Contents - Previous - Next

This is the old United Nations University website. Visit the new site at

Impacts of ENSO

Impacts on Soil, Vegetation and Wildlife

In Australia, there is clear evidence that much of the natural vegetation is adapted to the high variability and temporal 'rhythms' associated with ENSO (Nicholls, 1991). Many plants in areas strongly affected by the phenomenon are well adapted to a regime of frequent droughts and periods of heavy rains of, typically, 12 months duration. In these areas, the El Niņo-associated variability dominates the annual cycle. The relative strengths of the annual cycle of rainfall and the ENSO 'cycle' will obviously be different in the Tropics, so the vegetation there may be less clearly adapted. The effects of the phenomenon, however, might still be seen. Massive tree deaths. attributed to the drought, were reported from Borneo during the 1877 El Niņo event (Allen, Brookfield and Byron, 1989).

The high rainfall variability associated with ENSO, and the biennial cycle induced by the phenomenon, could increase the likelihood of fires, due to both natural and human causes. The long, wet periods followed rapidly by long, dry periods would enhance the likelihood of wildfires. Rapid growth during the pluvial season would dry quickly in the ensuing drought and then 'dry' thunderstorms could ignite the forests. Burning-off at inappropriate times, that is, during droughts, could also lead to wildfires. The major El Niņo of 1982-3 led to enormous fires in the Equatorial rain forests of Borneo (Malingreau, 1987).

Extremes in winds and highly variable rainfall caused by ENSO can also lead to severe soil erosion. The biennial nature of the phenomenon, again, can aggravate erosion further. Thus, heavy rains in a La Nina, following an El Niņo drought, would fall on bare soils, with less than normal interception by vegetation. In the New Guinea highlands, it appears that the biennial nature of ENSO leads to above average rainfall before and after El Niņo-related droughts (Allen, Brookfield and Byron,1989), as is the case in Australia (Nicholls, 1991).

Therefore, a similar phenomenon throughout Indonesia and other parts of South-East Asia affected by ENSO can be expected. In these areas, the potential for soil erosion might be high for the reasons outlined above, especially if forests are removed to allow expanded areas of cropping or grazing. Stronger-than-normal south-east winds through the dry season during an El Niņo could also lead to soil erosion, although wind erosion would be less of a problem in tropical South-East Asia compared with the semi-arid parts of Australia, for instance.

There is clear evidence from Australia, and elsewhere, that the climatic and ocean anomalies associated with ENSO affect wildlife. Limpus and Nicholls (1988) found that the number of green turtles nesting each year could be predicted by monitoring ENSO. Green turtles nest in various parts of South-East Asia, and in some areas, they are used for food. If their breeding elsewhere in the region is triggered by the phenomenon, the management of the turtle population might be facilitated by monitoring with the SOI.

Impacts on Agriculture

Malingreau (1987) showed that Indonesian crops are adversely affected by El Niņorelated droughts, such as those of 1972, 1976 and 1982. The severe drought of 1982-3 impaired the dry-season crop of 1982 and delayed the planting of the 1982-3 crop (Malingreau, 1987). At the national level, the main effect of the drought was to postpone the realization of Indonesian self-sufficiency objectives until 1984. Allen, Brookfield and Byron (1989) cite documentary evidence of crop failures and livestock losses throughout Indonesia in previous El Niņo episodes (for example, 1914, 1877 and 1804). The eastern Borneo rice crop failed after the El Niņo drought in 1877-8 and again following floods during the 1879 La Nina.

In the highlands of Papua New Guinea, El Niņo episodes are often accompanied by a series of severe frosts (Allen, Brookfield and Byron, 1989). These damage the aboveground growth of most of the crops, including sweet potato, the single most important food of the highlands. Again, frosts might be expected through the highlands of much of South-East Asia, where cloud cover is reduced during El Niņo episodes. Thus, crops may be vulnerable to frost damage during these periods. Heavy frost and even freak snowfalls in Irian Jaya led to food shortages during 1982 and early 1983 (Allen, Brookfield and Byron, 1989).

