Seasonal Mean Temperatures and Precipitation for the United States during Strong El Ninos

Two types of maps encompassing the United States are provided for overlapping three-month periods, the November/December period, and the December through March period. Both types of charts are based on ranking mean precipitation from the 102-year climate division record for a particular period (like January through March) from wettest (rank 1) to driest (rank 102) year, and mean temperature from coldest (rank 1) to warmest (rank 102) year.

One set of charts show the average ranks of temperature and precipitation for each three-month or other period for those years with strong El Nino episodes in progress during the period. These are straight forward to interpret.

The second set of maps contains more detailed information than the first set, but requires additional description to interpret. Specifically, these maps show the number of times in the past during strong El Nino episodes that the mean temperature (precipitation) ranked among the warmest (wettest) or coldest (driest) third of the 102-year climate division record dating back to 1895. This information is depicted only for those climate divisions where the distribution of occurrences in the tercile classes of above, near, and below normal departed sufficiently from a uniform distribution that the odds of the departure being an accident were less than about 10%.

Thus, this set of charts not only provides insight into what kind of conditions El Nino favors in a specific area for a specific time of year, but also how reliably those conditions were observed in past strong El Nino episodes.

The number of cases that fall in the tercile class that the distribution is skewed towards is color coded; for example on the seasonal precipitation charts when most of the El Nino years at a location were wet the number of years is denoted by a green shade covering the climate division, but when most were dry the number of cases is indicated by a brown to yellow shade. For each colored division on a chart the number of El Nino years that fall in the opposite tercile class (for example the driest third for divisions that are colored a green shade on the precipitation charts) is denoted by a number. Thus for every three-month or other period at every climate division the complete distribution among temperature and precipitation terciles for strong El Nino episodes can be determined from the diagrams.

For specific examples refer to the [October thru December] precipitation chart constructed from 11 strong El Ninos: For coastal Southern California seven of the El Nino years were among the wettest third, three among the middle third, and only one of the eleven among the driest third, while for Long Island the distribution was almost the opposite, i.e. two among the wettest, two among the middle, and seven among the driest. Thus, suitably smoothed, the information on the diagrams can be used to formulate a priori probabilities of the different temperature or precipitation classes conditional on a high-confidence El Nino forecast.


For each state, key periods have been selected to highlight El Nino effects on the particular state with two charts. The first type of chart is a map of selected statistics, division by division, for the state to contrast conditions during strong El Nino years with normal conditions. The other is a set of bar graphs for each division within the state depicting the same information about the distribution of temperature (precipitation) for strong El Nino episodes among the warmest (wettest), middle, and coldest (driest) thirds of the 102-year climate division record.

The years representing strong El Ninos change from period to period. This is because the part of the year for which the central Equatorial Pacific sea surface temperatures (SSTs) are well above normal differs from episode to episode. The cases that were included are those for which the average SST in a prescribed area was close to or greater than one degree Celsius above normal in all (two to four) months spanning a particular period. The key area used for case selection is bounded by the Dateline and 150 west longitude and 5 north and 5 south latitudes. This area was used because it approximates the region in the equatorial Pacific where tropical convection and rainfall (the major source of atmospheric energy in the tropics) are the most sensitive to relatively small changes in SST. Thus, the SST anomaly in this area should be a good index of how strong an El Nino's impact on the global atmosphere will be.

Note also that precipitation charts are constructed from more years than the temperature charts; for the former El Ninos prior to 1940 were included as candidates for selection but not for the latter. This is because there are obvious large-scale trends in the temperature data but not in the precipitation data. Even with this restriction to the 1940-1996 period the temperature series are still likely to be somewhat nonstationary, so the temperature distributions should be used with more caution than the precipitation distributions. In particular, we have observed that for the strong El Nino cases from the 1980s and 90s the northern United States has overall been considerably warmer in the January through April period than earlier in the post-1940 record. Thus temperature distributions for these cases are more skewed towards the warm tercile than shown here for JFM and FMA; areas where the full period distributions are skewed towards the warm tercile are more skewed towards warm for the recent period. However, for FMA in the Southeast this difference in skewness between pre- and post-1980 is only slight.

The diagrams shown here for both precipitation and temperature reflect, complement, and extend the information recently presented by Livezey et al. (1997: J. Climate, 10, 1787-1820; hereafter L97). For example, the area centered on Montana in the precipitation distributions for NDJ, DJF, and JFM was not highlighted in the text of L97 and only hinted at in their Figs. 14 and 15a-b therein, yet emerges as a prominent teleconnection in this analysis. Likewise, it was not possible from L97 to fully appreciate the remarkable consistency of some of the strong El Nino precipitation signals shown here, particularly for the wetness for different parts of the Southwest in a number of the charts, the wetness in Florida for NDJ and JFM, the aforementioned Montana dryness, and the dryness in the Ohio Valley for JFM. With the benefit of the L97 maps it is apparent that the OND precipitation signals are dominated by November and December while the FMA signals are largely a consequence of February and March. Thus, the probabilities of OND precipitation terciles implied by the OND chart should not be automatically applied to the month of October, etc. A final point to note in the precipitation charts is a modest tendency for bimodality in the mean precipitation in the center of the country for OND.

Temperature Distributions

Despite the caveat about the possible nonstationarity of the temperature data series mentioned in the discussion above about case selection, there is still very useful information in the effected JFM and FMA charts that is insensitive to this problem. Specifically, there are no reported instances of average temperatures in the above normal tercile during strong El Nino episodes for extensive areas of the Southeast in either period. In this same context, not only is there substantial skewness towards the above normal tercile over the north central United States in NDJ and DJF, there are also a very large number of climate divisions with no instances of average temperatures in the below normal class. The economic implications of these observations should be considerable. Lastly, the warning against use of three-month precipitation tercile probabilities for individual months within a period is equally applicable here, especially with respect to the OND temperature chart that is mostly dominated by November and, especially, December (note that detailed ND information instead of OND is provided for eastern states).