Professor, University of Texas at Austin
Outstanding displays of wildflowers are typical after an El Niño winter and spring in Central Texas
This article discusses the long-term rainfall patterns at the Fly Gap Division of the Double Helix Ranch, and the relationship between El Niño cycles and the droughts and wet periods common to this area of central Texas.
This article addresses questions such as: What caused the recent drought of 2005-2006? Why did it end? How severe was this drought compared to other droughts of record? After a wet year first half of 2007, why did dry conditions return in late 2007? What are the long-term trends for rainfall in central Texas? How will global warming affect our rainfall patterns?
The Fly Gap Division of the Double Helix Ranch is located in central Texas, close to the geographical center of the state. The long-term average rainfall is about 26 inches per year (66 cm). However, an "average" year is quite rare. Instead, the ranch cycles regularly between wet periods (up to about double the average rainfall within a 12-month period) and dry periods (often as little as half the average rainfall). These wet-dry cycles are fairly regular, with "extreme" wet and dry periods occurring roughly every five years (see the graph below):
In this graph, each plotted point shows the monthly totals of rainfall for the previous 12 months. Since this 12-month moving sum always includes a full year, the cycles have no relationship to any seasonality (in other words, each 12-month period includes each of the months of January through December).
I have marked some of the major droughts and wet periods over the past 55 years. In particular, notice the severe drought of record in the early 1950s. Also, notice that the wettest wet periods have been in the past two decades, and that the 2005-2006 drought was one of the worst droughts (in both intensity and duration) since the early 1960s. The 1950s drought actually consisted of three droughts of similar intensity in a row, with short periods of near-average precipitation in between each major dry period.
These wet-dry cycles in central Texas are caused largely by the fluctuations in ocean temperatures in the western Pacific Ocean (the El Niño Southern Oscillation). When the western Pacific Ocean temperatures rise, rainfall in central Texas increases (especially in the winter and spring). When these ocean temperatures fall, rainfall in central Texas decreases. Long periods of above-average Pacific Ocean temperatures are called El Niño episodes, and long periods of below average temperatures are called La Niña episodes. Specifically, the Ocean Niño Index (ONI) measures the departures in temperature from the long-term average in an area of the southern Pacific Ocean known as Niño 3.4. El Niño episodes are declared when the ONI is above 0.5 degrees (C) for five or more consecutive overlapping three-month periods, and La Niña episodes are declared when the ONI is below 0.5 degrees for five or more consecutive overlapping three-month periods. (If the ONI is above or below the 0.5 C range but it has not yet been there long enough to declare an El Niño or La Niña episode, then El Niño or La Niña conditions are said to exist.)
There are exceptions to the pattern of wet El Niño years and dry La Niña years, of course. Hurricanes also can have a major influence on our rainfall patterns in the summer and fall, and hurricanes tend to be suppressed in El Niño years and more active in La Niña years. La Niña years tend to be dry in the winter and spring, but may get late summer/fall floods associated with hurricanes. El Niño years are often wet in the winter and spring. The wettest years of all come when we have an El Niño winter and spring, and then the El Niño breaks to allow an active hurricane season in the summer and fall.
The following graph from NOAA shows the El Niño/La Niña cycles since 1950:
If you compare the wet/dry cycles at Fly Gap to the El Niño/La Niña cycles, you will see that there is a close correlation. In fact, if we plot the ONI against the rainfall deviation from the long-term average, on a month-by-month basis for the past 20 years, the correlation is quite clear:
In particular, note that when the ONI is above 1.0, Fly Gap is virtually always wetter than average. When the ONI is below -1.0, Fly Gap is likely to be in drought (although this pattern is somewhat more variable). Some of the variation stems from the fact that El Niño episodes (when the ONI is high) tend to produce wet weather in central Texas from fall through spring, but drier than normal weather in the summers.