The high variability of rainfall and the potential for severe frosts, in areas affected by ENSO throughout South-East Asia, suggest that care needs to be taken in planting appropriate crops. Crops able to thrive under a range of rainfall regimes, from dry to wet, might be more appropriate in these areas rather than crops more suited to consistent tropical climates where ENSO does not have a major impact.

Land-management practices also need to be developed with consideration to the high variability of climate and the temporal pattern of rainfall anomalies imposed by the conditions. Imposition of European land-management practices in parts of Australia where ENSO affects climate has led to long-term changes in vegetation. The best known of these changes is in the area now known as the Pilliga Scrub (Austin and Williams, 1988; Rolls, 1981). Much of this area of 400 000 hectares was open, grassy country with only about eight large trees per hectare when Europeans arrived in the 1830s. Frequent burning by the Aborigines and grazing by indigenous marsupials restricted the opportunities for trees and shrubs to establish themselves. Fire germinated the seeds of the trees and shrubs, but rat kangaroos ate many of the resulting seedlings before they could be established.

The introduction of sheep reduced the number of rat kangaroos by destroying their cover and food. A severe drought during the major El Niņo event of 1877-8 further reduced the number of indigenous marsupials. The following year, a major La Nina event (Wright, 1975), was very wet. The few large trees seeded well and when stock owners burnt to destroy grasses with seeds that got into their sheep's wool, seedlings came up thickly, unhindered by the grasses which would usually compete with them for space. This time, there were no rat kangaroos to eat the seedlings and the trees grew unchecked.

Over the next decade, there were several further periods of establishment, again synchronized with El Niņo-La Niņa oscillations. The European rabbit, also an enthusiastic eater of seedlings, arrived in the area in the late 1880s and prevented further establishment until myxomatosis reduced the rabbit population in 1951. The first successful release of myxomatosis occurred in 1950. Earlier releases of the disease had not led to widespread effects. Extensive rains and flooding in 1950, associated with a major La Nina, contributed to the successful establishment of the disease by providing ideal breeding conditions for the insects that spread it.

In 1917, the Forestry Commission stopped burning in the Pilliga Scrub and by 1950, large amounts of forest litter had accumulated. So had decades of seed production. The forest dried in the El Niņo event of 1951, following good growth during the La Nina of 1950, and a major fire started in November 1951. In the absence of rat kangaroos and rabbits, the new growth induced by the fire had nothing to stop it.

In less than a century, Europeans had unintentionally transformed the area from grazing land into the dense Pilliga Scrub supporting sustained timber harvesting. The ENSO phenomenon played a critical role in this transformation. The biennial nature of it, and of the associated climatic fluctuations, appear to have been especially important in determining these changes. In the establishment of the Pilliga Scrub, for example, the temporal pairing of El Niņo and La Nina events (1877 and 1878; 1950 and 1951) was a major factor, not just extreme rainfalls falling randomly in single years. The very large variability of rainfall caused by the ENSO phenomenon was also important. Heathcote (196:S) notes that there was considerable concern early in the twentieth century that the combination of grazing and the high variability of Australian rainfall was leading to progressive degradation of the native vegetation in semi-arid areas.

Austin and Williams (1988) and McKeon et al. (1990) cite other examples where the intense climatic events associated with both extremes of ENSO resulted in major longterm vegetation degradation. In western Queensland, there was a rapid increase in the sheep population during the above average rainfall years of the early 1890s. Major El Niņo events between 1899 and 1902 resulted in very low rainfall and a rapid drop in animal numbers. Heavy utilization of edible grasses and shrubs during this drought led to a spread of inedible plants and carrying capacities seem to have been permanently impaired (Heathcote, 1965; McKeon et al., 1990). In the subtropical grasslands of southern coastal Queensland, the rapid change in species composition to bunch spear grass appears to have resulted from overgrazing with sheep during the El Niņo-related drought of 1881-2 (McKeon et al., 1990). More recently, low beef prices in the mid1970s led to increased stocking rates in Queensland. These years were wet, the result of the 1973-5 La Nina, but attempts to maintain the high stocking rates into the 1980s with their drier, El Niņo-related conditions has led to pasture degradation, species changes and soil erosion (McKeon et al., 1990).

