I am using the EDGAR inventory to generate CMAQ-ready emission files with the SMOKE system. However, the runtime is extremely long. To address this, I am considering reducing the file size by cropping the global data (e.g., from (1800,3600) to (350,550)) to match my model domain. Is this possible?
I would like to apply different hourly profiles for weekdays and weekends. My current example amptref file includes only ALLDAY profiles. Is there a way to assign unique hourly profiles for WEEKDAY, SATURDAY, and SUNDAY?
I have not tried to extract the subdomain to run SMOKE faster but I guess it is feasible as long as you extract both EDGAR inventory as well as GRIDMASK input files. Just aware that there could be a high chance of issue or error occurred.
Thank you for your response ! I would like to ask if it is possible to configure different hourly profiles by day of the week when running SMOKE with the EDGAR inventory?
We haven’t tried windowing the EDGAR files but don’t see any reason why that couldn’t be done, although we are not sure how much of a difference that would make with runtime. We are a little surprised you are having runtime issues though. Is this happening with the sectors that have 3-D layering (Laypoint), e.g. energy and industry, or also the 2-D sectors, e.g. transport? How long are the runtimes?
For temporal profiles, yes, you can define different temporal profiles by day. For example, you could put in the TREF:
Thank you for your response !
As you may already know, EDGAR encompasses 27 SECTOR categories. These 27 SECTORs can be consolidated into 15 SECTORs, namely [‘TNR’, ‘AWB’, ‘AGR’, ‘BMB’, ‘TRO’, ‘RES’, ‘IND’, ‘NEU’, ‘PRO’, ‘SOL’, ‘TRF’, ‘ENE’, ‘LTO’, ‘CDS’, ‘CRS’, ‘WASTE’]. When executing the process under this consolidation, it typically takes approximately 7–8 minutes to run each SECTOR per day. Consequently, generating a CMAQ-ready emission file for all SECTORs in a single day requires around 1 to 1.5 hours to complete.
Your runtime is not much different than ours, but we can run our sectors in parallel streams. You could consider submitting multiple jobs at once if your computer system supports this. Also, you can consider using representative dates, e.g. you don’t need to run the transport sector for every day, just for 7 days for each month.