Applications of Satellite Data for Ozone Exceptional Events

Transcription

1 Applicatins f Satellite Data fr Ozne Exceptinal Events Gail Tnnesen EPA Regin 8 Western U.S. TEMPO Early Adpters Wrkshp April 11, 2018, Frt Cllins, CO This presentatin des nt reflect EPA plicy, guidance, r recmmendatins

2 Applicatins f Satellite Data Develp imprved air quality mdels: Imprved evaluatin f meterlgical and phtchemical mdels. Develp tp-dwn cnstraints n emissins inventries. Assimilatin f satellite data int meterlgical and air quality mdels: fr precursr cncentratins; O3 in the stratsphere and free trpsphere. Infrm air quality management decisins: Identificatin f exceptinal events (stratspheric intrusins, high wind events, wildfires) that cntribute t O3 and PM. Use f indicatr ratis t identify the limiting precursr fr prductin f O3 and PM2.5. Health Impacts studies: Current epidemilgical studies fr O3, xidants and PM are based n sparse mnitring data. High spatial reslutin TEMPO data shuld be useful fr new health effects assessments f O3 and ttal xidants: 2

3 Ozne is used as an indicatr f xidant air pllutin High zne day in suthern Califrnia High O3 days assciated with Stratspheric intrusin can have very lw cncentratins f ther xidants. Can TEMPO prvide better, multipllutant indicatrs f xidant air pllutin? High zne day in Gthic, Clrad 3

4 EPA, State and Tribe Planning Needs States, Tribes and EPA use phtchemical mdels t develp Implementatin Plans (SIPs, TIPs, r FIPs) that demnstrate: Attainment f Natinal Ambient Air Quality Standards fr O3 and PM2.5 Gd Neighbr SIPs fr interstate transprt f O3. Prgress tward reginal haze gal f natural visibility cnditins at Class I Areas. Air Quality gals are evaluated fr a subset f days: O3 attainment is typically evaluated fr the ten highest mdeled O3 days. Reginal haze gal is evaluated fr the 20% wrst anthrpgenic impairment days and fr the 20% best visibility days. TEMPO high tempral reslutin will be valuable fr episdic and day specific analyses. 4

5 Analysis f Specific Events r Episdes Mdels shuld accurately simulate backgrund O3, PM2.5 and natural haze levels n individual days: Backgrund O3, PM2.5 and natural haze levels are highly variable in space and time. D mdels accurately represent backgrund O3, PM2.5 and natural haze n the subset f days used fr air quality planning? Can the next generatin f satellite data help prvide mre accurate estimates f daily O3, PM2.5 and natural haze? Days with high O3 r PM2.5 caused by Exceptinal Events can be excluded frm the planning prcess. Fr O3 and PM2.5, states prepare an exceptinal event demnstratin fr review and apprval by EPA. Fr reginal haze, data analysis methds are used t estimate the natural haze level n each day. Hpe that TEMPO data will make it easier t identify exceptinal events. 5

6 Stratspheric Intrusin Wrkgrup Wrkgrup meeting semi-mnthly since 2012, include federal and state research and air quality management cmmunities. Lin et al. (2015) shw mre frequent late spring stratspheric intrusins when the plar jet meanders twards the western United States, such as ccurs fllwing strng La Niña winters. Few stratspheric O 3 events in 2015 and 2016 (during El Niñ cnditins) but winter 2016/17 had La Niña cnditins, and 2017 had several pssible intrusin events with maximum daily 8-hur (MD8) averages appraching r exceeding the NAAQS: March 18: 95 ppb (87 ppb regulatry) at Mt Washingtn NH. April 9: 74 ppb in Clrad Springs. April 22: 82 ppb Gthic, CO; 80 ppb Navaj Lake, NM; 75 ppb Centennial, WY; and five thers sites in UT, CO and NM exceed 70 ppb. April 23: 71 ppb NREL, CO; 71 ppb Rcky Flats, CO Several additinal days with MD8 between 65 t 69 ppb. 6

7 Analysis f SI O3 Exceptinal Events Clear Causal Relatinship between the Specific Event and the Mnitred Cncentratin Analyses shwing stratspheric-trpspheric exchange Analyses shwing stratspheric air reached the surface Air quality analyses shwing the impacts f the intrusin at the surface Gal is t use TEMPO fr these analyses. 7

8 Resurces fr SI O3 Analyses NOAA RAQMS simulatins and IDEA Trajectry analysis tl (Brad Pierce, NOAA). NCEP Glbal Frecast System mdel (GFS) analysis and frecast plts are generated each day at 12 hur intervals ut t 5 days. Trpspheric Ozne Lidar Netwrk (TOLNet) grund based lidars (Mike Newchurch, UAB; Andy Langfrd et al., NOAA/ESRL) Atmspheric Sunding, website hsted by Univ f Wyming AirNw surface O3 mnitring by states, tribes and EPA Additinal surface O3 sites by NOAA/GMD, NPS, FS and thers. NOAA Rapid Refresh (RAP) mdel simulatins: Satellite based measurements f O3, CO, NO2, HCHO, H2O and aersl ptical depth. NASA AJAX aircraft bservatins. Staff at state envirnmental agencies with expertise in meterlgical analysis and lcal cnditins. 8

