Using Climate Variability to Predict Annual Precipitation and Estimate the Persistence of Climate Extremes for Major Urban Areas and Regions within the United States

Abstract:

Relationships between climate variability and precipitation in several urban areas throughout the United States are developed using various global climate indices. Precipitation data for over 1200 stations are obtained from the United States Historical Climatology Network maintained by the National Climate Data Center, NOAA. All data are averaged over an extended period (up to five years) and correlated to several climate indices averaged over a period of equal length using lag times also up to five years. The period length and lag time are optimized in order to produce the highest correlation. The index that best correlates with precipitation for each urban area analyzed in the current study is identified and used to create regions within the United States that are predominantly affected by a particular index; strong correlations (r² values > 0.70) were found in all regions. The final result is a map of the United States that displays the spatial distribution of each region. These results, which include the specific relationships developed for each region and urban area, will not only allow a greater understanding of the major mechanisms that are responsible for rainfall variability throughout the United States, but will also result in improved predictability of precipitation over multiple time scales, including seasonal and annual. In addition, the ability to predict total rainfall for periods greater than one year will allow an estimate of the persistence of trends and extreme events, such as periods of drought or above-average rainfall, to be made in advance; how far these projections can be made in advance depends on the lag times used to create each site-specific and regional correlation. An example related to the California Drought is given.