Open-Meteo Solar Forecast

[dm82m (Dirk)]

and most importantly: this plugin here uses open meteo api and NOT solcast. it has nothing to do with solcast!

[Neil_Brownlee]

I get that, so if I put in 130 - the integration subtracts 180 and gives -50 which is wrong though? My actual azimuth is technically -130. Subtracting 180 only works if you are facing South with this logic.

[dm82m (Dirk)]

cant answer that, we need to wait for @rany

[rany]

If your azimuth is negative, just add 360. So if it was -130 (0=North), then just add 360 it would be 230.

[slipx06]

So I made a mistake when entering some details. Apologies for going down this rabbit hole. Doing the comparison now and its much closer.

Daily Meteo 33.84
Daily Solcast 34,3

[MelleD]

I don’t know if a comparison really makes sense.

I think Solcast isn’t 100% right either. Just because it’s different doesn’t mean it’s bad…

In my case, I think the forecast is quite good. Let’s see how it turns out in the long term :).

Today Open Meteo

Solcast 5 day ago

[yfsk (SK)]

Neither SolCast nor OpenMeteo will ever be 100% accurate. Guys! It’s a forecast!!! Nothing else.

And OpenMeteo is doing that extremely well. Congrats and thanks for the work!

[michal_p]

Hi, just wanted to share my to experience with Open-Meteo Solar Forecast. I originally was using Forecast Solar (good at the beginning, wildly unusable later) and Solcast (solid). Now, the Open-Meteo Solar forecast is giving me very accurate predictions (Czechia).

I mean it can’t be much better than this, right? It’s from couple of days ago. (There’s a tree casting shadow over my east string in the morning).

This is from today.

All in all, great integration and many thanks for your work! Much appreciated!

[rany]

Thanks a lot for the words of appreciation, it means a lot!

[yfsk (SK)]

Honestly, this works really, really good and I absolutely appreciate the work and time @rany spent for this. Many, many thanks! :o)

I have one suggestion, though:

https://doc.forecast.solar/damping

That would help to fight the overestimation especially in the morning and evening hours, due to trees or hills in the east and west. :o)

[bengy56]

Second that - really great work, much appreciated !

One question - maybe I missed something, but how do you view/extract the detailed daily forecast information (hour by hour) ?
In Solcast it is available as an attribute list on one of the entities.
I would like to use this information in a ‘bespoke’ Apex graph, but could not find this sort of information attached to any of the published entities.

[rany]

You can’t currently

[rany]

Hey friends, feel free to look into:

• Issues · rany2/ha-open-meteo-solar-forecast · GitHub
• Issues · rany2/open-meteo-solar-forecast · GitHub

If there is anything missing that you think you want, feel free to make an issue yourself or if you don’t have a GitHub account I could make it on your behalf

[rany]

Hey, I have a quick question for you. If you got this data from service response, would you be able to use it in ApexCharts (I’m not an ApexCharts user so I wouldn’t know)

[gieljnssns (Giel)]

I think if you need it for apexcharts you have to make a template yourself.

Apexcharts and Openweathermap Forecast examples

I am trying to get it right since hours, but I can’t figure out what is actually the problem. I setup the template sensor as described above. My sensors forecast attribute is looking like this: [{'datetime': datetime.datetime(2024, 6, 2, 17, 0, tzinfo=datetime.timezone.utc), 'precipitation': None, 'precipitation_probability': None, 'cloudcover_percentage': 100, 'cloudless_percentage': 0, 'cloud_area_fraction': 100, 'cloud_area_fraction_high': 100, 'cloud_area_fraction_medium': 2, 'cloud_area_fr…

[rany]

Perfect, I’ll do service response then

[michal_p]

What is the shortest time frame we can get, please? Since we’re migrating to 15 minutes long time frames (starting each 00, 15, 30 and 45) on the electricity spot market, it would be helpful to pair it to fitting prediction… Thanks!

[rany]

What do you mean by shortest time frame? The intervals for the data or when I’ll be able to work on this?

If you mean the intervals for the data, it’s always going to be 15mins. The average at every 15min interval and the instantaneous data at every 15min interval will be provided.

If you mean time frame for the feature complementation, hopefully by tomorrow.

[rany]

@rany

The average at every 15min interval and the instantaneous data at every 15min interval will be provided.

For regions that are not EU or the US, the irradiance data is interpolated but it works fine for me despite this.

[michal_p]

I meant the data (intervals). When you’ll have time for this is completely up to you and I’m grateful you’re putting your own time to this. Thanks!