Developing a Regional Correlation Function for Rainfall near Hamilton, Ontario
2001
Valeo, C. | Tang, D.U.H.
In this paper, we develop a model for the second order process of rainfall in an area of southern Ontario. Hourly rainfall data collected from 17 gauges over a period of five years were used to develop a correlation function for rainfall. Eighteen test cases were used to investigate the optimized range, the appropriate function form, and the impacts of rainfall type and gauge error on the estimated correlation function. An exponential model for the correlation function was selected over a spherical and gaussian model. It is suspected that gauge error greatly affects the size of the range, which decreases as the uncertainty in the data increases. Three distinct values of range were determined and the differences in range were attributed to the number of thunderstorms observed in the region during the test periods. The range spanned from 26 km in periods of heavy thunderstorm activity to 75 km in periods of mostly cyclonic activity. High values of the coefficient of determination were found when combining years of data with range values of just over 50 km. The difference in the standard error associated with a rain gauge network can be significant depending on the range chosen. With regard to devising rain gauge networks to determine average rainfall, an average range of 50 km is recommended.
显示更多 [+] 显示较少 [-]