Jump to content

User's feedback regarding Meteonorm vs PVGIS


julmou
 Share

Recommended Posts

Hello,

Here is my thought process: I thought Meteonorm was the most accurate in terms of irradiation data (mixture of satellite and weather stations data, accumulated over many years, great algorithm etc), so I was mostly relying on it to do my simulations.

Recently, I've been studying a site near Canberra (Australia), and these are the various annual Global Horizontal Irradiation I get depending on the source (always using same coordinates):

  • Meteonorm 8.0 ->                              1915 kWh/m2
  • PVGIS 5.2 (ERA5) ->                          1670 kWh/m2
  • Global Solar Atlas / SolarGIS ->         1750 kWh/m2
  • NASA-SSE ->                                     1693 kWh/m2
  • SMA Meteo. data / Sunny Design -> 1753 kWh/m2

As you can see, they are ALL around 1700, EXCEPT Meteonorm... which bothers me as I thought it was the most trustworthy.

Which is why I would be happy to share experience from fellow developers about your experience regarding accuracy of different weather sources.

When importing this Meteonorm data in PVsyst, it shows "Sat=93%" which means 93% of the data are from satellite sources, and only 7% from weather stations. I'm guessing, the bigger the part with actual ground stations (so the lower the sat. part), the better? I'm also guessing this has to do with the site being in the Southern Hemisphere. I'm pretty sure Meteonorm is great with Europe, but if some of you guys have feedback on the accuracy of these various meteo data sources in the S. Hemisphere, I would be very interested. Because now, I'm really doubting if I should keep using Meteonorm at all here (I get 1716 kWh/m2 if I average the 4 other sources, so Meteonorm in that case has a deviation of 12% from that value!!).

Link to comment
Share on other sites

Same thought on Meteonorm, i've always thought of it as the most reliable, came here to see what people thought of PVGIS 5.2

Ran a quick test in Perth, Meteonorm shows 100% Sat?

Source                Global                Diffuse

Meteonorm 8.0  2052.6                568.0

NASA-SSE         1836.3                565.2

PVGIS 5.2           1921.6                612.8

Gonna have a significant effect on results, so like to hear the consensus on what is more accurate?

 

Link to comment
Share on other sites

  • 1 month later...

Any more feedback from people using Meteonorm in the Southern Hemisphere compare to other sources of data?

So far, it would seem from my case and @cyoung 's case above that the Global Irradiation of Meteonorm in Australia tend to be quite higher than the other sources... (+ 12% as I detailed in my case above is quite significant!)

Link to comment
Share on other sites

Personally, I would be hesitant to trust Meteonorm data, except possibly in Europe.  If you read the documentation carefully, you will find that the data at each location is a mish mash of measured and satellite data.  Within the measured data, values can come from multiple locations different distances away from your desired location, and the number of years covered by the satellite data depends on location, with only two years used in many areas:

"The hourly pictures of the visible channel of the 5 geostationary satellites have been used (period 2008–2020 for MSG, 2019-2020 for Himawari and 2018-19 for GOES-E and Indoex)"  (Meteonorm 8, Handbook Part II: Theory, p. 4) 

(I have also never seen documentation of the quality of the satellite estimates, but it may be out there.)  The methods used to select and combine data from various sources is described in the same document.  As you are aware, more ground data is available in certain regions than in others.

I don't want to completely knock Meteonorm.  At the time it was first developed, it made excellent use of the data that was available, which is likely where its good reputation originated.  These days, though, it's possible to produce more consistent resource data sets using long satellite records and careful tuning.  You still can't beat a long-term, high quality measurement, though!

 

 

Edited by Lazare Fesnien
False URL
Link to comment
Share on other sites

  • 2 weeks later...
On 9/3/2022 at 8:41 PM, julmou said:

That's interesting, thanks for your answer!

Out of curiosity, which data source(s) do you trust most?

Also, what do you mean with careful tuning? Thank you!

You're welcome.  I spent quite a lot of time reading Meteonorm's materials.

At the moment, I would only use data from the major commercial providers, such as CPR, Vaisala, or Solargis.  Unfortunately, it's difficult to know which is the most accurate at any given location -- only validation work can prove this and the companies won't let anyone publish intercomparisons.  

By "careful tuning," I'm referring to the fact that it's not enough to start with good satellite data and throw some equations at it.  The provider must work on calibrating the satellite data, adjusting methods for different satellite instruments and different land surface types, and verifying that the results are consistent around the globe in order to produce good irradiance estimates.  

Link to comment
Share on other sites

Perhaps you thought I meant ground tuning.  I strongly support the use of ground tuning to reduce local biases in satellite resource data sets.  This can be very effective, if performed using quality measurements and appropriate tuning algorithms.  I lean towards letting the satellite data providers perform the tuning because they are the best experts on what affects solar irradiance in their models.

 

Link to comment
Share on other sites

  • 2 months later...

Hi all,

This is an interesting discussion, and reflects what we've observed as well in New Zealand. In the past we used Meteonorm, but when Meteonorm 8.0 was released we also noticed a significant increase in insolation. 

Due to this, we did an analysis where we compared the output of PVSyst modelling using Meteonorm 7.x, Meteonorm 8, SolarGIS, historical recorded data from local weather stations, and data that was recorded on site at a completed project (both irradiance & energy output, for 1 year).

Our findings showed that when compared to the actual output of the solar farm for 1 year, using Meteonorm 8 overestimated the modelled output of the farm by more than 10%, whereas the older Meteonorm was within 2%.

When the insolation estimate was compared to historical data from local weather stations, again Meteonorm 8 estimated significantly higher insolation compared to the measured & averaged historical data. In the end we chose to get a SolarGIS subscription, as their estimate was slightly lower than the measured historical average so it felt like a suitable conservative estimate for our commercial modelling. 

 

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
 Share

×
×
  • Create New...