WRAP Modeling Forum Meeting, March 8-9, 2005, San Francisco, CA Overview of RMC Activities in 2004 Gail Tonnesen, Chao-Jung Chien, Zion Wang, Mohammad Omary University of California, Riverside Zac Adelman, Andy Holland University of North Carolina Ralph Morris, Sue Kemball-Cook, Yiqin Jia, Bonyoung Koo, Steven Lau, Abby Hoats ENVIRON Corporation Int., Novato, CA WRAP Modeling Forum Meeting, San Francisco, CA March 8-9 18, 2005 WRAP Modeling Forum Meeting, March 8-9, 2005, San Francisco, CA MM5 Modeling • Completed 2002 36-km modeling: – Poor initial performance in the WRAP region. – Added a sensitivity study for 36 & 12-km MM5 • Completed 36 and 12-km MM5 modeling using configuration identified in sensitivity testing. • MM5 results are now available for distribution on both the 36 and 12-km WRAP domains. • Deliverables include updated modeling protocol, reports on sensitivity testing, peer-review comments and MM5 output. WRAP Modeling Forum Meeting, March 8-9, 2005, San Francisco, CA Emissions Modeling • Developed several iterations of the Preliminary 2002 emissions inventory, Pre02 A – F, 36km: – Pre02a: initial test version – Pre02b: – Pre02: included fire emissions – Pre02e: included only natural fire emissions. • Pre02d 12-km emissions completed in March 2005. • Deliverables include reports, ppt files and QA results on the RMC webpage. WRAP Modeling Forum Meeting, March 8-9, 2005, San Francisco, CA CMAQ Modeling • Ran CMAQ and completed model evaluation for each emissions version. • Ran comparisons with new CMAQ releases: – Versions 4.2.2; 4.4beta and 4.4final • CMAQ 12-km Pre02d is still in progress: – Initial version had bad NH3 soil EI, – currently have results for Jan/July with no soil NH3 EI. – Corrected Pre02d 12-km case to be completed end of March. • Deliverables include reports and ppt files on the RMC webpage. WRAP Modeling Forum Meeting, March 8-9, 2005, San Francisco, CA Source Apportionment • Early in 2004 decided to port the tagged species source apportionment algorithm to CMAQ v4.4beta. • Major problems in the algorithm with mass conservation: – Attempted several different schemes to renormalize to account for lost mass. – Error terms accumulated and were substantial. • Completed an annual simulation with source apportionment for SO4 and NO3 – Error term accumulated in “other” category. – Results are qualitative, not quantitative WRAP Modeling Forum Meeting, March 8-9, 2005, San Francisco, CA Natural Background • Reviewed availability of natural emissions data – list of references included in report. • Still missing major categories of natural emissions – uncetain if current EI would be a substantial improvement ove our previous 1996 natural modeling. • Develop natural EI in 2005/6 work plan. WRAP Modeling Forum Meeting, March 8-9, 2005, San Francisco, CA Compare Alternate Models • Selected CAMx as alternate model. • Decided to include PSAT, the CAMx source apportionment algorithm which does not suffer from mass conservation problems. • Model results delayed due to wait for latest 2002 MM5 and pre02d emissions results and bug in emissions converter – results available for this meeting. WRAP Modeling Forum Meeting, March 8-9, 2005, San Francisco, CA Model Evaluation Software • Many updates made to the evaluation software: – Recoded to improve efficiency and clarity of code – New evaluation plots and metrics implemented – Much uncertainty over site locations (“no right answer”) so we now use expanded site list. • Experimented with new methods of presenting model evaluation – stacked bar plots and bugle plots. • Plan to release “Open Source” version of the model evaluation package. WRAP Modeling Forum Meeting, March 8-9, 2005, San Francisco, CA Fire Sensitivity Modeling • Performed fire sensitivity studies, cases included: – vertical distribution of ag buring – Natural and anthropogenic fires – Analysis of previous base and optimal smoke cases. • Difficult to obtain data sets for all planned sensitivities studies: – Deferred studies of small fires and point source vertical distribution sensitivity. WRAP Modeling Forum Meeting, March 8-9, 2005, San Francisco, CA Alaska Modeling • Completed MM5 modeling for Alaska – Uses polar stereographic projection – Use of sea ice limits in winter PBL/LSM options • Preparing Emissions and data sets needed for CALPUFF modeling – Assisting Alaska emissions contractor WRAP Modeling Forum Meeting, March 8-9, 2005, San Francisco, CA Training/Tech Transfer • Developed introductory training material for managers • Provided some support to state agencies in systems configuration – Decided to support only a single, standardized modeling platform. Preparing FAQ to assist in systems and model set up • Data sets are being transferred to WRAP members. WRAP Modeling Forum Meeting, March 8-9, 2005, San Francisco, CA 2004 Problems and Solutions • Significant staff turnover at UCR during 2004 – New programmer on board – still need project management help • Included funding for Tech Writer at UNC to assist in reporting. • Major problems with computer hardware and disk storage needs for 12-km modeling. – Need to redesign SMOKE 3-D emissions data structures. – Need expanded and more reliable disk storage. WRAP Modeling Forum Meeting, March 8-9, 2005, San Francisco, CA 2004 Problems and Solutions (cont.) • Mass conservation errors in CMAQ 4.4beta: – EPA funding to revise source apportionment algorithm – CAMx available as alternate model for source apportionment modeling. • Significant improvements in evaluation tools for presenting modeling results: – Still lack formal guidance on “acceptable” model performance.