Relationships between ENSO and Daily Precipitation in the United States

Relationships between warm (El Nino) and cold (La Nina) episodes and winter weather features over the United States have been documented in publications too numerous to mention. An excellent tutorial which summarizes ENSO related effects on United States weather is found on the CPC Home Page http://nic.fb4.noaa.gov/products/analysis_monitoring/ensocycle/enso_cycle.html. There is relatively little in the literature, however, relating El Nino and La Nina to daily precipitation statistics (i.e. "storminess") over the United States.

There is a perception that "storminess" has increased in the United States in recent years. It remains unclear whether such an increase (if it exists) is attributable to more frequent warm (El Nino) episodes, and / or to secular trends (independent of El Nino). In many areas of the United States there has been a notable increase in the number of catastrophic flooding events in recent years. Familiar examples include the spring/summer floods of 1993 along the Mississippi, the winter floods of 1997 in the Pacific Northwest and Northern California and the spring floods of 1997 along the Ohio River and Red River Valley. In general, these floods are not the result of a single extreme precipitation event, but rather a series of extreme events over an extended period. Consistent with this, there is evidence for a recent increase in the proportion of the area of the United States affected by above-normal frequency of extreme daily precipitation events (e.g. Karl and Knight 1998).

Changes in precipitation are often quantified in terms of changes in the total precipitation over long averaging periods, e.g. annually, seasonally or monthly. However, most precipitation events in the midlatitudes last a few days at most, which implies that finer temporal resolution (e.g. daily) is required to reveal important aspects of how precipitation changes within the long averaging periods. For example, a fundamental question that can be addressed with daily data is how much of any precipitation increase or decrease over a period of years is attributable to changes in the number of wet days versus the amount of daily precipitation? That is, increased precipitation could be derived from an increase in the number of wet days during the year, and they may be equally distributed for light, moderate and heavy daily precipitation amounts. Alternately, one could envision a situation where the number of days with precipitation does not increase, but the amount of precipitation changes for, say, the heavy rain days.

In this ATLAS we examine whether there have been detectable changes in total precipitation and / or in the number of daily precipitation events since 1948 and, if so, how large these changes have been compared to precipitation anomalies typically associated with El Nino. Changes in total precipitation and in the number of precipitation events are assessed by examining differences between the periods (1977-1996) and (1948-1976). These two periods were chosen because there is a sizable body of literature that argues for an abrupt climate shift in the mid-1970's (studies too numerous to mention). Comparisons of differences using all years to differences in which warm and cold ENSO episodes are ignored (i.e. the "neutral" years) yield a rough estimate of the secular trend in precipitation. When differences for all years and for the neutral years are similar, we argue that contributions due to ENSO are relatively small and vice-versa. In order to determine whether these changes are large or small, we compare them to mean precipitation anomalies associated with moderate / strong El Nino episodes. There are several reasons why it is appropriate to compare such changes to anomalies associated with El Nino: (1) there have been two 100-year warm episodes (1982-83 and 1997-98) in the period since 1977; (2) the 1990's have been characterized by warm episode conditions more often than not, and (3) there has been only a single strong cold episode (1988-89) in the period since 1977.

In order to make the assessments described above, however, it is first necessary to document seasonal relationships between warm and cold ENSO episodes and daily precipitation statistics over the conterminous United States. Such an analysis allows us to place changes in total precipitation in the proper historical perspective. For this purpose we employ a set of gridded daily precipitation analyses over the conterminous United States (Higgins et al. 1996). These analyses were developed from hourly observations for approximately 2500 stations obtained from the NWS / Techniques Development Laboratory, who compiled and quality-controlled station data archived at the NOAA National Climatic Data center. The analyses were gridded to a horizontal resolution of 2o latitude by 2.5o longitude using a Cressman (1959) scheme with modifications (Glahn et al. 1985; Charba et al. 1992). The time domain for this data set is 1 July 1948 to 31 December 1996.

Section 2 describes the season-by-season classification of warm and cold ENSO episodes in the tropical Pacific since 1948. Section 3 provides an in depth analysis of the seasonal effects of warm and cold ENSO episodes on the number of wet/dry days and on the total seasonal precipitation over the conterminous United States. In the analysis wet days are defined as those on which measurable precipitation occurred (precipitation amounts greater than or equal to 1.0 mm day-1). A few aspects of the annual cycle of daily precipitation in the conterminous United States. are explored in section 4 using selected major cities. Decade-scale changes in total seasonal precipitation and in the number of wet days are examined in section 5.