Similar long-term vegetation changes could be anticipated to occur in the parts of South-East Asia affected by ENSO, unless land-management strategies are designed to take into account the phenomenon. Such major, long-term changes make sustainable development more difficult than might be the case in less variable climates. On the other hand, the predictability of the climatic fluctuations, afforded by ENSO, may allow some of its deleterious effects to be offset by tactical land-management responses. Attempts are under way in Australia to combine climate forecasts prepared by monitoring ENSO with pro-active land-use strategies, both for cropping and grazing. These should lead to higher, more consistent yields and to less land degradation.

Impacts on Human Health

The ENSO phenomenon appears to have considerable potential for affecting human health in various parts of South-East Asia. During the 1982-3 El Niņo, there were reports of isolated outbreaks of cholera throughout Indonesia related to the effect of drought on water supplies. A total of 340 deaths from starvation were also reported (Canby, 1984), although reliable information on this is difficult to obtain. In Irian Jaya, food shortages also apparently led to some deaths (Allen, Brookfield and Byron, 1989). There were, many reports of gastrointestinal diseases due to the shortage of good water in Indonesia during the severe drought related to the 1877 El Niņo (Allen, Brookfield and Byron, 1989). Other nineteenth-century El Niņo episodes also led to food shortages throughout Indonesia. In the highlands of Papua New Guinea, frost damage to food crops during El Niņo periods has led to famines in the past (Allen, Brookfield and Byron, 1989). During La Nina episodes, with their concomitant heavy rains and flooding, an increase in the prevalence of arboviruses (for example, Japanese encephalitis) might be expected. In Australia, such a pattern exists with epidemics of Murray Valley encephalitis occurring only during La Nina events (Nicholls' 1986).

All countries suffer from droughts and floods. It is sometimes suggested that, since not all countries have food shortages as a result, factors other than climatic variability are required to explain why droughts in some areas lead to famines. The higher climatic variability (that is, more severe droughts and floods) in countries affected by ENSO may provide a partial explanation of why droughts in these countries can lead to severe food shortages whereas elsewhere they do not. Simply put, the droughts in such countries will, in general, be more severe than in other areas.

This is not to say that the relatively high climatic variability in low-latitude countries affected by ENSO will necessarily lead to highly visible health impacts. Many observers have noted that drought itself cannot be considered the sole cause of famines. Digby (1901) reported an investigation of the Indian famines of the late-nineteenth century by The Revd J. T. Sunderland published in 1900. These famines occurred during El Niņorelated droughts. Sunderland dismissed the failure of the rains as the cause of the famines. He noted that when drought was severe in some parts, other parts had plenty of rain; that irrigation, which was widespread, provided certainty in cropping; and that transport was good across the country, allowing ready conveyance of food from areas of abundance to areas of scarcity. Sunderland decided that the real cause of the famines was the 'extreme, the abject, the awful, poverty of the Indian people'. This extreme poverty kept most of the population on the very verge of suffering even during years of plenty and prevented them from storing anything to tide them over years of scarcity. The Indian, thus, 'finds starvation invariably staring him in the face if any disorder overtakes that little crop which is the only thing which stands between him and death'.

Sunderland then went on to determine the cause of the poverty. He attributed it to the 'enormous foreign tribute', which 'drains away her wealth in a steady stream that is all the while enriching the English people, and of course correspondingly impoverishing the helpless people of India' (Digby, 1901: 166). Digby agreed that the nineteenth-century famines occurred 'not because rains fail and moisture is denied; always, even in the worst of years, there is water enough poured from the skies on Indian soil to germinate and ripen the grain, but because India is steadily and rapidly growing poorer'.

Glantz ( 1987) reviewed recent studies on the famine-drought nexus in Africa and reached a similar conclusion; namely, that drought cannot be regarded as the major or sole source of the agricultural crisis existing in most African countries. Drought is very often a contributing factor to other underlying problems afflicting societies that are dependent on agricultural production. Garcia (1981: 219) in a study of droughts and famines in 1972-3 (another major El Niņo event) also concluded that his 'case-studies provide confirmatory evidences that droughts . . . are not the sole or even primary cause of internal disequilibrium in the society. They merely reveal a pre-existing disequilibrium.'

