El Niņo Implications for the 1997 Southwest
Drought and Fall/Winter Precipitation

revised August 4, 1997

1. El Niņo and its importance to us

El Niņo is one part of a multi-year cycle of the atmosphere and ocean interacting together primarily in the tropical Pacific. The atmospheric circulation in this region depends on the distribution of tropical sea surface temperatures (SSTs). However, the distribution of SSTs is partly determined by the overlying atmospheric circulation; hence, the term "coupled interaction" or a "chicken-egg" kind of relationship between the ocean and atmospheric circulations. One phase of the cycle occurs when SSTs in the central and eastern Pacific are warmer than usual. This is called an "El Niņo." When they are colder than normal, this period is called a "La Niņa". The shift from warm to cold SSTs and back again occurs irregularly every three to five years.

El Niņo (and La Niņa) are important for us in the United States because the atmospheric circulation changes associated with them extend globally even though the ocean-atmosphere interaction that produces them occurs mostly in the near-equatorial regions. This can easily be demonstrated using results from a computer experiment (Figure 1) that mimics nature. At the beginning of the experiment a typical El Niņo SST pattern is introduced into the tropical Pacific. By the second day (top panel) the atmosphere has warmed up in the vicinity of SST changes. By day six (middle panel) this heating has become quite pronounced and the impacts of this are clearly seen spreading eastward along the equator and northward (and southward) into the subtropical and mid-latitude Pacific. There already is evidence of weak changes over North America. By day twelve, changes are evident over much of the global atmosphere; however, the largest changes remain in the vicinity of the initial changes to SST in the tropical Pacific and to the northeast over North America.

This link between events in the tropical Pacific Ocean and resultant influences in regions outside the tropics occurs in nature on a variety of timescales. As in this computer experiment, it can happen over a multi-week period. This kind of process was thought to be important in the heavy rain events in northern California and Oregon in late 1996/97. It can also evolve over a several month period. This is the El Niņo-related phenomena. We think that multi-year to decadal changes in the tropical Pacific SSTs can also generate related patterns.

The temperature changes in the atmosphere (Figure 1) result in modifications in the atmospheric circulation in mid-latitudes and hence the tracks storms take. The impact of this is to change the normal pattern of rainfall. The nature of the rainfall changes in this numerical experiment is shown in (Figure 2). Over the United States, El Niņo reduces the rainfall in the Pacific Northwest and increases it over much of the southern part of the country. Another prominent change is the enhanced area of rainfall off the West Coast of the United States at about the latitude of northern California. Larger changes in tropical SSTs than used in this particular numerical experiment extended this rainband eastward into northern California. The likelihood of rain in this region depends on the size of the El Niņo SST changes. This dependence on the amplitude of the El Niņo is one reason why a few events result in near- or below-normal rainfall in California. This coming winter the forecast SST changes are very large; hence the forecast for California is for above normal rainfall. The overall pattern that is shown here is typical of what might be expected for December through March. Patterns for the warm half of the year are not as clear as this. Also, in actuality, there is more regional detail in the rainfall distribution than is shown in this figure.

One way we learn about the impacts of El Niņo over the United States is to examine historical data sets and ask the simple question, "when SSTs in the tropical Pacific are quite warm, what is the average rainfall (or temperature) pattern over the United States?" By averaging over seven El Niņo's during the past 45 years, one gets an El Niņo related rainfall pattern which is quite similar to that in (Figure 2). When the tropical Pacific is cold, i.e. La Niņa, roughly the opposite distribution of rainfall is observed, i.e. the northern part of the United States experiences above normal rainfall and the southern part below normal rainfall. However, there can be differences in detail.

Over the last 65 years, about twelve El Niņo's and La Niņa's have been observed. These are respectively indicated by the circles and triangles in (Figure 3), where it is apparent that not every El Niņo engendered above-normal rainfall, as might have been expected. This was true only in about three-quarters of the cases; hence, the link between rainfall and El Niņo is this region has a degree of uncertainty, i.e. it is probabilistic. Interestingly, almost every time it was cold in the tropical Pacific, i.e. a La Niņa, less rainfall than average was observed. This link seems to be more reliable than the El Niņo one. The most recent La Niņa was during the fall and winter of 1995/96. At that time, the Southwest suffered one of its stronger droughts on record (Figure 3). Not all of the above- or below-normal rainfall years were related to El Niņo or La Niņa. We suspect these are related to the mechanisms discussed in Figures 1 and 2 that happened either on weekly or decadal time scales.

2. Current Status of El Niņo and Official Forecasts for the Next Several Seasons

Much has already appeared in the news media about the appearance of the current El Niņo. Based solely on what has been observed, this event even now (in terms of the size of the departures from normal SSTs) is as large as the 1972/73 event was during its peak, the second largest event of the past 50 years. Typically such events peak in the late fall/early winter time frame. However, all of the CPC/NCEP objective forecast models indicate that this event will actually not reach peak strength until well into the winter before starting to lose amplitude. When tested against events during the past 15 years, each of these forecasts have proven to be independently quite accurate out to about six months. This is particularly true when forecasts are initiated in the summer season for winter conditions. Since all of the forecast models agree so closely, CPC confidence in this forecast is especially high. Hence, it is likely that this event will be at least the second largest event in recent history, or perhaps as large as the one that took place in 1982/83, considered the "event of the century," at least through the winter. What happens in the spring of 1998 and later is less certain. The forecasts indicate that the event will decay; however, forecasts nine or more months into the future are less certain and forecasting conditions at that time in any case is difficult.

