Charlie A. Liles
National Weather Service Albuquerque, NM

Introduction

Part of the Southern Oscillation Index (SOI) cycle involves cool episodes, known as La Niņa, in which the equatorial Pacific Ocean region becomes cooler than normal.  The causes and effects of the cycle are still not well understood, but the cooling comes about because of strengthening of the east-west circulation known as the Walker circulation.  During La Niņa episodes, the Walker circulation is strong, which leads to strong downward motion over the eastern equatorial Pacific, and strong upward motion and increased convection (and precipitation) over the western equatorial Pacific.  The southern oscillation becomes positive, leading to increased pressure gradients over the Pacific, and enhanced trade winds and upwelling of colder water from beneath the ocean surface.

For the Southwest U.S., there is less moisture available for precipitation production for several reasons.  Just as El Niņo leads to a pooling of mid and high tropospheric moisture over the eastern Pacific that may be tapped for enhanced precipitation from storms, the downward motion over the eastern Pacific associated with La Niņa dries out the atmosphere.  However, another even important reason for less precipitation in the Southwest U.S. during La Niņa is related to changes in flow patterns above the oceans.  During La Niņa, mid and high level tropospheric ridges develop over the Pacific, and also across the southern U.S. The affects of this ridging are especially pronounced during the winter, as they shunt storms to the north and considerably weaken storms farther south that manage to reach the Southwest U.S.

Methodology

Just as with El Niņo, various authors identify different periods associated with La Niņa.  Most authors agree on the major, main events, but not on the weaker ones.  Most researchers use a combination of the SOI and sea surface temperatures to define episodes.  Some research has produced convincing evidence of relationships between La Niņa and weather in the United States.

Since reliable pressures are available for a longer period of time, I utilized only the Southern Oscillation Index to identify cold episodes.  In general, the SOI is available for the period 1866 into 1998, although data before 1920 had to be estimated in some cases using points other than Papeete, Tahiti, and Darwin, Australia.  Cold episodes were defined using the following methodology: .

Once a period of at least five months was flagged as being a cold episode, the episode was said to begin at the first month of the five.  For events in progress, once the five month running SOI was less than 2.5, the episode was said to end at the beginning of the first month in which the SOI became negative.   Finally, cold episodes that did not persist for a minimum of two complete seasons were thrown out.  Episodes that were initially identified as contenders and then thrown out included: March-July 1902, July-November 1908, May-October 1915, and March-July 1931.  The first two of these were initially identified as "weak" events, while the second two were in the moderate category. Seasons were considered to consist of the following time periods: Winter (December-February), Spring (March-May), Summer (June-August), and Autumn (September-November).

Once cold periods were identified, seasonal precipitation at Albuquerque during cold events was compared to the long-term seasonal precipitation averages (1892-1997) to determine what relationships existed between La Niņa and Albuquerque precipitation.

La Niņa Periods 1892-1997

Seasonal Precipitation During La Niņa Events

Spring

Summer

Autumn

Winter

Annual (Calendar Year)

Discussion of Results

Of the 14 springs during cold episodes, Albuquerque's precipitation was below normal during all but one of those seasons.  The exception was 1918, when 1.80 inches of precipitation was measured.  This was actually during perhaps the strongest La Niņa of the century, but it should be noted that that particular cold episode was winding down.  In fact, the SOI fell sharply and even went negative in March of 1918, before rebounding strongly to 2.20 in April.  The SOI then fell below zero in June of that year, and the cold episode was over. The driest spring of the cold episodes was 1956 (0.18 inches of precipitation), which was in the middle of an extended La Niņa.

Consequently, it's impressive that 13 of 14 cold-episode springs exhibited below normal precipitation, with an average precipitation amount only a little over half of the normal "non-La Niņa" amount.

Cold episodes produced 5 summers with above normal precipitation, 9 with below normal, and one that was basically normal. Average precipitation during cold episodes was 2.97 inches, below the 106 year average of 3.42 inches and the non-La Niņa average of 3.50 inches.