The conclusions reached by Garcia and Glantz differ substantially from the view of famines prevailing in the 1960s. At that time, many observers regarded the drought itself as responsible for food shortages and famines in various parts of Africa. Late in the nineteenth century, notwithstanding the analysis by Sunderland, drought was widely regarded as the primary cause of Indian famines. Such views led to technological 'solutions' to famine. Often such technological solutions tend to favour cash-crop production for export which can actually worsen local food availability (Glantz, 1987). The conclusion that social, economic and political problems (what Garcia refers to as a state of 'disequilibrium') lead to vulnerability to drought implies that removal of the source of this disequilibrium is necessary to prevent climate-related health problems.

The influence of social, economic and political structures in determining whether climatic anomalies in a particular region will lead to famine or other health problems can explain why some countries, with a climate strongly affected by ENSO, do not suffer food shortages that lead to famines, while in others (for example, northern China and India), the inexorability with which famine followed drought during the nineteenth and early twentieth centuries has now been broken. The conclusion that non-physical factors may determine whether the possible health problems associated with climatic anomalies will be realized in a specific country or at a specific time does not, however, invalidate the fact that increased climatic variability heightens the potential for health problems. A highly variable climate can provide the frequent severe droughts and floods which set the scene for potential food shortages and even famine. But other factors must operate to realize this potential.

Implications for Sustainable Development

The substantial effect of ENSO on interannual climatic variations in many areas of the world, including at least parts of South-East Asia, suggests that its role needs to be considered in the development of a sustainable environmental future. In this section, a few examples of how ENSO may impact on sustainable development are provided. The list is by no means exhaustive; the widespread effects of ENSO mean that it will impact on sustainable development strategies in many ways, some of which are unpredictable.

The ways in which information about ENSO can be used are categorized here according to whether they represent the development of long-term procedures (strategies) or short-term responses (tactics). Thus, a decision to plant a different crop because a drought has been forecast (by monitoring the phenomenon) would be a tactical response to the information provided. A decision to always plant crops suited to a highly variable climate would be a strategic response to the problems.

The impacts of ENSO on human health are relevant to sustainable development. As noted above, the high climatic variability increases the potential for climate-related food shortages and epidemics of gastrointestinal illness and arboviruses. Affected countries need to be aware of the higher probability of climatic anomalies on the potential threat to human health. Tactical responses and strategic action might be necessary to minimize the consequences of climatic anomalies related to ENSO. A tactical response could be the use of greater prophylactic measures (for example, spraying to control mosquitoes) when a La Niņa event (with concomitant heavy rains) is under way, because of the greater likelihood of arboviral epidemics. A strategic approach would be to stockpile greater food supplies than might be necessary in an area of lower climatic variability. The wide spatial scales of the phenomenon, which means that food shortages might occur over large parts of South-East Asia concurrently, also need to be considered. These spatial aspects would complicate the use of internal food-relief strategies to overcome shortages unless adequate stockpiles are developed.

Strategic and tactical approaches to the problems raised by ENSO in agriculture and land management can also be devised. Since, as noted above, soil erosion and wildfires may be more frequent and severe in areas affected by ENSO, strategic responses, such as the use of tillage practices appropriate to areas of highly variable rainfall, might reduce these deleterious effects. Similarly, recognition of the higher danger of wildfires calls for the development of forest-management strategies to minimize the risk of fires. The choice of crops suitable to a highly variable climate, rather than those suitable for more consistent climates, would also be a strategic response.

Tactical approaches would include the use of seasonal climate outlooks that ENSO can provide to determine whether to burn off, for example. Thus, slas-hand-burn operations might be restricted during an El Niņo event to minimize the risk of fire. Tillage might be reduced at the end of major El Niņo events when the heavy rains associated with the breaking of the 'El Niņo' drought might increase soil erosion. Leaving residual vegetation and crop stubble in the fields at such times may also reduce erosion. Tactical use of seasonal climate outlooks may also be useful in crop planning; for example, in some years, two crops may be possible when the wet season starts early. The date of the onset of the wet season could be predicted with ENSO.

