Solar Radiation Data

[s2t2]

We have an example solar radiation TMY3 data file for Greensboro, NC in the codebase. it is used for testing.

Data source:

pvlib-python repository

Data for Mountain View

To work with our historical simulator building SB-1, which exists in the Bay Area (Mountain View, CA), it would be helpful to have historical data from the Moffett Field location. We could add that data to configs/resources.

Alternative Data source:

https://energyplus.net/weather

https://energyplus.net/weather-location/north_and_central_america_wmo_region_4/USA/CA/USA_CA_Mountain.View-Moffett.Field.NAS.745090_TMY3

Pipeline

While we are on the topic, it would be helpful to add an automated script to download historical solar radiation files for any given weather location provided by the data source.

Format

The format of the CSV files we add should be able to be read programmatically by pvlib.read_tmy3,

[ryux1]

I’d be happy to work on this. I checked the current EnergyPlus catalog and the linked Moffett Field TMY3 package contains .epw, .ddy, and .stat files, but no TMY3 CSV. pvlib.iotools.read_tmy3 reads the NREL TMY3 CSV format; the EnergyPlus weather file would instead be read with pvlib.iotools.read_epw.

Which contract should the contribution follow?

1. Download the official EnergyPlus EPW and use read_epw, or
2. Keep read_tmy3 compatibility, in which case we need a different authoritative source for the Moffett Field CSV.

Once that is clear, I can add the data and a tested downloader without relabeling converted EPW data as source TMY3 data.

[s2t2]

Hi @ryux1 thanks for the info.

My thinking at the moment is that because energy plus provides lots of different regions and locations, we probably want to use it as our data source (unless we are able to identify an alternative data source that has as many different locations).

If the provided files are a different format (.EPW), that's fine, as long as we can read the files via the pvlib library (using the other function you mention). We will probably want to provide some brief explanation in the documentation about the format we are using.

@ecosang what do you think about us using .EPW files instead of .TMY3? Or did you have any notes to share about how you did a manual conversion from TMY3 to TXT to CSV or something? We probably prefer a pipeline with as few translation steps as possible, so maybe EPW could be ok?.

[ecosang]

I think eventually the typical workflow is to use tmy3 or real weather data for their simulations.
In that sense, the energyplus website is the good source to get epw format of tmy3 data, so it is good to have an epw converter to sbsim weather format. Since pvlib provides both read_epw and read_tmy3, it is not a bad idea to have both.
EnergyPlus has epw to csv converter, but this is not really tmy3 format (https://bigladdersoftware.com/epx/docs/8-6/auxiliary-programs/using-the-weather-converter.html). It just convert epw to .csv ,so user can open it via csv. Once we delete the several headers for metadata, the format is close to the tmy3. But, this one should come with EnergyPlus installation, so I don't recommend using this format.

In the pvlib's example, the tmy3 format is like this.
https://github.com/pvlib/pvlib-python/blob/main/tests/data/Burlington%2C%20United%20States%20SolarAnywhere%20Time%20Series%202021%20Lat_44_465%20Lon_-73_205%20TMY3%20format.csv

whitebox technology sell TMY3 dataset a while ago. Not sure how people use weather data nowadays.
http://weather.whiteboxtechnologies.com/

This one (https://github.com/IMMM-SFA/diyepw/blob/main/docs/tutorial.rst) say we can get epw format from the weather station data.

In my case, I download a epw data and then convert it to a CSV-readable format. Then, update the weather data with real observation and convert it to EPW again.

So, to sum up, using epw is good starting point. And then we think about how we can let users put their own weather data. For this, I think two ways.

• using read_tmy3 if the user has a tmy format.
• using the sbsim-specific format. User can put their weather data into a simple .csv format and convert it to sbsim-compatible weather data. (Because TM3 has lots of data like solar, precipitation, wind speed, etc., which we don't need in the sbsim simulation. The user doesn't need to make good tmy3 data for the simulation).