Where El Niño Makes It Rain And Where It Doesn’t

The 2015 El Niño continues to grow in strength and California is once more suffering flash floods and mudslides. It’s widely recognized that California gets wetter when a strong El Niño is running in the Equatorial Pacific, and the plot of the San Diego rainfall record against the Niño3.4 Index shown in Figure 1 confirms that this is indeed the case:

Figure 1: San Diego GHCNv2 rainfall record vs. the Niño3.4 Index. The data are 12-month running averages of 1870-2013 monthly means downloaded from KNMI Climate Explorer. I define a 12-month period when the Niño3.4 Index averages 1.0˚C or more or minus 1.0˚C or less as a “strong” El Niño or La Niña. The usual threshold for defining an El Niño/La Niña event is five consecutive months with the Niño3.4 Index over 0.5˚C/under minus 0.5˚C.

But what of the rest of the world? Does El Niño affect rainfall everywhere or just in a few places? And what does its widely ignored twin sister La Niña do? I looked into this question a few years ago, and here is a brief summary of the results.

First some more rainfall records. According to GHCN v2 data annual rainfall at Guayaquil, Ecuador, increases more than anywhere else on Earth during El Niños, and with a high proportion of unusually wet years:

Figure 2: Guayaquil, Ecuador GHCNv2 rainfall record vs. the Niño3.4 Index. 

But there are stations where rainfall decreases during El Niños and increases during La Niñas, such as Fortaleza, Brazil:

Figure 3: Fortaleza, Brazil GHCNv2 rainfall record vs. the Niño3.4 Index. 

And some where rainfall decreases during both El Niños and La Niñas, such as Tahiti. (Other stations that show this convex-up shape, including Honolulu, Christchurch and Ouagadougou, but there are none that show a well-defined concave-up shape):

Figure 4: Tahiti GHCNv2 rainfall record vs. the Niño3.4 Index. 

And others that show no significant changes, such as Manchester, Seoul and New York (shown below):

Figure 5: New York GHCNv2 rainfall record vs. the Niño3.4 Index. 

To summarize the impacts of El Niños on global rainfall I plotted the percent change in annual rainfall during strong El Niño events relative to average annual rainfall at 27 stations roughly evenly spaced across the Earth’s surface. The results are shown in Figure 6. Notable features are:

*Rainfall during strong El Niños increases by more than ten percent at Cape Town and Choilbasan in Mongolia, but only Guayaquil and San Diego record large increases.

*Rainfall decreases across most of the rest of the Pacific and Pacific Rim area, from Nome, Alaska to Tahiti and from Arequipa, Peru to Townsville, Australia. Significant decreases also occur at Fortaleza, Brazil and Jakarta.

*Rainfall impacts over the rest of the world are minimal. (Changes of less than 10% are judged to be within the limits of estimation error).

Figure 6: Percent change in annual rainfall relative to average annual rainfall when a strong El Niño is running , 27 stations

Figure 7 shows the same plot for strong La Niñas. The impacts are comparable to the impacts of strong El Niños. Townsville, Fortaleza and in particular Arequipa all experience rainfall increases which with the exception of Guayaquil are as large as any experienced during El Niño episodes.

Figure 7: Percent change in annual rainfall relative to average annual rainfall when a strong La Niña is running , 27 stations

A feature of El Niños is that they frequently transition directly into La Niñas, meaning that large variations in rainfall can occur in short periods of time. Figure 8, which plots the percentage differences between strong El Niño and strong La Niña average rainfall, shows the range of total variation at each station. This range provides a measure of the combined impact of El Niños and La Niñas, or if you like the overall strength of the Niño/Niña rainfall effect:

Figure 8: Percent change in annual rainfall between strong El Niño and strong La Niña conditions, 27 stations

Combined Niño/Niña effects are strongest in Northern South America and Southern California and appreciable in Northern Australia, Indonesia and Mongolia, but weak to negligible elsewhere.

These results are not totally definitive because a high-rainfall year does not necessarily guarantee floods and a low-rainfall year does not necessarily guarantee no floods. Nevertheless we can conclude that El Niños and La Niñas have a regional rather than a global impact on rainfall. But the world sees the hand of El Niño in extreme precipitation events everywhere. El Niño has already taken the blame for floods (and droughts) in areas where it has minimal impact on rainfall, as claimed in this 2002 BBC article, and it’s now being blamed for disasters that have yet to happen, such as the coming devastating floods in Africa and the coming snowy winter in UK . Curiously, however, its twin sister La Niña rarely takes the blame for anything.

