Climate variability on multiple time scales: Monsoon bursts and El Nino clash over SE Asia


Matthew C. Wheeler

Invited presentation for an international science roundtable for the media, a side-event to the PrepCom4 meeting for the World Summit on Sustainable Development held in Bali, Indonesia. June 2002.

Power Point slide show of presentation



Link for Programme and Media Releases

Background Paper


Background text of presentation

In 1997 Indonesia was gripped by a severe drought exacerbating forest fires burning across the country. This drought was caused by the extreme El Nino taking place in the Pacific; ocean surface temperatures were much warmer than normal in the east and cooler than usual near Indonesia, shifting rain away from Indonesia. By the time the drought-induced forest fires were finally out in 1998, huge losses had not only occurred for the millions of people affected by the smoke haze, but large areas of natural wildlife habitats and forest resources had been destroyed. Conservative estimates of the economic loss that occurred for Indonesia are over US$4 billion. This chain of events is a clear example of the susceptibility of the natural and human-developed world to climate variability. El Nino is just one source of this climate variability on the interannual time scale, recurring every two to seven years.

Important climate variations also exist on much shorter time scales. February this year saw devastating floods in southern Indonesia arising from a burst in the monsoon. Changes in atmospheric conditions caused the monsoon to undergo phases of "active" and "break" periods, each lasting for a week to a month. During active periods rainfall is greater than normal over a large area, but can hit some local areas more strongly than others, as was dramatically seen around the city of Jakarta in February. At one stage, 20% of the city's surface was covered by muddy water, killing at least 100 people in the area, and leaving thousands homeless for many days. The destruction of crops and food shortages were also felt. Such destruction and losses, like those of the 1997-98 El Nino, are a shocking set-back for a country's development.

Whereas most year-long droughts in Indonesia are linked to El Nino and associated changes in sea-surface temperatures to the east, recent scientific analysis of the shorter time-scale intraseasonal monsoon bursts has revealed links with changes in atmospheric conditions to the west over the Indian Ocean.. The February active monsoon period over southern Indonesia appears to have been linked to what scientists call the Madden-Julian Oscillation (MJO), a large-scale eastward propagating wave in the atmosphere with a period in the range of 30 to 60 days. The February floods were clearly associated with such an oscillation in the weather propagating from the Indian basin across Indonesia and into the western Pacific, a sequence that usually takes around one month for the MJO. As the active phase of the oscillation moves across the area the rain rates increase followed by a shift of the surface winds to strong westerlies.

So, as opposed to El Nino, for which scientists look toward the Pacific for advanced warning, it is the atmosphere over the tropical Indian Ocean that forewarns of an impending MJO, information that can be used for a prediction up to 20 days in advance. Indeed, there was quite a build-up of active storms and rain over the equatorial Indian Ocean in early January such that accurate predictions, using a new technique, were made of the impending rain over Indonesia. Presumably, much could have been done to avert some of the destruction and loss of life had the warning signs from the prediction been readily available and heeded.

Predictions of the longer time scale El Nino, and its opposite cold-phase sister La Nina, are also continuing to be developed and improved. There has not been an El Nino since 1998, but current predictions indicate that one has a good chance of appearing in 2002. The latest (of May 2002) predictions from Australian climate scientists suggest the probability of an El Nino occurring this year are at 50%, double the normal chance, but still not a certainty. The most likely scenario, scientists agree, is that there will be a slow evolution toward weak El Nino conditions during the remainder of the year. Such a weak El Nino would have considerably weaker global impacts than were experienced during the very strong event of 1997-98, but its effects would still be felt in many parts of the world, including Indonesia. This prediction of a possible 2002 El Nino has remained essentially the same as that suggested at the beginning of the year, a considerably long lead-time for such a forecast. Interestingly, much of the great speculation made at the beginning of the year of a brewing 2002 El Nino was the result of the considerable strength of the shorter time-scale MJO seen during that time. Indeed, the 30- to 60-day MJO has been implicated in the generation of past El Ninos, an example of a complex interaction between different time scales. Thus the MJO event that contributed to the February floods in Indonesia is not unrelated to the possibility of an El Nino and droughts there later this year.

And such natural climate variability as bursts of the monsoon and El Nino can augment or diminish the variations resulting from human induced climate changes. El Nino years, such as 97-98, are globally warmer than non-El Nino years, hence increasing global temperatures above those arising from human induced changes alone. This, along with the examples of El Nino and MJO/monsoon interactions demonstrate that the climate system interacts on all time and space scales and emphasizes the need to study the system as a whole. This is the approach taken by the World Climate Research Programme and its project on Climate Variability and Predictability in particular.

Notwithstanding this new understanding and prospects for improved prediction of these significant weather and climate events, there is much to be gained just from knowledge of their existence. These forms of climate and weather events do exist, and will continue to happen around the globe. Often, the decisions and changes undergone by humans in the process of development have made us more susceptible to such events. In the case of the 1997 drought in Indonesia, better forest management practices could have helped to reduce the impact of the El Nino. In the case of the February floods in Jakarta, improved urban planning for water run-off from such rain events could be employed. Not only is the climate science important, but the partnership of that science with the human dimensions side is crucial. Continued development of many countries will be difficult to sustain without regard to the occurrence of climate variability on all time scales and its impacts.