What are Cooling Degree Days?

cooling degree days

Sales of air conditioners in the US are largely dependent on weather.  I know that sounds obvious but what you may not know is that there is government tracked metric that shows how hot one summer was relative to past summers.   The US Air Conditioning industry uses this metric to determine how many sales were due to hot weather and how many might have been due to other factors.

The national Cooling Degree Days (CDD) metric is based on a population-weighted, national average of all regions compiled by the National Oceanic and Atmospheric Administration (NOAA).  This metric tends to correlate well with how much air conditioning is being used as the season goes on.

For example, from 2010 to 2013, the CDD ranged from 250 to 274 degree cooling days above normal when the U.S. saw some  of its hottest summers since the mid-1980’s.  In comparison, 2009 was very cool, at only 33 cooling degree days.  In 1997, we were 119 cooling degree days below normal, and in 1992, we were down to an amazing 177 cooling degree days below normal.


Was this helpful?

Vote This Post Up 74Loading...

4 thoughts on “What are Cooling Degree Days?

  1. Very interesting perspective that proved to be very useful to help my clients forecast residential AC Market demand. There are however some important questions you need to answer before comparing CDD to Residential AC sales:
    1- What is the base degree? Is the threshold based on empirical or scientific conclusions? I usually opt for 20° as it seems according to several researches to be the mid of the comfort zone.
    2- How can I end up with a representative pool of regions whose CDD needs to be computed? Should it be the regions whose sales are covered? or regions representative of the different weather profiles? or both of them?
    3- Are Residential AC sales exclusively B2C Sales or includes B2B sales (which are far from being responsive to temperature increase but rather depend on construction landscape and other factors…)?
    4- Once the relationship established and proven to be statistically significant (and powerful), Has the designed model showed satisfactory outcome when tested on historical data?
    Again thank you very much for this pretty interesting article.

    • Mehdi,
      Glad you liked the article. There is a detailed explanation of the degree days and how it is calculated on the National Weather Center Service website, here is a link to it – http://www.cpc.ncep.noaa.gov/products/analysis_monitoring/cdus/degree_days/index.shtml. The site defines degree days as this – “Degree day is a quantitative index demonstrated to reflect demand for energy to heat or cool houses and businesses. This index is derived from daily temperature observations at nearly 200 major weather stations in the contiguous United States.”

      Residential AC sales includes B2C and B2B however, B2C is much larger. Therefore any modelling should account for variability with the different drivers for each sub segment. Statistical correlation has proven to be very significant and should take into consideration both seasonality and regional factors. Do let us know if you have any more information worth sharing.

  2. Thank you for your question. The data we use comes from the National Weather Service Climate Prediction Center. They provide a very detailed break out of both cooling and heating degrees by location and time period. Specifically we are interested in the cooling degree days (CDD) deviation from nominal as opposed to total CDD, as the deviation provides a better correlation to actual air conditioning demand. Although it is not stated in the article, we also are primarily interested in the period from April to September as that is the time of year most air conditioners are purchased.

    As related to 2009, the CDD were higher than nominal that year, but significantly lower than any of the prior four years or next five years. If you are interested in viewing the detailed data it can be found at the following address: http://www.cpc.ncep.noaa.gov/products/analysis_monitoring/cdus/degree_days/index.shtml

  3. ” 2009 was very cool, at only 33 cooling degree days.”

    You mean 33 cooling degrees above normal (or below normal)? Normal is >1000, so 33 would be quite a dramatic event.

Let us know your thoughts

Your email address will not be published. Required fields are marked *