These are just a few examples indicating how knowledge of the climatic effects of ENSO in South-East Asia, and their impacts on the ecology, environment, agriculture and economy might be used to help attain a sustainable environmental future. Many others may present themselves later. Future use of the expanding knowledge may prevent the long-term environmental changes that have occurred elsewhere through the interaction between ENSO and land-management practices designed for less variable, but less predictable, climates. But whether ENSO will continue to affect South-East Asia as it has done in the past needs to be monitored. If, for example, ENSO stops operating, this would mean a less variable climate. A change in the global climate might conceivably lead to a change in the behaviour of the ENSO phenomenon.

ENSO in the past and future

It is clearly important for South-East Asia to be able to predict how ENSO might be affected by global warming. Several suggestions have been made about how this might be done, including the use of paleoclimatic information, simple theories of ENSO and numerical models. Most of these approaches are flawed and produce conflicting predictions. Therefore, no reliable predictions of a likely ENSO reaction to global warming can be made at this stage. In the long run, a coupled ocean-atmosphere model capable of realistically simulating the present-day behaviour of the phenomenon must be developed. It could then be used to predict any reaction to global warming. Such a simulation is not yet available, although several groups have developed models which reproduce some aspects of ENSO. Some models of ENSO are sensitive to the background model 'climates' (Enfield, 1989). If the models are adequately simulating the real behaviour of ENSO, this sensitivity would suggest that global warming may produce marked changes in its behaviour.

Until a credible model is available, perhaps the best approach to predicting the phenomenon's future behaviour is to examine its past behaviour. If it is sure that change has not been marked in, say, the last 10,000 years, it can be confidently assumed that it is strong enough to survive global warming on the grounds that it has survived past climatic changes. On the other hand, evidence of considerable variations, such as frequency of occurrence, would indicate that it was not very robust with respect to climatic change, as suggested by the rudimentary ENSO models presently available.

The SOI is plotted for over a century, 1880-1990, in Figure 7.8. The time series shows little evidence of major changes or instability in behaviour, despite the 0.5 °C global warming observed over this period.

Unconventional data sources must be used to plot the behaviour of ENSO prior to the middle of the nineteenth century, because of the paucity of appropriate meteorological data. Nicholls (1988b) used documentary sources to show that ENSO was operating and affecting both sides of the Pacific in the first half of the nineteenth century. Quinn, Neal and Antenuz de Mayolo (1987) plotted the strong and moderate El Niņo events (using documentary records of heavy rainfalls in northern Peru) that have occurred since 1525. Enfield (1989) demonstrated that El Niņo intensities and intervals between events seem no different now than they were four centuries ago. Lough and Fritts (1985) reconstructed the SOI back to 1601 from tree-ring data and also found no clear trends in its behaviour. Murphy and Whetton (1989) found an ENSO signal in Java tree rings from between 1514 and 1929. They compared tree rings in years identified as El Niņo events by Quinn, Neal and Antenuz de Mayolo (1987) with those of other years and found a tendency for reduced tree-ring width (implying drought) in the year before El Niņo events were observed in the eastern Pacific. The heavy eastern Pacific rains associated with El Niņo normally come at the end of an event (for example, 1983 at the end of the 1982-3 event). All these different lines of evidence imply that ENSO has been robust with respect to climatic changes of the magnitude of the 'Little Ice Age', a period of globally lower temperatures that extended roughly from 1500 to 1850.

Whether ENSO exhibited its present-day behaviour earlier than this is difficult to say, although several studies of proxy data such as ice-cores, coral cores and sediments suggest that at least some El Niņo events occurred thousands of years ago. Enfield (1989) reviews some of this evidence. A different approach was suggested by Nicholls (1989). As noted earlier, ENSO causes highly variable rainfall in the areas it affects. The wildlife and vegetation in at least some of the areas (for example, Australia) is very well adapted to highly variable rainfall, implying that ENSO may have been operating over evolutionary time scales in much the same way as today. If so, it has survived major climatic changes and might be expected to survive global warming of the magnitude predicted from the enhanced greenhouse effect.

