____________________The State of the SMOKE Modeling System 2001 2002

____________________The State of the SMOKE Modeling System 2001/2002 Marc Houyoux MCNC Environmental Modeling Center October 21, 2002 http: //www. emc. mcnc. org/products/smoke 1 ___________________________Environmental Modeling Center

____________________The State of the SMOKE Modeling System 2001/2002 Overview l l l l Provide a snapshot of SMOKE Status to date Emissions modeling trends Recent SMOKE enhancements Ongoing SMOKE development Recent SMOKE applications Release and support issues 2 ___________________________Environmental Modeling Center

____________________The State of the SMOKE Modeling System 2001/2002 Status to Date l l l l l Idea for SMOKE from Dr. Carlie Coats Vector-matrix multiplication, I/O API, integer-based matching Prototype based on limited EPS 2. 0 Compared to EPS 2. 0 results for NC Added EMS-95 format readers and processing Compared to EMS-95 results during OTAG Update and integrate with Models-3/MIMS Update based on applications (ongoing) Update for toxic emission inventories (ongoing) Stand-alone versions 1. 3 and 1. 4 -beta currently in use 3 ___________________________Environmental Modeling Center

____________________The State of the SMOKE Modeling System 2001/2002 Current Trends in Emissions Modeling l l EPA – annual and national cases More regional modeling centers starting work – l l WRAP: annual and regional cases Processing for both ozone and haze MOBILE 6 – even precomputed emissions have been favored over episode-dependent MOBILE 5 b Increased interest in SAPRC chemical mechanism PC-Linux platforms 4 ___________________________Environmental Modeling Center

____________________The State of the SMOKE Modeling System 2001/2002 Improvements in version 1. 4 beta l Inventory Import (Smkinven): – – – l (1) PMC computation allowed when VMT in inventory Added CEM data reader and automated matching with IDA inventories Repaired EMS-95 hour-specific usage when trying to match with IDA inventories Updated to permit computation of “annual” emissions when only ozone-season emissions provided Corrected assignment of weekday normalization Controls (Cntlmat): – – – Updated to permit cross-referencing at all 4 levels of SCCs Repaired application of pollutant-specific controls Repaired generating multiple projection/control matrices 5 ___________________________Environmental Modeling Center

____________________The State of the SMOKE Modeling System 2001/2002 Improvements in version 1. 4 beta l Inventory growth (Grwinven): – – l Added support of user-defined selection by stack parameters, emissions rank, plant ID, source ID, analytical plume rise, and any combination of these items Spatial allocation: – l Correctly output IDA-formatted inventories Updated to apply any number of projection or control matrices Elevated source selection (Elevpoint): – l (2) Updated for grid extractions (Grdmat, Rawbio, Laypoint, Temporal, and Premobl) Temporal allocation (Temporal): – – For mobile, allowed use of min/max temperatures from previous day Corrected bug caused when output time zone > source time zone 6 ___________________________Environmental Modeling Center

____________________The State of the SMOKE Modeling System 2001/2002 Improvements in version 1. 4 beta l l Biogenics – released BEIS 3! Layer Fractions: – l (3) Added usage of hour-specific plume rise computed outside of SMOKE (Smkinven, Laypoint, and Smkmerge) Quality Assurance: – – – Updated to include reporting by speciation or temporal profile codes or gridding surrogate codes Added selection of inventory records based on noncontiguous subgrid, region codes, and elevated status Added normalization by population and cell area 7 ___________________________Environmental Modeling Center

____________________The State of the SMOKE Modeling System 2001/2002 SMOKE since version 1. 4 -beta l Main reporting utility (Smkreport) – – l l l Now includes growth and control SCC descriptions included with area and mobile “by source” Merge program (Smkmerge) applies mobile control matrices 3 -d sea-salt model added Corrections – – – Corrected calculation of mobile control matrices (Cntlmat) Corrected reporting by layer fractions (Laypoint) I/O API fix corrected Jan 1 and Oct emissions (Temporal) PC temporal allocation will now run (Temporal) ASCII elevated files fixed to not cause abort or drop emissions in specific cases (Smkreport) 8 ___________________________Environmental Modeling Center

____________________The State of the SMOKE Modeling System 2001/2002 Known problems with 1. 4 beta l l l l Linux IDA-format readers broken Missing CEM data not handled properly Links with zero-length cause intended abort Possibly error in temporal cross-reference files Tmpbio crashes on Sun platforms unless all options files used State/county “controlled” totals from Smkmerge have zero emissions for any pollutant, state, or county that is not controlled Smkreport cannot report speciation codes for hourly emissions when MOBILE 5 b has been used to compute mobile emissions 9 ___________________________Environmental Modeling Center

____________________The State of the SMOKE Modeling System 2001/2002 Ongoing SMOKE development l l l EPA ORD developed new plume rise EPA OAQPS having MCNC update SMOKE for toxics processing WRAP RMC having MCNC make SMOKE more robust for Linux. – l l Seeking about $12 K to release version 1. 4 for Linux EPA STAR grant program has resulted in version of SMOKE that includes inventory uncertainties ACC having MCNC and UC-Riverside update SMOKE for support of dynamic SAPRC mechanism 10 ___________________________Environmental Modeling Center

____________________The State of the SMOKE Modeling System 2001/2002 l l l l l Support real-time MAQSIP ozone forecasting in eastern U. S. Apply new sea-salt model in Gulf Coast for input to CMAQ Develop CMAQ emission inputs for BRAVO Develop 36 -, 12 -, 4 -km inputs in eastern US for CMAQ Annual CMAQ inputs for Clear Skies Initiative Annual CMAQ inputs for western U. S. for WRAP Support NHx cycling research project over eastern U. S. U. Tennessee applying SMOKE in China and Taiwan Canadian effort in northwestern U. S. and W. Canada for CMAQ Northwest US demonstration of SMOKE/MM 5/CMAQ Illinois water survey: 90 -km U. S. and 10 -km midwest for CMAQ NOT MCNC l MCNC Recent SMOKE applications 11 ___________________________Environmental Modeling Center

____________________The State of the SMOKE Modeling System 2001/2002 Release and Support Issues l Two primary challenges for SMOKE users: – – l l Erratic timing of releases and bug fixes Overall quality and testing of release content To date, MCNC has supported most of the cost of releases and testing, which has resulted in lessthan-optimal time spent on these efforts Seeking community funding for support of: – – – Providing semi-annual, high-quality SMOKE releases Creating and using appropriate test cases Correcting bugs and release the fixes in a timely manner Maintaining the SMOKE website Annual cost estimate for this is $35 K 12 ___________________________Environmental Modeling Center

____________________The State of the SMOKE Modeling System 2001/2002 Conclusions l l l Lots of development and applications are ongoing with SMOKE! The community is adopting SMOKE and using it A community solution for funding releases and maintaining the system is sorely needed Last notes: l User-support is also needed, and the approach for this is being worked out through the CMAS l Thanks for your interest in and support of SMOKE 13 ___________________________Environmental Modeling Center