Premium
Simulation of Normal Annual and Diurnal Temperature Oscillations in Non‐Mountainous Mainland United States
Author(s) -
Fernández C.J.
Publication year - 1992
Publication title -
agronomy journal
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.752
H-Index - 131
eISSN - 1435-0645
pISSN - 0002-1962
DOI - 10.2134/agronj1992.00021962008400020023x
Subject(s) - amplitude , environmental science , diurnal cycle , elevation (ballistics) , latitude , atmospheric sciences , equinox , climatology , diurnal temperature variation , annual cycle , sine wave , meteorology , mathematics , geology , geodesy , geography , physics , geometry , quantum mechanics , voltage
Simulation of air temperature data may add flexibility to many computer programs for modeling crop growth. An empirical model was developed to simulate both average annual air temperatures and average diurnal oscillations of air temperature, at any location within the USA east of the Rocky Mountains and with elevation below 1500 m above sea level. The target region was divided into four zones according to average annual rainfall and proximity to the Atlantic Ocean. Annual oscillations of the daily mean, maximum, and minimum temperatures for a given location were simulated with single sine‐wave equations. The average and amplitude parameters of these equations are calculated from estimates of the monthly maximum and minimum temperature normals of January and July. These four temperature normals were estimated using regression equations on latitude. Different sets of equations were used for each location zone. Diurnal oscillations of temperature on a given day at the given location were simulated by joining two sine waves to give a single smooth asymmetrical wave. The average and amplitude parameters of these diurnal oscillations were obtained from the annual curves previously simulated. The accuracy of the model in simulating annual oscillations of daily maximum and minimum temperatures was tested. Simulated and observed average monthly temperatures recorded at 27 weather stations correlated with r = 0.993 (significant at Ρ = 0.01). Plots of residuals against months showed that the model underestimated the monthly temperatures around the vernal equinox (average residuals ranged from −1.76 to −2.66°C). The accuracy of the model in simulating diurnal oscillation was also tested. Simulated and observed temperatures at 3‐h intervals for 18 cloudless days and 18 overcast days correlated with r = 0.992 (significant at P = 0.01). Plots of residuals against time showed that the model underestimated the temperature during the warming phase in clear‐sky days (average residuals were about −3°C.