9 Satellite Data Example Plts Water vapr imagery frm the GOES-West satellite 9

10 Satellite Data Example Plts Map f OMI satellite-estimated ttal zne clumn data frm the day befre the event 10

11 TEMPO data prducts Satellite based measurements f O3, NO2, HCHO, H2O. Expect very lw mixing ratis f H2O, CO, NO2 and HCHO t be assciated with stratspheric intrusins. Mixing f wildfire r Asian transprt plumes with stratspheric intrusins can make it mre difficult t interpret. High winds and wind blwn dust assciated SI events might r might nt cause high with particulate matter and aersl ptical depth. Will hurly TEMPO day aid in visualizing transprted plumes? Supplemented by trajectry analyses as Brad shwed? 11

12 Wildfire Exceptinal Event Cmplexity Denver O 3, June 23, 2015: 91 ppb (hurly 122 ppb). Smke frm Califrnia Fires nrth and Arizna fires suth f Denver. Urban 8-hur O 3 at 91 ppb was 35 ppb higher than the highest rural site in Clrad, at 56 ppb. Smke impact, if any, is ambiguus. Wildfire Plumes, June 23, 2015 Smke Impacts n PM 25? 7

13 Nested 36/12/4-km Mdel Dmains WRF and CMAQ r CAMx Vertical Grid: Typically 20 t 40 layers frm the surface up t the lwer stratsphere at 50 mb. Lateral BC frm Glbal Mdels: MOZART GEOS-Chem AM3 RAQMS 13

14 Ozne Attainment Planning Mdel evaluatin is cmpleted fr histrical O3 episdes t assess if the mdel is reliable fr prjecting future changes in O3. The base year O3 design values is calculated as the weighted average f 5 years f the 4 th highest daily 8-hur average O3. Mdel O3 relative respnse factr (RRF) is used t prject the future O3 design value. RRF = Future mdel O3/Base year mdel O3 (average rati fr the ten highest mdeled O3 days. Future O3 design value = (RRF)(Base Year O3 Design value) The Relative Respnse Factr apprach is adpted t crrect fr mdel bias. The state successfully prjects attainment f the O3 NAAQS if the future design value des nt exceed the NAAQS. 14

15 CMAQ biased lw at rural sites at 7 am, May 30, The mdel simulates a ridge f elevated zne but is biased lw by 30 ppb 15 11

16 1-hur O3 mdel and bs at 1 pm LDT July 23, 2011 Mdel perfrms well in the Denver area but is biased high fr reginal zne n July

17 17 July 23 Mdel respnse ( ) fr 1-hur average O3 Hurly O3 NOx disbenefits in mrning and benefits in afternn. 8-hur average has small benefit. Delta O3 at 11 am LDT Delta O3 at 3 pm LDT

18 HCHO/NO2 Indicatr Schreder et al., New Insights int the Clumn CH2O/NO2 Rati as an Indicatr f Near-surface Ozne Sensitivity: Clumn measurements f HCHO/NO2 are a pr indicatr f nearsurface zne sensitivity due t uneven vertical mixing in the lwer trpsphere. Lng and shrt-term bservatins f changes in clumn HCHO/NO2 ratis may still prvide useful infrmatin abut the relative imprtance f O3 precursrs Jin et al., 2017, Evaluating a space-based indicatr f surface zne-nox- VOC sensitivity ver mid-latitude surce regins and applicatin t decadal trends: trends in HCHO/NO2 ver the past decade indicate that a NOx emissin cntrl prgram will imprve zne air quality mre nw than it wuld have a decade ag. 18

19 CAMx mdeled PO3 Sensitivity Hpe t use TEMPO data t evaluate if the mdel accurately simulates PO3 sensitivity t VOC and NOx 19

20 CAMx mdeled vs GOME2 HCHO/NO2 GOME2 HCHO/NO2 evaluated by Pat Reddy at CDPHE, used 2 mnth average t imprve S/N rati. Nt quantitatively cmparable t mdel hurly rati but GOME2 indicates greater relative sensitivity t VOC in Denver Urban area, cnsistent with mdel PO3 sensitivity. 20

21 Summary State, Tribe and Federal Air Quality planners need: High spatial reslutin hurly measurements f O3 and precursrs fr urban and reginal O3 planning. Reginal scale measurements f O3 and precursrs fr evaluatin f O3 exceptinal events and backgrund O3 levels. Measurements f speciated PM t evaluate natural haze levels and internatinal transprt f visibility impairing pllutants. 21