Summary

Background for This Climate Outlook

The forecast is expressed in a probabilistic manner, in terms of the estimated probabilities that the mean temperature or the total precipitation for the two 3-month periods will be in each of the following three categories: below normal, near normal or above normal. The normal is based on the observations over a recent 30-year period (1971-2000), and the three categories are defined such that the observed temperature or precipitation over the 30-year period has been in each category in one-third of the cases (i.e., 10 years out of the 30). Forecasts for three-month periods are used to give a broad view of the climate averaged over the given months. Individual weather events within the season may sometimes act in the opposite direction from the climate forecast, while the average of the weather over the 3 months may still be in agreement with the forecast. The forecasts have greater accuracy when averaged over a long period such as 3 months than when they are for shorter periods, because the individual weather events are not predictable. The longer the period, the less likely the average will be affected by a small number of strong but unpredictable individual weather events. The factor causing there to be some predictability in a longer-term average is a weak but consistent tendency caused by departures from normal of sea surface temperature in various locations, such as in the tropical Pacific, tropical Atlantic, and Indian Ocean. 

The severe drought that occurred over much of CSWA during the three-year period of 1998-2001 is believed to have been associated with the very large-scale pattern of the sea surface temperature (SST) that occurred in the tropical Pacific and Indian Oceans during that period. In each of the three winters (1998-99, 1999-2000 and 2000-01), and to a lesser degree during the subsequent winter (2001-02), this SST pattern was characterized by below normal SSTs in the central and east-central tropical Pacific, from about 165E longitude to at least 140W. The pattern has also included above normal SST in the western Pacific extending from 160E to Indonesia, and additionally including portions of the tropical eastern Indian Ocean. Much of this SST pattern, particularly that across the Pacific, has been observed many times previously and is associated with the cold phase of the El Niņo/Southern Oscillation ENSO (i.e., La Niņa). In fact, the 1998-99 winter had a moderately strong La Niņa, the 1999-2000 winter had a weaker La Niņa and the 2000-01 winter had a cool-normal or a borderline La Niņa, depending on one's definition of La Niņa. What was unique about the SST pattern during the three most severe drought winters of 1998-2001 was in the Indian Ocean portion. Usually, much of the Indian Ocean tends to follow what happens in the central and eastern tropical Pacific, lagging it by a few months: During El Niņo it becomes warm by about October and during La Niņa it becomes cool around that time, since both El Niņos and La Niņas often first develop during the period of April to July. During the three winters of 1998-99, 1999-2000 and 2000-01, much of the Indian Ocean followed the western Pacific (i.e. it was warm) rather than the central and eastern Pacific. That is, it went in the opposite direction from what was expected. The warmth of parts of the Indian Ocean that usually become cool during La Niņa episodes may reflect global climate change, since the Indian Ocean has shown a tendency to warm gradually over the last two to three decades. If this is the case, there is a risk that this pattern could develop with greater likelihood in the future than it has in the past. However, scientists do not yet know enough to be able to state conclusively that this particular climate change is in fact occurring. 

There has historically been a weak tendency for wintertime precipitation in CSWA to be below normal during La Niña episodes and above normal during El Niņo episodes. Because the tendency is weak, the relationship has been observed only in some of the cases, and exceptions have occurred. However, what is noteworthy is that never during the last 50 years have several years of drought been as pronounced as observed during the 1998-2001 period, even with previous consecutive La Niña years such as those of the middle 1970s or the weaker series of La Niña years in the middle 1980s. The observational record indicates that the drought of the 1998-2001 winters is unprecedented in recent history. 

Based on a study of observational data (Barlow et al. 2002; Tippett et al. 2003, 2005), and on the behavior of dynamical models in forecasting winter precipitation in CSWA during the last several decades, the following two factors appear to be critical in causing CSWA to have dry winters: 

1) The SST needs to be below normal in the tropical Pacific in the vicinity of the international date line, in the region of 170W to 165E. 

2) The SST needs to be well above normal in the far western tropical Pacific and far eastern tropical Indian Ocean, in the region of 115-155E. (Note that the SST in this region does not have large year-to-year variation, so only 0.6 deg C above normal can be considered well above normal.) 

The strength and areal extent of the warmer-than-normal SST described in the second factor seems to be the ingredient present during 1998-2001 that was not found to as great an extent in the La Niņa episodes of earlier years. Whether this stronger warm pool in the western Pacific is a reflection of global climate change (warming) remains to be confirmed. 

During winter of 2001-02, La Niņa conditions were largely absent, and the warmth of the SST in the western Pacific was weaker. While precipitation in CSWA remained deficient during the months of December, January and February, it began to improve in March and in some locations it actually became above normal in late March and during most of April. Thus, the drought that dominated the previous three winters continued through much of the fourth winter before relief arrived near the end of the winter. The total precipitation during the winter and spring seasons of 2001-02, while still below normal, was less severely below than for the previous three winters. 

During winter of 2002-03, there was a moderate El Niño in the Pacific, and the SSTs in the far western Pacific and eastern Indian Ocean were less warm than they had been during the previous four winters (although they still were not below normal as would be expected during an El Niņo). Observed climate featured greater than normal precipitation in the far northern portions of the CSWA region, climatological probabilities in the center of the central and north-central parts, and for a tendency for below normal precipitation in the south-central and southern portions.

During winter of 2003-04, ENSO conditions were neutral, but leaning very slightly toward El Niño. Climate observations showed a pattern quite similar to those of winter of 2002-03: above normal precipitation in only the northern portions of the CSWA region just north of Pakistan and Afghanistan, approximately climatological amounts in far northern parts of those countries, and below normal precipitation in the central and southern portions.

During winter of 2004-05, ENSO conditions were that of a weak El Niņo, but one that was centered near the dateline rather than having a more traditional pattern of warmer than normal SSTs extended much farther east. The precipitation observations were mixed and averaged near normal over the area as a whole. There was a tendency for above normal precipitation in extreme northern Afghanistan, Pakistan and in immediately adjacent parts of northern India and western China. Some above normal precipitation was also unexpectedly observed in a small part of southern Afghanistan. Slightly below normal precipitation occurred over central Afghanistan.

Forecasts for Winter 2005-06: Nov- Dec-Jan 2005-06 and Feb-Mar-Apr 2006

Back  Top