衛(wèi)星遙感技術在空氣質量模型中應用院

1、衛(wèi)星遙感技術在空氣質量模擬與預測系統(tǒng)中的應用 唐偉、孟凡、徐峻、何友江、劉世杰 中國環(huán)境科學研究院環(huán)境空氣質量遙感監(jiān)測技術研討會2014.12.8 北京Particulate Matter can be characterized on basis of particle size (aerodynamic diameter): TSP (Total Suspended Particle): 0-40 m PM10: 0-10 m Coarse Particle: 2.5-10 m PM2.5 (fine particle): 0-2.5 m Ultrafine particle: 0.1 m

2、2Particulate Matter (PM) Cause of Concern and Health Effects of Particulate Matter Regional haze Affect overall radiative balance of the earth Influence chemical properties of droplets, surface water acidity Erosion and staining of structures Human health effect Lung cancer Bronchial asthma Eye and

3、skin irritation Cardiac hospital admission34“Good Up High, Bad Nearby” U.S. EPASource: Good and Bad OzoneAbsorb UV-BAir pollutant大氣污染研究野外觀測實驗室試驗 數(shù)值模擬模式的重要作用1)解釋觀測現(xiàn)象2)污染演變規(guī)律 3)預測與調控工具Photochemical Air Quality models Community Multi-scale Air Quality Modeling System (CMAQ) (US. EPA) Comprehensive Air

4、Quality Model with extensions (CAMx) (ENVIRON) Weather Research and Forecasting (WRF)-Chem (NOAA)7Photochemical Air Quality Modeling ,PM2.5 concentrationO3 concentrations and its sensitivities to NOx and VOCsO3 is sensitive to NOx over most of China, except ShanghaiO3 sensitivity to NOx emissionsO3

5、sensitivity to VOC emissionsO3 concentrations 2010-08PM2.5 concentrations and its sensitivities to NOx, VOCs, SO2, and NH3PM2.5 concentrationsPM2.5 sensitivity to NOx emissionsPM2.5 sensitivity to VOCs emissionsPM2.5 sensitivity to SO2 emissionsPM2.5 sensitivity to NH3 emissionsIn 京津冀, reducing VOCs

6、, SO2, and NH3 emissions may bring down PM2.5 level, except NOx emissions 2010-01為公眾服務為管理服務空氣質量預報預警的作用（能干什么）提供空氣質量預報信息提供空氣重污染預警信息發(fā)生空氣重污染時的應急措施發(fā)生空氣重污染時的重點源管理全國范圍省/自治區(qū)/直轄市范圍市/縣范圍市區(qū)范圍長期預報：5-7天中長時預報：2-5天短時預報：1-2天空氣質量預報尺度范圍空氣質量預報時間Meteorology prediction NCEP ( )14Chemical mechanisms.Reaction rateconstant

7、sEmission inventoryPhotolysis ratesUncertaintyStructural Uncertainty(imperfections in the models numerical representations of atmospheric processes)Parametric Uncertainty(model inputs or parameters errors) Vertical transport schemes.Air Quality ModelUncertainties in Air Quality ModelsAssimilated Pho

8、tolysis rates using GOES satellite observationsInverted NOx emissions using OMI satellite and ground observations Modeled PM2.5 and Ozone concentrations Air Quality ManagementOverview15Modeling domainEastern Texas Episodes from Texas SIPsAug 13 Sept 15, 2006 (during TexAQS-2)16Impact of the Geostati

9、onary Operational Environmental Satellite (GOES)-derived photolysis rates on Ozone modelingPart I17Photolysis rates (J) simulated in Model18Clear-sky photolysis rates (Jclear)Predicted by MM5Satellite (GOES)JNO2 19 JNO2NO2O3Differences inNO2O3More photochemical activitiesImpact of GOES-photolysis ra

10、tes on NO2 and O3 simulationsand vice versaInverse modeling of NOx emissions using space-based NO2 observationsPart II20+.Emission factors * Emission activitiesBottom-up NOx emission inventory21（自下而上）ModeledNO2Bottom-upTop-downInverseTop-down NOx emission inventoryMeasuredNO2Forward22（自上而下）Ozone Mon

11、itoring Instrument (OMI) NO2 dataNO2 spectral range (405-465 nm)Overpass time: 13:30 local time dailySpatial resolution: 13 24km2 (nadir)NASA Standard Product version 2 gridded to 0.1 0.1 resolutionSource: OMI Guide 2009; Courtesy of Lok Lamsal23Observed vs. Modeled NO2 column OMI NO2 is higher than

12、 modeled NO2 columns in the rural areas, but lower in the urban areas24Modeled NO2 OMI NO2DifferenceDiscrete Kalman filter (DKF) inversion25(A priori)(A posteriori)NOx emissions: A priori vs. A posteriori26Sector-based inversionRegion-based inversion27Influence of GOES-corrected photolysis rates and OMI-constrained NOx emissions on O3 simulationsModel performance in O3 simulations Ground O3 simulations get improvedFuture worksInverse modeling of aerosol precursor emissions using satellite-observ