Also, notice that for the last 20 years, the ONI and Fly Gap rainfall have tended to be above their long-term averages, consistent with the world-wide patterns of global warming. In fact, global warming is likely responsible for a slow but consistent increase in our local average annual rainfall, as shown in the following two graphs (the first shows a linear trend line, the second a ten-year moving average):
Although the average across all these years is about 26 inches, note that the annual average has shifted upward by almost 8 inches over the past 55 years (from about 22 to 30 inches per year). Even though overall global warming appears to have contributed to this trend of locally increasing rainfall, the same effect has contributed to slightly more moderate temperatures locally. Increased cloud cover reduces high temperatures (shading effect), but also raises low temperatures (insulation effect). Thus, over the past 55 years, Fly Gap has seen a slight decrease in the average high temperatures in summer, and a slight increase in the average low temperatures in winter. The changes are slight but significant:
Will these long-term patterns continue in central Texas with increased global warming? There is considerable debate about that point, with different models showing different outcomes. A moderate degree of ocean warming is likely to increase El Niño events, which tend to make central Texas wetter. However, a major increase in temperatures could cause a shift in the Pacific jet stream, which supplies us with much of our moisture. An increase in temperature would increase evapotranspiration (if average windspeed remains constant), so it would take more rainfall to produce the same amount of soil moisture. However, worldwide increased evaporation also must lead to increased precipitation, since evaporation and precipitation have to be in equilibrium in the long term (otherwise, atmospheric moisture would continue to increase or decrease indefinitely). Some climate models predict increased variation in rainfall for the southwestern United States, which could increase the intensity of both droughts and floods. One recent analysis of the worldwide rainfall patterns over the past twenty years shows a slow but significant increase in global precipitation associated with global warming (How much more rain will global warming bring?; Science 317: 232-235; 2007). Although this is consistent with the effects seen locally in central Texas over the past few decades, the rainfall patterns in any one region could change quickly depending on a large number of climatic complexities. The bottom line is that the long-term predictions are all over the map for central Texas. The one thing that is clear is that global warming will produce major effects on precipitation patterns. Whether that means that central Texas will become wetter, drier, or more variable is not yet entirely clear, however. To date, recent global warming seems to have made central Texas wetter, but a switchpoint in the climate could end that trend suddenly.
The following grasph from NOAA documents the reality of global warming, and how quickly and suddenly it is happening. Note that global temperatures have been consistently above average, and steadily rising, since about 1980. One can see the beginnings of this trend dating back to about 1940:
Getting back to the effects from El Niño/La Niña cycles: The effects of this cycle are particularly evident from fall through spring. If we compare El Niño years to La Niña years, and look at the rainfall from October to March, most El Niño years are wetter than average and most La Niña years are drier than average. In the following graph, El Niño years are shown in blue, and La Niña years in red:
Note that almost all very wet years were El Niño years (the La Niña year of 1984-1985 was the exception), and all the very dry years were La Niña years. Some El Niño years and some La Niña years are about average, but they are in the minority.
When I first wrote this article, in summer 2006, the ONI was climbing out of negative territory it had been in since fall 2005 (when our 2005-2006 drought began). The ONI was predicted to increase, and to keep climbing for the remainder of 2006, suggesting moderate El Niño conditions developing in late 2006 and early 2007, and therefore forecasting an end to the severe 2005-2006 drought. This happened exactly as predicted. We had a wet spring and summer in 2007, which pulled us out of the 2005-2006 drought:
In August 2006, over 90% of Texas was in one of the drought categories, with 73% of the state in severe, extreme, or exceptional drought. After the rains in the first half of 2007, the entire state was free of drought, with many areas experiencing record rainfall. This map from the National Weather Service shows the rainfall (as a percent of normal) for 1 January to 21 August 2007:
All this rain stimulated substantial vegetative growth, as seen in this NASA map of vegetative growth in June of 2007:
Most of Texas showed far more vegetative growth than usual in the spring and summer of 2007, and the areas in eastern Texas that are not green on this map were under heavy cloud cover when the satellite data for this image were collected.
During the latter half of 2007, we returned to La Niña conditions, as seen in this summary of sea surface temperatures from the Pacific presented by NOAA (in particular, notice the return to negative deviations in the critical El Niño 3.4 region):
This return to La Niña conditions resulted in a return to our dry weather pattern for the latter part of 2007 and beginning of 2008. From September 2007 to March 2008, we were significantly below average rainfall in central Texas. Note, however, that the negative sea-surface temperature anomolies have lessened recently, and positive deviations in sea-surface temperatures have returned in the eastern Pacific (the 1+2 region in the graph above). Forecast models predict that this weakening of the La Niña pattern will continue into the summer of 2008, with a return to near-neutral conditions likely:
The black line shows the observed departures of sea-surface temperatures in the El Niño 3.4 region. The heavy blue line is the average of the various forcast models (shown by thin lines). Thus, the current La Niña episode is likely to end in 2008. There is now about an equal chance of wetter than average or drier than average conditions for the second half of 2008.