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

Reliability of ground temperature data in EPW files

asked 2018-01-18 08:26:37 -0600

dhollman gravatar image

updated 2018-01-18 08:28:53 -0600

I'm wondering how reliable the ground temperature data is in the EPW weather files on the EnergyPlus website.

For instance I was looking at Calgary, AB, Canadafrom

Ground temperatures are generated by the EnergyPlus Weather Converter program as far as I am aware; it does not originate with the rest of the data in those files (though it seems to be based on annual DB values).

Original file contents:

GROUND TEMPERATURES,3,.5,,,,-5.32,-0.71,4.61,8.57,14.09,15.26,13.38,8.88,3.00,-2.46,-6.33,-7.35,2,,,,-3.94,-1.30,2.37,5.40,10.36,12.27,11.91,9.37,5.31,1.02,-2.60,-4.38,4,,,,-1.75,-0.59,1.57,3.57,7.30,9.21,9.67,8.57,6.20,3.29,0.48,-1.33

Formatted more readably:

             Jan   Feb    Mar    Apr    May    Jun    Jul    Aug    Sep    Oct    Nov   Dec
0.5 m       -5.3  -0.7    4.6    8.6   14.1   15.3   13.4    8.9    3.0   -2.5   -6.3  -7.4    
2.0 m       -3.9  -1.3    2.4    5.4   10.4   12.3   11.9    9.4    5.3    1.0   -2.6  -4.4    
4.0 m       -1.8  -0.6    1.6    3.6    7.3    9.2    9.7    8.6    6.2    3.3    0.5  -1.3

So in January at 4m depth (about 13 ft.) it is below freezing.

There is some evidence that calls this 4m temperature into question. According to the City of Calgary:

Frost depth This is the depth to which the ground is frozen. In Calgary, frost depth usually reaches an average of seven feet every year. It takes several months to reach this depth and usually occurs in late February or early March.

Though they say "average" depth I'd actually expect in normal practice that a local frost depth guideline would rarely actually freeze.

Calgary's statement seems to be roughly consistent with other sources as well.

So not only does the 4m temperature seem to be far colder than would be expected by the frost depth statement, but also as the peak at that depth it occurs the wrong time of year, 1-2 months early.

What procedure is used to generate this data? Are there any references or documentation?


edit retag flag offensive close merge delete

4 Answers

Sort by » oldest newest most voted

answered 2018-01-18 20:47:39 -0600

Joe Huang gravatar image

updated 2018-01-22 01:00:18 -0600

To give credit where credit is due, the simple algorithm used in the EnergyPlus WeatherConverter program was lifted by yours truly from a 12-line GTEMP routine in DOE-2 and provided to the EnergyPlus Development Team, back in 2001 when I was still part of the Team. The GTEMP routine was coded up in the early 1980's utilizing the equation in the Kusuda report that gives the ground temperature as a simple sine curve with a lag and amplitude depending on the depth below ground and the soil diffusivity. Both of these parameters - 5 ft down, 1.0 diffusivity for moist soil (IP units) - were held fixed in the DOE-2 implementation, but in the EnergyPlus implementation ground temperatures are given at three different depths.

At the time I supplied this algorithm to the EnergyPlus Team, I had suggested that they just use the equation rather than the calculated monthly values, because I didn't like the staircase appearance of the monthly temperature and thus heat flows through the foundation in DOE-2. I'm not sure if that's also true for EnergyPlus, or whether somebody added an interpolation to the routine at some point.

It's always been known that these ground temperatures are very approximate. If there's a high water table or a geothermal hot spot, the temperatures could be off a lot. Also, these temperatures are meant for undisturbed soil away from any building. In truth, any construction above grade will change the ground temperature below.

edit flag offensive delete link more



I learned a couple of days ago of a new, much improved ground temperature model developed by Jeff Spitler of Oklahoma State University et al. Instead of adding that information as another comment on this thread, I decided to start a new one in hopes of getting this significant work the attention that it deserves. See thread on "Has there been any improvement in calculating ground temperatures since the 1973 Kusuda study?"

Joe Huang gravatar image Joe Huang  ( 2018-01-25 18:13:52 -0600 )edit

answered 2018-01-19 03:47:10 -0600

RCulham gravatar image

Environment Canada in their Canadian_Climate_Normals_1981_2010_Calculation_Information.pdf document states that they take daily measurements of the soil temperature in accordance with the WMO recommendations at the standard depths or 5, 10, 20, 50, 100, 150, and 300 cm. I happened to send an email today asking Environment Canada if they are recording this data at our local weather station for which there is an epw file. It would be interesting to see how this compares to the Kusuda formulation.

edit flag offensive delete link more


Yes please follow up here when you get a reply from them that may be really interesting! Thanks

dhollman gravatar image dhollman  ( 2018-01-19 06:24:20 -0600 )edit

answered 2018-01-18 17:55:52 -0600

See this note in the Auxiliary Programs Reference:

The ground temperatures in .epw files are calculated from "a heuristic, time lagged calculation based on dry bulb temperature and location. References on the topic are found in Kusuda (see references)."

References are:

Kusuda, T., “Earth Temperatures Beneath Five Different Surfaces”, Institute for Applied Technology, NBS Report 10-373, 1971, NBS, Washington DC 20234.

Kusuda, T., Least Squares Technique for the Analysis of Periodic Temperatures of the Earth’s Surface Region, NBS Journal of Research, Vol. 71C, Jan-Mar. 1967, pp 43-50.

I was able to find .pdfs of both Kusuda documents with a google search.

edit flag offensive delete link more

answered 2018-01-18 08:47:31 -0600

Avi gravatar image

Try reading that post.

Generally you probably want to consult the Engineering Reference and the Input Output Reference.

Specifically your observations are supported by what discussed in the referred post and in the Input Output Reference.

edit flag offensive delete link more

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: 2018-01-18 08:26:37 -0600

Seen: 872 times

Last updated: Jan 22 '18