FIGURE 7.8 Annual Average Southern Oscillation Index, 1880-1990

From the evidence available up to the early 1990s, there seems little reason to believe that major changes in ENSO will occur due to global warming over the next few decades. Certainly there is no way, yet, of predicting the changes. A definitive answer, however, requires the production of a model capable of simulating ENSO's present-day behaviour.

Future work

Further work is needed to document the exact influence of ENSO throughout South-East Asia. Exhaustive studies, such as those by Allen (1989) and Allen, Brookfield and Byron ( 1989) for the highlands of Papua New Guinea should be completed for the entire region. Improvements in the accessibility and quality of climate data are needed if this is to be done. Some interesting interactions between ENSO and the background climate can be expected, because of the strong seasonal climatic variations in this region. The effect of ENSO on rainfall at Kota Kinabalu (Sabah), for example, is illustrated in Figure 7.9. Rainfall here usually exhibits twin peaks, as the Intertropical Convergence Zone (ITCZ) passes over the station in June and October. During La Nina episodes, these peaks are exaggerated; while in El Niņo episodes, they are almost removed and a new peak rainfall appears in August.

Further work is also needed to assess impacts of climatic anomalies associated with ENSO in the region. Effects on agriculture and human health, in particular, need to be better documented. The implications of the high variability and temporal patterning of ENSO-related climatic anomalies for forest management, soil conservation and wildlife management also have implications for the development of a sustainable environmental future for the region. Finally, ways of using the information provided by ENSO about the region's climatic variations (including the seasonal predictability) to enhance the environment, economy and lifestyle of the region need to be developed.

FIGURE 7.9 Mean Rainfall and Composite Rainfall in El Niņo and La Nina Years at Kota Kinabalu (Sabah)

The South-East Asian region is likely to continue to be affected in much the same ways as it is now, at least for the foreseeable future. As populations increase, greater pressures will be placed on the environment. These can combine with the highly variable climate produced by ENSO to yield further rapid and irreversible changes to the environment. Amplification of climatic variations caused by ENSO in South-East Asia and other areas will make sustainable development more difficult than will be the case in the unaffected areas. Both the knowledge of the way ENSO influences the climate and the predictability provided by the phenomenon need to be improved and utilized in the quest for sustainable development.

A successful prediction using unconventional data

Prediction of the 1991 ENSO event in Indonesia
Editorial comment


NICHOLLS has prepared an excellent and comprehensive review on the El NiņoSouthern Oscillation (ENSO) with particular emphasis on its impact on drought and flooding rain in South-East Asia. Therefore, there is little room for discussion. However, some additional information on the ENSO phenomenon from an oceanographic viewpoint can be offered, and also a plan of integrated inquiry and the development of capabilities in the region, both for research and for predicting ENSO.

Located between the Pacific and the Indian Oceans and between the Asian and Australian continents, the South-East Asian seas are strongly influenced by the monsoonal climate pattern. The South-East Asian seas, dominated by the Indonesian seas, form the only tropical interoceanic link between a reservoir of warm surface water in the western Pacific and the eastern Indian Ocean. The heat and water flux between the two oceans through this link is estimated to be considerable and has a large, perhaps even global-scale impact on the ocean-atmosphere system.

The ENSO phenomenon generates adverse climatic effects regionally over the whole Pacific basin, even globally. The constant westward Equatorial tradewind pushes the warm surface water of the Pacific, piling it up in the western Pacific ocean just north of the Indonesian archipelago. Apparently, a 'western Pacific warm pool' is a key factor in triggering ENSO. Therefore, an international programme called COARE (Coupled Ocean Atmosphere Response Experiment) is now under way to study the interaction of oceanography and meteorology in an area between 140 and 180 °E, 10 °S and 10 °N, just off Papua New Guinea. This experiment is part of an International Tropical Ocean and Global Atmosphere (TOGA) programme. Some of the results of studies and modelling have been presented in the last International TOGA Scientific Conference held in Honolulu, Hawaii during 16-20 July 1990.

