Question-and-Answer Resource for the Building Energy Modeling Community
Get s tarted with the Help page
Ask Your Question

What is the best way to create an EPW weather file from real data?

asked 2015-06-22 16:22:15 -0600

zachl1220 gravatar image

updated 2017-05-29 15:04:08 -0600

I've been trying to create an EPW weather file for EnergyPlus for a building in the Cardiff region, and have looked through various methods of gathering actual measured data per hour for approximately 1 year. It seems to me that if I don't want to purchase this data from a source such as Weather Analytics, then the best way would be to try and utilize Weather Underground's meteorological data. I've seen suggestions of utilizing BeautifulSoup methods to scrape the hourly data from Weather Underground, and I imagine that would be the best way given the length of the simulation. My current issue, though, is the lack of cloud cover/solar radiation information in the WU database. Any suggestions where I may reliably find this data on an hourly basis that I can supplement to the WU data for a sufficient EPW? Also, any other suggestions to improve my methods would be greatly appreciated, as this is my first time generating weather files for EPlus. Thanks

edit retag flag offensive close merge delete



Hi Neal,

Hadn't seen the second post you added, which helped a bit, but I still haven't been able to figure out a way to get more reliable solar radiation data besides converting verbal sky descriptions to numerical data. Does anyone have any suggestions for that?

zachl1220 gravatar image zachl1220  ( 2015-06-22 16:42:16 -0600 )edit

@zach1220 You may want to edit your question so that it is more focused on exactly that problem (creating solar data for EPW files). I know @Joe Huang has done this before, and might have ideas for you.

Neal Kruis gravatar image Neal Kruis  ( 2015-06-22 17:20:43 -0600 )edit


I'm actually working on the same topic for Cardiff as well. Therefore, I would like to know which technique advised by Joe Huang worked for you.

Thank you in advance !

Idhem gravatar image Idhem  ( 2016-07-12 08:54:04 -0600 )edit

2 Answers

Sort by » oldest newest most voted

answered 2015-06-22 20:47:26 -0600

Joe Huang gravatar image

updated 2016-09-23 01:24:35 -0600

(some of this repeats what I had posted on EnergyPlus_Support in answer to @zach1220)

When I work with WU data, I "borrow" the Cloud Cover data from the nearest Integrated Surface Database (ISD) station. The ISD is a large database maintained by the US National Climatic Data Center ( that contains roughly 8,000-9,000 active weather stations around the world.

In your case, there is a weather file for the Cardiff IAP in the ISD, which may have the same data as in the WU file, since WU contains both Personal Weather Stations (PWS), mostly in the US, and official weather station, i.e., the same ones that report to the ISD.

You might also look at what Lukas Rokka in Sweden did in 2012 with his "Real-Time Weather Converter" that can create a historical weather file combining the ISD data with satellite-derived solar from the Swedish Radiation Safety Authority (STRÅNG) for any place in Northern Europe, including the entire UK. I have never had need to try it, but you might find it useful ( unmaintained link - now at Shiny weather data). If you do, do me a favor and let me know what you find out. BTW, satellite-derived solar has been found to be very accurate, better than any cloud-base model.

edit flag offensive delete link more

answered 2016-09-23 12:59:30 -0600

Ricardo GAllegos gravatar image

There is a programm free to download called "Elements" that allows you to create weather files, or modify existingweather files, here is the link . Hope it helps.

edit flag offensive delete link more



Elements is a fine program but it's primarily for viewing and modifying an existing weather file, or converting the format of the file between EPW and BIN (DOE-2 format). Creating a weather file from weather records requires more work, especially if the recorded weather does not contain solar radiation, which is true 99% of the time (see previous answers for details).

Joe Huang gravatar image Joe Huang  ( 2016-09-26 05:42:58 -0600 )edit

Hi Joe, have you ever created a file and used E+ to split the global radiation into direct and diffuse ? I tried with a .csv and .def but it does not work for me. I think something is wrong in the way I create them. Any help would be really appreciated because the documentation is really lacking on that topic. thanks very much in advance!

Aurore gravatar image Aurore  ( 2019-04-03 04:55:12 -0600 )edit

Hi, Ricardo, No, I have not used E+ to do the direct/diffuse split. I didn't know that EnergyPlus has this capability, or are you referring to the weather.exe preprocessor that's a separate utility program in the PreProcess subdirectory? I've not tried using this preprocessor program either to do this, but another person who did 8 years ago found that it didn't work and produced results that didn't add up. I had notified the program developer, but don't remember whether it ever got fixed. There are many direct/diffuse split models. The main challenge is to calculate the sun angle.

Joe Huang gravatar image Joe Huang  ( 2019-04-05 18:51:47 -0600 )edit

Hi Joe, I am referring to the WeatherConvertor auxiliary program. Could you let me know if the program developer got back to you? Do you know why is this written in the documentation if it actually does not work? Do you know an equivalent program that I could use to do the split? Thanks very much and best regards.

Aurore gravatar image Aurore  ( 2019-04-16 02:16:03 -0600 )edit

Your Answer

Please start posting anonymously - your entry will be published after you log in or create a new account.

Add Answer


Question Tools



Asked: 2015-06-22 16:22:15 -0600

Seen: 5,513 times

Last updated: Sep 23 '16