This entry was posted in Climate change and tagged , , . Bookmark the permalink.

11 Responses to Where El Niño Makes It Rain And Where It Doesn’t

  1. Dave Rutledge says:

    Hi Roger,

    Thanks for a great post. I would like to show my students a couple of the graphs on Tuesday.


  2. Good and interesting topic Roger.

    One thing that might be entertaining to speculate about is whether or not projected climate change would look anything like an El Niño or a La Niña. After googling up a storm on this, I sort of conclude it would not be the same, because El Niño/La Niña is a repeating cycling of heat exchange between the ocean and the atmosphere, and it is “temperature neutral” – that is to say, it doesn’t heat or cool the earth, but simply redistributes heat around the globe, kind of like turning an electric fan on and off inside a house without opening the windows. Sorry if that sounds simplistic.

    Dimly related to this topic, I think I’ve mentioned before that I live in Taiwan, and we are about to get hit by approaching Typhoon Koppu which has been battering the Philippines for the past few days:


    Actually, we can use the rain, so I’m sort of looking forward to it.


  3. Euan Mearns says:

    For those not familiar with El Nino, it is a quasi-periodic phenomenon of the equatorial Pacific manifest as a band of warmer than normal water stretching westwards from the equatorial coast of S America through the Galapagos Islands. It forms in response to a weakening of the west blowing Trade Winds. The UK Met office provides a good description with a short vid.


  4. Javier says:

    As we say in my country, “It never rains to everyone’s taste”. Also the plurals of “El Niño” and “La Niña” are “Los Niños” and “Las Niñas”.

    Thanks for the interesting post.

    Aparently, “Los Niños” are a feature of the Neoglacial period of the Holocene after the Mid-Holocene Transition. See for example: Cabarcos et al. 2014:
    Figure 5e. The world was previously dominated by “Las niñas”.

    It seems that the colder the Earth gets, the more “Los Niños” take place.

    • the plurals of “El Niño” and “La Niña” are “Los Niños” and “Las Niñas”.

      “Spanglish” es un idioma reconocido aquí en México 😉

      It seems that the colder the Earth gets, the more “Los Niños” take place.

      I think it might be the other way round.

      • Javier says:

        I think it might be the other way round.

        That’s what I thought too, but doesn’t seem so. The Earth has been cooling progressively from about 7000 yr BP until 50 yr BP (to leave current warming out of the discussion) following the decay in the obliquity cycle. At the Mid-Holocene Transition (6000 to 4500 yr BP) the ITCZ migrated South towards its current position due to the progressive reduction in northern summer insolation associated to the precession cycle, This change of ITCZ position made possible a higher frequency of El Niño.

        In essence El Niño is oceanic subsurface heat of solar irradiation origin that comes out to the sea surface (El Niño index is defined by changes in SST) and warms the atmosphere on its way to the space. Although to us, atmosphere critters, Los Niños appear as global warming, for the planet they represent a net loss of energy that otherwise would remain longer in the system. The more Los Niños, the faster the planet cools.

        As the planet is cooling following the obliquity redistribution of insolation as its ultimate cause, this suggests that the increase in Los Niños responds to the cooling, and not the other way around. In my opinion, the temperature gradient between the tropics and the poles, that constitutes the ultimate engine for the energy transfer within the Earth system, is increasing as the Holocene nears its end and higher latitudes cool down more than the tropics. The Los Niños then would constitute like a release valve for excessive temperature gradient. A way to reduce it without having to take that heat to the polar regions.

        In terms of global warming an El Niño appears as more global warming as it gets reflected in increased lower atmospheric temperatures. However after the El Niño there is less heat in the system, so the global warming appears to stall or even reverse. This has confused some people like Bob Tinsdale into thinking that Los Niños are the cause of the global warming, while in reality they are just the opposite, a fast track towards less warming (or more cooling).

        This is my current view 😉

        • Javier:

          This has confused some people like Bob Tisdale into thinking that Los Niños are the cause of the global warming, while in reality they are just the opposite, a fast track towards less warming (or more cooling).

          Well, I guess it confused Roger Andrews into thinking more or less the same thing:


          But feel free to poke holes in my argument.

