Dreariness Index (2015)

How do you define dreary weather? Is it the amount of rain/snow? How about the frequency of precipitation? Many people feel that cloudy weather is dreary. Of course dreary does not have a scientific definition so some arbitrary measure must be developed.

In previous posts, I have looked at total rainfall, number of wet days, and cloud cover independently of one another. Now seems like a good time to combine these variables to come up with a single composite value.

Methodology

Three different variables are used in this analysis to come up with Dreary Index – total annual precipitation, number of days per year with measurable precipitation, and average annual cloud coverage. An inverses distance weighted surfacing technique was used to generate a gridded data set for the entire U.S. for each of the three variables. 

Grid cells are 25,000 meters on a side and a total of just over 2,500,000 grid cells for the entire U.S. are created. For each data set, the 250,000 grid cells with the lowest raw values are assigned a rank of 1, the next highest 250,000 grid cells are assigned a rank of 2, and so on. This creates 10 categories with an equal number of members. The procedure is then repeated for the other data sets. Afterward, there are three ranked data layers that each contain values of 1 to 10.

Since the grid cells for the different layers occupy the same geographical space, a raster addition calculation may be performed. In the absence of literature describing a dreariness methodology, I give equal weighting to all three categories. The variables are described in greater detail below.

Variable 1: Total annual precipitation. This was generated directly from the National Climate Data Center's (NCDC) published normal values for the 1981-2010 climate normal period. A total of 8,526 stations were used in this analysis.

Variable 2: Days per year with measurable precipitation. This was calculated from Global Climate Historical Network (GHCN) daily climate data. All days between 1981 and 2010 for the 1,751 "first order" stations in the U.S. were evaluated on whether or not measurable precipitation was observed. Stations with 15 or more complete years of data were utilized. A total of 952 stations met this criteria.

Variable 3: Annual cloud coverage. This data were obtained by analyzing 31 years of GHCN daily climate data (1966-1996) that contain entries for daily cloud coverage. A much smaller number of stations had a sufficiently robust (10 or more years) level of data coverage to use in the analysis. A total of 221 stations met the minimum data criteria. Because of the small number of stations, the gridded data are more generalized than for either Variable 1 or Variable 2.