• correlation;
  • precipitation;
  • southeast United States;
  • storm;
  • spatial and temporal;
  • weather observation network


Enhanced understanding of rainfall variability requires high resolution observations, yet most weather networks are unable to capture rainfall gradients that are important for modelling and decision-making. Using a network of 46 rain gauges over an area of 3100 km2 in southwest Georgia, this study evaluated different rainfall characteristics, including duration, intensity and correlation, and the impact of network density on spatially-aggregated rainfall. Distances among the observation locations ranged from 1 to 60 km, which makes the network significantly denser than most state-level weather networks in the United States. Results of the analysis of storm characteristics indicated that summer and winter portrayed the highest contrast. Most summer storms had an approximate duration of 1 h and were characterized by high rainfall intensities. Fall and spring appeared to exhibit a mix of characteristics from summer and winter. During the summer, rainfall occurrence at a location was quasi-independent of rainfall events at other locations in the study area while during the winter rainfall tended to occur or not occur over the whole study area simultaneously. During the summer, the correlation of rainfall amount was only high (0.90) in the immediate neighbourhood of the site considered (less than 3.3 km) while during the winter, this distance extended to approximately 8–18 km. To allow a coefficient of variability of 5% in the distribution of aggregate rainfall amount over the study area, the required network density sizes would be 16 gauges for summer and 4 gauges for winter, which translates into 5.2 gauges per 1000 km2 and 1.3 gauges per 1000 km2, respectively. These results provided insights into the characteristics and importance of rainfall variability at spatial resolutions ranging from local to regional. Results from this study may be applicable in the southeast United States and other areas with similar weather systems.