Solar Arbiter (maximum distance limit exceeded)

[Marvin Valentin]

Hello everyone:

I am learning how to predict AC power from a certain location in Europe with PV system. I have learned through this group the solar arbiter.

Every time I change the longitude and latitude of the code from the GitHub, it always result to an error which says 

ValueError: Maximum distance limit exceeded

Thanks for the response in advance.

Marvin

[cwh...@sandia.gov]

Hi Marvin,

That's not a pvlib error message, it may be from one of the dependencies. Can you post the code that produces this error?

Cheers,

Cliff

[Will Hobbs]

Marvin,

Since this isn't directly related to pvlib, it might be best to move this discussion to https://github.com/SolarArbiter/solarforecastarbiter-core/discussions. Others here please comment on that, though. 

That said, here is some info that might be helpful: 

The solarforecastarbiter package comes with small, regionally subsetted NWP files (in NetCDF, or .nc file format) that are intended for testing/demonstration. 

For example, the workshop (https://github.com/SolarArbiter/workshop) file reference_forecasts.ipynb uses the HRRR subhourly NWP file here: ...\solarforecastarbiter\io\tests\data\hrrr_subhourly\2019\05\15\00 (I created a virtual environment called "sfa_venv" in a "python" folder in my user directory in Windows, so the full path for me is "C:\Users\willh\python\sfa_venv\Lib\site-packages\solarforecastarbiter\io\tests\data\hrrr_subhourly\2019\05\15\00".

If I open that file, hrrr_subhourly.nc, in a tool like NOAA's WCT (https://www.ncdc.noaa.gov/wct/), I can see that it only covers a small area around Tucson, AZ, USA (a regional subset).

Since Europe (and everywhere that isn't right around Tucson) isn't covered in the included NWP files, you would need to get additional files with the correct coverage. This also means that you will need to pick the correct NWP - the HRRR (used in the workshop demo) does not include Europe, and I would recommend the GFS model. I think two NetCDF NWP files for the GFS are included in the sfa package (look in the same folder as the "hrrr_subhourly" directory mentioned earlier), so you can start experimenting with the GFS right away. Note that they are also limited to a small region in Arizona. I would not recommend the GEFS for a beginner - it gets extra complicated.

Next, you will also need to get recent NWPs to get a "real time" forecast (in contrast to the 2019 data included with the sfa package). To do all this, you will need to setup and run the process to "fetch" the NWP files from NOAA's NOMADS server. This may require one of the following: 1) a Linux machine, 2) the ability to run Docker, or 3) customizing solarforecastarbiter-core for another use case. Introductory pointers are in the FAQ here https://forecastarbiter.epri.com/faq/#core. 

In my experience, generating forecasts with solarforecastarbiter-core requires a good bit more effort than setting up and running pvlib. That isn't to discourage anyone from trying, but just to set expectations. I've documented my efforts here: https://github.com/williamhobbs/RPi-Solar-Forecast. 

Will

[Marvin Valentin]

Dear Will

Thank you for always responding to our queries. I have attached herewith the code to the end where it gives the error.

Thanks,

marvin

MARVIN T. VALENTIN

PhD Student

Department of Applied Bioeconomy

Wroclaw University of Environmental and Life Sciences

FanPage: https://www.facebook.com/profile.php?id=100083067135867

--
You received this message because you are subscribed to the Google Groups "pvlib-python" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pvlib-python...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/pvlib-python/cd435925-0f70-441b-9aee-0f5d2b35a7e6n%40googlegroups.com.