regions in 2 NOx, CO 2 Northern America 21 % SO 2, CO, BC, NH 3 East Asia 25 % CH 4, OC Other Asia 26.5 % A2 SO 2, NOx East Asia 23 % CH 4 Other Asia 35 % B1 NOx Africa 21 % SO 2, CH 4 Other Asia 28 % 1% 9% 7% 6% 5% 4% 3% SO 2 NOx CH 4 1% Other 9% 8% 8% Oceania 8% Oceania Oceania 2% 1% % A2_21 2 B1_21 7% East Asia 6% Other Asia 5% Middle East 4% Former USSR 3% Europe 2% Africa 1% Southern America % Northern America A2_21 2 B1_21 1% Other 9% 7% East Asia 6% Other Asia 5% Middle East 4% Former USSR 3% Europe 2% Africa 1% Southern America % Northern America A2_21 2 B1_21 Other East Asia Other Asia Middle East Former USSR Europe Africa Southern America Northern America
Computational performance Computational performance Programs Processing Processing Total process speed error hours of operation volume of data number of data Yearly emission projection factor 2.8 MB / second % 4 second 112 MB 2,4 extraction Yearly emission 16 MB / second - Projection Monthly emission 14 MB / second <. % allocation 1,8 second (18 minutes) 18,224 MB (176 GB) 28,8 second 4,194,34 MB (78 hours) (4 TB) 6, 24, CPU: INTEL XEON E543 Harpertown 2.66GHz 2CPU (8 cores) Specification of RAM: 16GB DDR2 computer OS: Centos 4.7 Compiler: Portland Group Fortran compiler
Summary & on-going work Summary We developed an emission inventory yprocessing gprogram to integrate & automate the process. Emissions were estimated using a developed emission inventory processing program for A2 and db1 scenarios. In case of A2 scenario, all pollutants emissions increase 59%~47% On the other hand, emission estimates of B1 scenario shows relatively less increase or decrease, from -84%~56% East Asia and other regions of Asia show up as main contributor regions in the base year and in 21 for the A2 and B1 scenarios.
Summary & on-going work On-going work Emission estimation Scenarios: 6 scenarios(a1b, A1FI, A1T, A2, B1, B2), Period: 4 years(year 1996~25, 216~225, 246~255, 291~21) Scale: global & regional emission inventory New scenario study RCP (Representative Concentration Pathway) : inquire the scenario & compare and analysis with earlier study Source: IPCC, CLIMATE CHANGE 21:THE SCIENTIFIC BASIS
Thank you for your attention! ACKNOWLEDGEMENT This research was performed under the support of "National Comprehensive Measures against Climate Change" Program by Ministry of Environment, Korea (Grant No. 17-1737-322-21-13)