          • Javier says:

            Oops. I am sorry, Roger. Had I known it was also your theory I would have not posted my opinion so bluntly. It is impolite. I apologize.

            But since you ask, I took a look at your article above. After reading it I did a bit of checking, since your analysis is for 4 solar cycles, and we have data for at least 10.5 cycles. This is the result:


            First let me say, as you know but others may not, that it is obvious that the heat that is transferred during an El Niño has a solar origin, but solar variability is thought to be too small, just 0.1% change in TSI during a solar cycle. UV change is somewhat higher, about 7%, but only a small part of the solar energy comes in the UV spectrum. So we have the usual problem with the sun variability – climate variability connection, that we lack a mechanism. We must rely then on strong correlations.

            The correlation that you present is not strong at all. If a case has to be made that Los Niños or Las Niñas fall preferentially at a certain phase of the solar cycle, the numbers do not support it. You certainly have not proved it. Los niños and Las Niñas fall all over the cycle and you will find a lot of difficulty in demonstrating a significant deviation from a random distribution.

            Now let’s use a little bit of reasoning. The energy that drives an El Niño is the heat from warm subsurface waters. How does that heat get there in the first place? The Sun heats surface waters in the tropical areas. The heat that is not radiated back finds its way to the subsurface through two mechanisms: vertical mixing by wind and tides, and the sinking of oceanic currents due to increased salinity from evaporation.

            Eventually that heat will come back to the surface and will be radiated to space over the course of decades or centuries, usually at the polar regions, as polar regions have an energy deficit and radiate more energy to space that what they receive from the Sun.

            El Niño does not generate heat and therefore cannot warm the world. It just redistributes the heat in a temporally and spatially localized way. During and after an El Niño, the Earth is radiating more heat to space, as the satellite measurements clearly indicate, while it is not taking in more energy from the Sun. The energy lost to space through an El Niño, will create a heat deficit for decades or centuries, but meanwhile it appears as warming to us because the heat has to go through the atmosphere and creates all sort of side effects.

            Los Niños are a byproduct of a cooling world. This is demonstrated by the evidence that shows that during the Holocene Climatic Optimum there were no Los Niños, and the climate was dominated by Las Niñas. The world was warmer then than now, yet no Los Niños.

            This theory is consistent with the Holocene paleoclimate record and present situation.

            Even though we are living through a warming world, yearly polar insolation continues to reduce every year, as the obliquity cycle advances. The world is getting into high gear to get rid of all the excess warming that is creating a huge polar to tropical gradient. The overabundance of Los Niños is one of the responses of the planet to global warming. We have gotten 6 El Niños over the last 12 years, one every two years (I think it is an Holocene record, or as some would say, unprecedented), yet the world has not warmed significantly. The theory that Los Niños produce global warming is on shaky ground. Los Niños are a significant factor in the world not warming.

  5. Javier says:

    It is a pity that images cannot be displayed. Perhaps Euan or Roger could display it for me:

    From C.M. Moy et al., 2002. Variability of El Niño/Southern Oscillation activity at millennial timescales during the Holocene epoch. Nature 420, 162-165:

    We observe that Bond events tend to occur during periods of low ENSO activity immediately following a period of high ENSO activity, which suggests that some link may exist between the two systems.

    That link cannot be that high ENSO activity produces warming. Then one would expect high ENSO activity after Bond events, when most of the warming is produced. It cannot be solar activity either, as there is no increase in solar activity immediately preceding Bond events.

    Without a need for understanding how El Niño takes place, and knowing that ENSO activity has been on the raise during the 20th century, and knowing that it continuous on the raise in the 21st, if past is prologue, I would expect ENSO activity to continue increasing, regardless of what temperatures do, until the next Bond event (i.e. abrupt cooling) comes around.

    These evidence is also consistent with the hypothesis that high ENSO activity produces delayed and accumulated cooling by robbing energy to the planet and throwing it out to space, and that high ENSO activity takes place when the difference between what the global temperature is and what it should be (as determined by the obliquity cycle) is maximal.

    If allowed to speculate, by increasing atmospheric CO2 we are stealing energy that should have gone out to space and warming the planet. But the planet is not going to let us get away with it for too long. It has cooling mechanisms, like increased ENSO activity, and it is activating them.

  6. Pingback: Where El Niño makes it rain and where it doesn’t | Climate Collections

Comments are closed.