Click to Go Back to Data Page

Population Weighted Degree Day Indices

Methodology #1
The first methodology uses 120 U.S. cities, with the weight assigned to each city just being the fraction of that city's population to the sum of the population for the 120 cities used in the calculation. All weights are based on 2000 census estimates.

Methodology #2
This technique is similar to #1 except that the entire U.S. population (of the 48 contiguous states) is assigned to the 120 cities such that the population of a given city is the population of that city, plus surrounding areas. All weights are again based on 2000 census estimates.

Methodology #3
The most recent and most accurate population weighted degree day index assigns the population of each county to a given point based on both nearness to that point and climate conditions in that county. Coastal counties will be assigned to coastal city locations whenever possible, even if it is not the closest city, for example. Population weights vary by year based on yearly Census Bureau population estimates from 1970 to present. The result is that some southern and western cities have seen an increase in their weightings over time while some northern cities have seen a decrease in their weightings as populations across the U.S. shift. The main advantage of Methodology #3 is the ability to more accurately compare data between distant years.

Natural Gas Heating Weighted Degree Day Indices

Note that Methodology #4 is the most recent and most accurate of all the NG DD indices. Be aware that regional or U.S. composite degree day estimates are meant to be used as an index and so the degree day values on a given date may vary significantly from methodology to methodology (both between our methods and methods from other vendors or NOAA). All degree day values are effectively weighted averages, with the weighting techniques being the difference.

Methodology #1
This was the first attempt Frontier Weather made at producing a Degree Day index that would be useful in projecting weekly NG storage numbers. About 120 U.S. cities are used in the calculations and each city is weighted within one of four census regions based on the population of that city. The Midwest and Northeast regions are then combined to produce a proxy for the East Consuming region. The resulting three regions are then gas weighted (based on home heating demand) to produce the U.S. index value.

Methodology #2
The methodology in the 2nd version of the Degree Day index grouped cities by EIA gas region instead of census region. Also, the entire U.S. population was assigned to the 120 cities instead of just using the metro area populations for each city as in methodology #1. This resulted in some cities being weighted more and others less than in the Methodology #1.

Methodology #3
Gas weighted degree days are calculated in this methodology in a similar manner to that in Methodology #4 below except that cities are grouped by census region with the Midwest and Northeast census region grouped together to serve as a proxy for the East Consuming region. The main difference between this set of data and Methodology #4 is that the census regions do not line up exactly with the EIA regions. So, the regional values may not be as good of a match, though the U.S. data, when composited provides a good indicator for total U.S. gas demand. This methodology exists as it is calculated in the same manner as Methodology #1 (which was the first degree day estimate technique produced by Frontier Weather) but uses the varying weights by year based on population changes.

Methodology #4
Gas weighted degree days are calculated similar to Methodology #2 above, but with improved weightings that vary by year. The population of every county in the country is assigned to one of about 120 cities across the country. The population assigned to each city was done more precisely than in Methodology #2. Gas weights for HDDs for those points are then determined by a combination of population and the percent of the population in that area that uses NG for home heating. So, the population used to assign weights to cities in New England is reduced by the percent of the population that uses heating oil, for example. Cities are weighted within each of the three gas regions (East Consuming, West Consuming and Producing) and then the three regions are weighted and added together to come up with a single value. Gas weights for CDDs are determined by a combination of population and the natural gas power generation in the region. All weights vary by year based on U.S. Census Bureau population estimates from the year 1970 through the present. A combined HDD + CDD value is also present in some data files. The weighting of the CDD values is effectively determined by finding the fraction of the CDD value that produces the best fit between the resulting combined value and EIA storage numbers. This is done for each year of data, with a smooth curve fit to the resulting coefficients. That curve is then used to determine what weight the CDDs should have for each calendar year. Since power production from natural gas fired power plants has increased over time, the weighting of CDDs generally increases with time.

Other Degree Day Indices
Population weighted data is also available in state by state format as well as by power region.

Oil Weighted Degree Days
This index effectively is an average of degree days across New England as home heating oil consumption is sparse across the remainder of the country.