The impacts of El Niņo on U.S. temperature and rainfall variability are also a strong function of the season. During the summer, several regional signals exist, one of which is mentioned in Section 3. Other discernable impacts of precipitation begin to appear by October, but overall winter and early spring impacts are the most pronounced. The official climate forecasts for the October - December and January - March periods are shown in Figure 4a & Figure 4b, respectively. The tools that CPC uses attempt to take into account other effects in addition to El Niņo; however, the forecasts for the next 9 months were primarily based on the historical El Niņo signal.

The most robust aspects of the U.S. prediction are the above-normal temperatures for the north-central states and the above-normal precipitation for Texas and coastal areas of the Southeast during winter and early spring. Nevertheless, the fall precipitation patterns historically linked to El Niņo are also quite reliable in some regions, including the Southwest (Section 4). A more detailed discussion of the developing precipitation deficit in the Southwest and what is expected to happen to this deficit follows.

3. History of the 1997 Southwest Drought

During the late winter and early spring of 1997 much of the southwestern United States experienced below normal precipitation (Figure 5a). During the 11 months through April 1997 parts of southeastern California, southern Nevada and western Arizona received only 28%-65% of normal precipitation (not shown). Below normal precipitation has persisted in this region up to the present time (Figure 5b). As discussed above, strong warm episode conditions have developed in the tropical Pacific during the spring and early summer. Typically, such strong warm episode conditions are associated with a delay in summer monsoon precipitation in the southwestern United States, which usually begins in early July (Figure 6).

4. Outlook for Southwest Precipitation

Given the high level of confidence in the El Niņo forecast for the next 6 to 8 months, it is likely that typical warm-episode precipitation anomalies should be observed in the Southwest during this period. Generally, El Niņo favors a return to near-normal precipitation by late summer in those regions (AZ & NM) that normally experience monsoonal precipitation (Figure 7). This rainfall near the end of the monsoon season may not alleviate the precipitation deficits in Arizona.

Southern California does not benefit from the summer monsoon rainfall, so it is expected to remain dry through early fall. By late fall, however, an enhanced North Pacific jet stream should develop and extend farther east than normal, favoring above-median precipitation throughout the Southwest (Figure 8a) and (Figure 8b). Note in particular that 10 out of 11 El Niņo episodes coincided with October - December precipitation among the wettest one-third of the historical distribution in Arizona. Thus, precipitation deficits should decline as we progress though the late fall and early winter. Confidence that wetter-than-usual conditions will continue in the Southwest remains high through early 1998 (Figure 9a) and (Figure 9b).

The Climate Prediction Center (CPC) provides climate products and services consisting of operational prediction of climate variations, monitoring of the climate system and development of databases for determining current global and regional climate anomalies and trends, and analysis of their origins and linkages to the complete climate system. These services cover climate time-scales ranging from weeks to seasons, extending into the future as far as technically feasible, and over the domain of land, ocean, and atmosphere, extending into the stratosphere.

Services and products are distributed to users in the government, the research community, private industry, and the public, both in this country and abroad. The CPC supports and stimulates the use of climate data for applications in agriculture, energy, transportation, water resources, and health.

To support these services, the CPC engages in diagnostic research and studies of observations and forecast model output to improve the prediction, monitoring, and analysis of the physical climate system.

This publication monitors current and developing climate variations and is issued as conditions warrant. Questions should be directed as follows:

     Area of Concern	               Source	                          Telephone Number

Current Climate Information  Climate Prediction Center, NWS
	                     Climate Operations Branch	                    (301) 763-4670
	                     Analysis Branch	                            (301) 763-8227
                             Prediction Branch	                            (301) 763-8155
Regional Climate Centers     Northeast RCC                                  (607) 255-5950
                             Southeast RCC                                  (803) 737-0800
                             Midwest RCC                                    (217) 244-1488
                             Southern RCC                                   (504) 388-6184
                             High Plains RCC                                (402) 472-6706
                             Western RCC                                    (702) 677-3103
Public Information	     NWS Public Affairs	                            (301) 713-0622
Historical Perspective	     NESDIS National Climatic Data Center           (704) 271-4800
Water Supply/Forecasts	     NWS Office of Hydrology*	                    (301) 713-1630
Wildfires	             National Interagency Coordination Center       (208) 387-5512
Agriculture	             USDA World Agricultural Outlook Board          (202) 720-3508
	                     USDA National Agricultural Statistics Service  (202) 720-2157
	                     Joint Agricultural Weather Facility	    (202) 720-7917
Drought Planning             National Drought Mitigation Center             (402) 472-6707

*There are 13 National Weather Service River Forecast Offices throughout the U.S. that provide water supply forecasts. Call the above number for the River Forecast Center with responsibility for your area of interest.