The South-East Asian countries participate actively in various international programmes to study the impact of global change on ENSO. Aside from TOGA, some examples of other programmes are:

Committee for Climate Changes and the Ocean, both the Indian Ocean Panel as well as the Pacific Ocean Panel;

Global Sea Level Observing System, which uses a network of tide gauges installed in the region;

Regional Ocean Dynamics (an ASEAN-Australia co-operative programme);

Subcommission of the Western Pacific (of IOC); and

World Ocean Circulation Experiment.

Prediction of the 1991 ENSO event in Indonesia

With all the limitations of lack of technological capabilities and equipment, researchers in the South-East Asian countries themselves have started to develop capabilities for predicting ENSO and sea-level rise. In co-operation with the Regional Office for Science and Technology in South-East Asia of UNESCO in Jakarta, Indonesian oceanographers offered an early prediction of an ENSO event in 1990-1. They relied particularly on anomalous coastal flooding that took place on the north coast of Java in November 1989, and again in December 1990, interpreting this as the consequence of a wave projected eastwards through the Indonesian archipelago from the Indian Ocean. In January 1991, using this information and rainfall data, they issued a statement predicting an event in 1991 that would be less intense than that of 1982-3, but stronger than the small event in 1986-7. This turned out to be a correct prediction, and it is something of a coup for the Third World scientists involved, who succeeded in making a long-range prediction without the help of high-tech monitoring capabilities and little in the way of computer capabilities.

The story began in mid-November 1989, when the old city of Jakarta was suddenly flooded by sea water to a depth of 30 40 centimetres above the normal high-water mark. This happened in fine, calm weather, and it imposed massive damage on food storages thought to be safe from flooding. By coincidence, a United Nations Development Programme-United Nations Educational, Scientific and Cultural Organization (UNDPUNESCO) training course at a sea-coast laboratory was delayed by the flooding, and the assembled Indonesian scientists turned the occasion into an on-the-spot workshop. The extent of the flooding implied that a very great deal of energy was involved, and suggested that the event was therefore of global significance.

Efforts to attract international scientific interest in the November 1989 phenomenon proved fruitless; most oceanographers dismissed the event as a consequence of coastal subsidence, and of no climatological significance. The Lembaga llmu Pengetahuan Indonesia (LIPI) and UNESCO then held an international workshop in March 1990, with a principal focus on global climatic change as it affected the coastal environment. This meeting, and a national meeting in October, concluded that no such standard explanation fitted the facts.

By mid-October 1990, however, it was becoming clear that the onset of the annual wet season was delayed. Then, in November, a further unusual phenomenon was noted by a UNESCO scientist studying behaviour patterns among birds in relation to predicted global climatic change. Collard or Oriental Pratincoles, last seen in Jakarta during the 1982-3 drought, reappeared in the city. These plover-like birds move slowly on the ground but fly like swallows, catching much of their insect food on the wing. They winter in dry, open country in northern Australia, but in November 1982 and again in 1990, small flocks were seen hawking insects in the southern suburbs of Jakarta.

By mid-December 1990, the wet season was clearly abnormal, not only in lava but also in Peninsular Malaysia and Sabah. By this time, an anomalously warm pool in the central Pacific, accumulating since early 1989, was becoming pronounced, with cyclogenesis further east than normal in the Pacific, leading to cautious international prognostications that an ENSO might be developing. Then the city of Jakarta again experienced coastal marine flooding on 3 December 1991. LIPI and UNESCO issued their prediction on 4 January 1991.

Editorial comment

Soegiarto went on to appeal that much greater attention be paid to unconventional indicators, such as bird behaviour and unusual levels of the sea, and that Third World scientists should be involved in an extensive monitoring network, especially in the critical region of South-East Asia and northern Australia. At present, many significant anomalies go undetected by the remote-sensing and instrumental network, on which prediction principally relies. Nicholls, whose own paper had stressed the use of unconventional data, and of biological responses in particular, welcomed this suggestion which was taken up in discussion as a proposal that should be adopted by the conference as a whole.

