mmb-carb / esta Goto Github PK
View Code? Open in Web Editor NEWEmissions Spatial and Temporal Allocator model
License: GNU General Public License v3.0
Emissions Spatial and Temporal Allocator model
License: GNU General Public License v3.0
Is there anything to do here? Or will our handling of EMFAC2014 files be fine?
Just adding a single loop could make it a lot easier to run a series of ESTA jobs in series.
They would be totally separate ESTA runs. If users want parallel ESTA runs, there will be system-dependent load-balancing issues and they can handle that themselves.
It would be nice to add user functionality for the output file name to add the speciation profile used when creating netCDF output?.
For example, the current default name is:
st_4k.mv.v0220..2013.2013365..e14..ncf.gz
If we used 2013 speciation gsref with vintage 11jan17, it would be nice to be able to keep track of this in the file name to have :
st_4k.mv.v0220..2013.2013365..e14.gsref_13_11jan17.ncf.gz
The eic_info.py file has a single dictionary with EICs as keys, and three columns of data as values. The question is: do we still need that first column? e.g.
71073411000000: (0, 'vmt', 1.0), # ('LDA', 'CAT', 'RUNEX')
That first column 0
is a relic of the DTIM link files. However, now it is tied into the CALVAD_VEHICLE types in some strange way. But that is all certainly unnecessary and complicated. So it needs to be removed.
I think the class name CalvadSmoke4SpatialSurrogateLoader
is no longer accurate. This class should probably be renamed to just SmokeSpatialSurrogateLoader
.
Users may want DPM in their NetCDF files for toxics studies.
The PMEDS output format (based on the MEDS format) has proven too restrictive. In particular, it does not allow for the California 1km modeling domain. As such, it is time to remove support for the PMEDS format, in favour of a headerless CSV-like format.
We are using the CalTrans trucking network as our linehaul spatial surrogate. Are there future years available for this network? If so, we should grab those to make future-year versions of that surrogate.
It would be worth researching.
Right now there is a hard-coded mapping from month to season, but users are asking for more flexibility so that seasons can also depend on the region code.
To make this, more complete, mapping it will have to be moved to an input file.
Patch for CALVAD spatial surrogates pending the following updates:
Would it be nicer if we had full examples of ESTA run on different grids? Perhaps a SCAQMD and BAAQMD example, with their own surrogates?
There are some lingering examples of run configs and code that was original used to test ESTA against DTIM for on-road. These tests are quite old, unsupported, and only partially functioning. It is time to remove them and make the simple examples for new users easier to follow.
What happens if we delete an arbitrary line from the eic_info.py
file? Can the program keep running? (Probably not.) So, does ESTA fail with any helpful error message?
Do we want the model to use magical defaults if an EIC is missing? Or do we want it to drop the emissions related to unknown EICs? Should warning messages get thrown to the screen, or should the model fail?
What would be the best possible behaviour?
We currently use Calvad for ITN/VMT spatial surrogates. We should try ITN, to support future years and different grids.
Upgrade the SMOKE spatial surrogate version to the most recent SMOKE 4.x. The major difference here is the second column should have something more like "06003" as opposed to something like "0SC0060300SC". This will simplify the input information needed for the ESTA model significantly.
I believe ESTA should already be able to process toxics. You just need to provide it different GSPRO/GSREF files. BUT, is that the case?
We should try it and find out!
Why do we have this file? can we replace it or create it on the fly?
Update the documentation for the two new input CSVs (NOx-by-GAI and season-by-GAI).
Currently, the Heavy-Duty diesel NOx fractions are given by year, but the defaults aren't. It would be better if the defaults also varied by year.
A declarative, efficient, and flexible JavaScript library for building user interfaces.
π Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. πππ
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google β€οΈ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.