The three wettest cold episode summers (1910, 1916, 1973) were all during strong La Niņa events.  This makes it clear that we can't count on cold episodes, no matter how strong, to produce below normal precipitation during Albuquerque's summer. Still, the fact that precipitation was below normal during 9 of 15 summers suggests cold episodes tend to produce dry summers in Albuquerque.

Of the 20 cold episode autumns, precipitation was below normal 13 years.  For all 20 seasons, the La Niņa average of 1.72 inches was below the non La Niņa average of 2.37 inches and the 106 year average of 2.24 inches.  The driest autumns were during the 1956 strong episode (0.39 inches), and the 1924 moderate episode (0.43 inches).  Autumn precipitation exceeded three inches during the weak episode of 1893, the moderate episodes of 1938 and 1971, and the strong episodes of 1916 and 1988.   During the 1988 episode when much of the middle of the U.S. was experiencing moderate to severe drought, western New Mexico continued to receive abundant moisture.

Consequently, conclusions for Albuquerque's autumn precipitation related to La Niņa are similar to those of summer.  On the average, cold episodes produce below normal precipitation during autumn, and the precipitation is below normal about two thirds of the seasons.  However, there is still a lot of variability in the results for autumn, making definitive forecasts based on La Niņa alone not such a good idea.

Of the 19 winters during cold episodes this century, precipitation was below normal during 15 of those seasons.  In a previous paper on La Niņa (by the author) and Albuquerque precipitation, it was found that winter data exhibited the smallest standard deviation.  Without even looking at standard deviations, a glance at the winter precipitation distribution from the tables above suggests that variability was certainly less than for summer and autumn.  However, on average, precipitation was only about 25 to 30 percent below normal for what is normally the driest season in Albuquerque anyway.  The winter of 1910-11 was somewhat of an exception, and precipitation was about 200 percent of normal.  It should be noted that the 1910-11 cold episode was waning rapidly during that winter, and the SOI had returned almost to zero.

By summing the seasonal precipitation averages for the four seasons during cold episodes, one comes up with a total of 6.72 inches.  This value is 21 percent below the 106 year average of 8.50 inches.  If one looks at the annual precipitation for years in which a cold episode was in progress for the entire year, we are limited to the following years: 1910, 1917, 1950, 1956, and 1971.  Precipitation for those years, respectively, was: 7.41, 3.29, 4.10, 4.06, and 8.05 inches.   Consequently, we see that the average precipitation during a year in which a cold episode was in progress for the entire year was only 5.38 inches!  This is only 63 percent of the 106 year average, and 62 percent of the non-La Niņa average.   Additionally, we can note that three of those years (1917, 1950, 1956) were among the driest 4 years in Albuquerque's written records that date back to 1892.  The year 1917 was the driest year on record, while 1956 was 2nd driest, and 1950 was fourth.

Conclusions

There appear to be some fairly strong signals between cold (La Niņa) episodes and Albuquerque precipitation.  Seasonally, there is a tendency to receive below normal precipitation during these cold events, with the strongest signals relating to the spring season.  During prolonged cold events in which the event spans the length of a year or more, the data strongly suggest Albuquerque will experience below normal precipitation.  During the five years of the 20th century in which cold episodes have persisted through an entire calendar year, precipitation was below normal during all five events.  Furthermore, three of these five years produced totals that are recorded among the driest four years of the century.  The average precipitation during these "La Niņa years," was a mere 5.38 inches, or 63 percent of the long-term (106 year) average of 8.50 inches.

Even though I categorized cold episode (La Niņa) strength in this paper according to a five month running total of SOI magnitudes, seasonal precipitation distribution did not appear to be correlated with these magnitudes.   Instead, it appears that the worst consequences of dry periods in Albuquerque associated with cold episodes may be more related to longer duration of the events.  However, it should be noted that the events that persisted the longest were categorized "strong" events, with the exception of the 1971 episode, which rated near the top of my "moderate" category.