The discussion was also extended to correlates with the anomalously wet, anti-ENSO or La Nina events which commonly either precede or follow an El Niņo, as described in Nicholls' Chapter 7. James Fox drew attention to stem borer damage in the rice crop, the correlation of which with low dry-season rainfall in East Java was first studied by Van der Laan (1959). This, it seems from subsequent correspondence between Nicholls and Fox, can tentatively be shown to be particularly severe one year after a La Nina event, that is, in a dry year following an unusually wet one.

A similar correlation between the same pairing of rainfall anomalies and low production of sweet potatoes in the highlands of Papua New Guinea was remarked upon by Allen; the wider question of the agricultural significance of this pairing was taken up in a subsequent paper by Brookfield and Allen (1991). At the meeting, Allen stressed the need to pay attention to extremes, rather than averages. Many social systems in the region contain elements which are adaptations to extremes, with continuing institutions that are like ghosts, responsive to extremes in the past.

As is noted in the Introduction to Part II, the anomalous wet season of 1990-1 was followed by a severe and prolonged 1991 dry season in Indonesia, as also in a large part of north-eastern Australia. By August 1991, large forest fires had again broken out in Kalimantan and Sumatra, burning some 30 000 hectares in Kalimantan by October. The rice crop in lava was also seriously affected by the drought.


8. Climatic change and agriculture: Problems for the Asian tropics

Climatic change in tropical Asia, 1910s-1980s
Problems in climate-agriculture relationships: Rice yields in three areas
Flow of impacts of climatic change on agriculture
Concluding remarks


PROBLEMS for agriculture arising from climatic change are reviewed in this chapter. In 1984, the author summarized similar problems (Yoshino, 1984b). This chapter is an updated version, in the context of new findings on global climatic change. Parry, Mendzhulin and Sinha (1990) reviewed, globally, the potential impact of climatic change on agriculture and forestry. They point out that present-day vulnerability to climate is substantial. Of the land area of the developing countries, 63 per cent is climatically suited to rainfed agriculture, and the corresponding figure is about 84 per cent in South-East Asia. However, the topographically and pedologically suitable area is far smaller. After a review of the facts of climatic change in the region, the chapter presents as examples some data on crop climate relationships in Sri Lanka, Indonesia and Hainan Island, South China. This leads to a more general discussion of relationships between climatic change and agriculture, forestry and fisheries.

Climatic change in tropical Asia, 1910s-1980s

Yamamoto (1990) has calculated rates of secular change of air temperature during the 1910s-1980s in the various regions of the oceanic areas. There are several important findings related to tropical Asia. During this period, the range between maximum and minimum temperature is the largest in the zone 50 70 °N and the smallest in the zone 10 30 °N. Since 1970, curves show a warming tendency in the zones 50-70 °N and between 10 °N and 10 °S, but a cooling in the zone 30-50 °N, and no clear tendency in the zone 10 30 °N over the Pacific. A recent study for Indonesia (Sutamihardja and Sutrisno, 1991) revealed evidence of a clear increase of minimum and maximum air temperatures at Jakarta during 1866 1989 (Figures 8.1-8.2). Although there are problems with the data, so that those for 1944-51 in Figure 8.1 should be questioned, and data for the period after 1970 in Figure 8.2 must be rejected, the increases during the last 50 years are striking.

The heat-island effect of a rapidly growing city should, of course, be deducted from the warming due to global tendencies. There are also long-term variations of rainfall in the Asian Tropics, but they differ from region to region, so that it is hard to draw a general conclusion. Year-to-year variations in rainfall have important effects on agricultural activities but, in the Asian Tropics, their range of fluctuation, cycles or periodicities and abrupt changes, have not yet been clearly analysed. Henderson-Sellers (Chapter 6) has discussed the large differences in the GCM modelling results for tropical Asia, and clearly further information is needed before any prediction is possible concerning the regional consequences of doubling carbon dioxide (CO2). Meanwhile, it is worthwhile to analyse present problems, especially where these are of a nature that could be modified significantly by global climatic change.

FIGURE 8.1 Secular Change of Annual Minimum Air Temperature at Jakarta, 1866-1989

FIGURE 8.2 Secular Change of Annual Maximum Air Temperature at Jakarta, 1866 1989

Contents - Previous - Next