Date of Award




Document Type

Master's Thesis

Degree Name

Master of Science (MS)


Department of Atmospheric and Environmental Sciences

Content Description

1 online resource (ii, 44 pages) : color illustrations.

Dissertation/Thesis Chair

Paul E Roundy


Electric power consumption, Energy consumption, Electric utilities, Weather forecasting

Subject Categories

Atmospheric Sciences


A seasonal harmonic linear regression approach is presented to model the seasonal and diurnal relationships between surface weather observations and electricity load data in New York State. The relationships between dry-bulb, dew point, and wet-bulb temperatures with electricity load were evaluated using the correlation coefficient to test the strength of the relationships, while the regression slope coefficient provided an interpretable scale for those relationships. We found that the strongest seasonal and diurnal relationships occur during boreal summer from the afternoon through the overnight hours, while similar but negative relationships are observed during the winter. Using the same seasonal harmonic linear regression approach a statistical model was constructed to predict electricity load using only ambient weather predictors for the four largest electricity consuming regions in the state. This model is unique to other electricity load models as its intent was to simulate the weather/electricity load relationship by including only meteorological predictors. Using the model we were able to conclude that 66% of the population-weighted variance in New York State’s electricity load can be explained by variability in the ambient weather predictors.