Contents - Previous - Next


This is the old United Nations University website. Visit the new site at http://unu.edu


Climatic change and public policy

SHAM SANI

BOTH Chapters 6 and 7 are basically related and are therefore equally relevant in terms of policy response considerations. The comments here are more apposite to the first of these chapters, but they do not refer to the detailed climatology discussed by HendersonSellers. They focus mainly on matters pertaining to policy responses-an issue which is raised in both chapters but not treated in any detail. However, with regard to sea-level rise, Tjia (1989), using more than 150 radiometrically dated shoreline indicators, suggested that actual sea level within the South-East Asian region is expected to decline in the near future at rates between 1.5 and 2.0 millimetres per year. This decline is expected to compensate for the projected 20-centimetre rise in sea level due to the greenhouse effect so that the net rise by 2025 will only be between 13 and 15 centimetres. Further, it is interesting to note that a number of the observations made in Chapter 7 have been similarly observed by the Malaysian Meteorological Service, and they were reported in its Technical Reports (Cheang, 1990; Quah, 1984, 1988).

It is evident that a great deal of work is needed not only to document the exact influence of the El Niņo Southern Oscillation (ENSO) phenomenon in South-East Asia but also to improve the existing climate models. Consideration of the more important climatic features like tropical cyclones, monsoons and ENSO events which have so much influence on the region's weather is especially important.

It is doubtful that anyone would dispute the views expressed about the generally poor climate predictability in the region. Detailed and accurate predictions on the magnitude of climatic change at regional and local levels are not possible on the basis of the state of knowledge in the early 1990s, and it will take a long time yet before precise predictions can be achieved in this part of the world. However, it is important to note that the governments of at least some South-East Asian countries are aware of global climatic change and its likely implications on human activities. They are making efforts to improve understanding of the nature and mechanisms of regional climate, evidence of climatic change, the likely climate scenario given a doubling of CO2 by 2030, the impact of such a climate scenario on agriculture, water resources, coastal and marine resources and policy options.

Detailed features of the generated scenarios may be somewhat exaggerated, or even underestimated, but they are nevertheless useful first approximations upon which policies and strategies can be based. Such policies can gradually be refined as more information becomes available. One good effort towards such an objective is reflected in a recent United Nations Environment Programme (UNEP) project on 'SocioEconomic Impacts and Policy Responses Resulting from Climate Change: A Regional Study in Southeast Asia'. While this project is probably not going to be the answer to prayers regarding climatic change, it is certainly a step in the right direction. The UNEP project was jointly undertaken in 1989 by Indonesia, Malaysia and Thailand. The objectives of the project were to generate the climate scenario, assuming a doubling of CO2 by 2030; to assess the impact of such a scenario on some important activity sectors; and to select appropriate policies and strategies in order to respond to future climatic change.

TABLE 6.2 Examples of Climatic Change Impacts

  1. Rice yield decreases by 12-22 per cent, but may be offset by CO2 increases.
  2. Maize production is not significantly affected. It is more sensitive to solar radiation changes.
  3. Palm-oil yield will be affected if dry seasons and several months of reduced sunshine occur.
  4. The limitation to rubber cultivation is negligible if temperature increases by only 2 °C. A 3-15 per cent decrease will occur if there is increased drought. A rainfall increase of 10 per cent can cause a 13 per cent decrease in yield.
  5. In the Kelantan River basin, an increase of flood peaks and duration is forecast. A 30-35 per cent increase in water deficits in the dry season can also be anticipated.

Source: Condensed from Chong (1990).

TABLE 6.3 Examples of Possible Policy Responses

Agriculture and Water Resources

  1. Breeding new crop varieties.
  2. Maintenance of broad genetic base.
  3. Policy on more efficient control and use of water resources.
  4. Review policy on subsidies; possible increase in subsidies.
  5. Encourage more intensive agriculture; reduce land fragmentation.
  6. Diversification of employment opportunities among farmers.
  7. Awareness programmes for planners and project implementors.
  8. Comprehensive monitoring programme regarding climate change.
  9. Water resource use and management policy-priority of water use; water pricing; water regulation and distribution.

Coastal Resources

  1. Review existing structural measures to prevent erosion.
  2. Relocation of population and important infrastructural facilities from areas likely to suffer immediate inundation.
  3. Monitoring and assessment.

Source: As for Table 6.2.

In Malaysia, as a result of the project, a number of interesting prognostications have now become available with regard to climatic change. Tables 6.2 and 6.3 provide examples.

7. Enso, drought and flooding rain in south-east Asia

Introduction
The El Niņo-southern oscillation
ENSO and south-east Asia
Effects of ENSO on climate
Impacts of ENSO
ENSO in the past and future
Future work

NEVILLE NICHOLLS

Introduction

THE El Niņo-Southern Oscillation (ENSO) affects the climate, natural vegetation and wildlife, agriculture, human health and economies of many of the countries bordering the Pacific and Indian Oceans. Because of its pervasive influence on many aspects of life, it needs to be considered in discussions of sustainable development. There have been instances where ignorance of its effects has led to land degradation or long-term vegetation changes. It has implications for the health of the population in the areas it affects and these also need to be considered.

The influence of ENSO on South-East Asia has not yet been comprehensively mapped. It has an important and well-documented role in controlling the interannual climatic variations of Indonesia, but much work remains to be done to determine its climatic effects and its ecological, environmental, social and economic impacts elsewhere in the region. Throughout this chapter, therefore, examples of its effects and impacts will be taken from other areas, especially Australia and New Guinea (Irian Jaya and Papua New Guinea) (from where extensive documentation of the phenomenon and its effects are available), to supplement information from South-East Asia. The relevance of ENSO, and this chapter, to sustainable development in South-East Asia, varies over the region. These variations depend on the strength of the phenomenon's influence in different parts and also on the use of the area by the local population.

The El Niņo-southern oscillation

Variations in climate from year to year appear at first glance to be random. Examination of historical data, however, reveals a coherent global pattern of oceanic and atmospheric fluctuations called the Southern Oscillation. Extreme anomalies in this pattern involve dislocations of rainfall distribution in the Tropics, bringing drought to some regions and torrential rains to others (Ropelewski and Halpert, 1987, 1989). These anomalies typically last about a year. Related anomalies of the atmospheric circulation extend high into the atmosphere and polewards into the temperate zones, especially in the southern hemisphere.

FIGURE 7.1 Circulation during an El Niņo Phase

FIGURE 7.2 Circulation during a La Nina Phase

FIGURE 7.3 Southern Oscillation Index (monthly means)

Some major changes in the ocean currents and temperatures are also related to the Oscillation. The best known of these is the El Niņo, a marked temperature increase that occurs every few years in the eastern Equatorial Pacific with catastrophic effects on marine ecosystems along the west coast of the Americas. Because El Niņo usually occurs with an extreme anomaly in the Oscillation, the two phenomena are often referred to jointly as 'El Niņo-Southern Oscillation' or 'ENSO'. Periods with very warm sea surface temperatures (SSTs) in the eastern Equatorial Pacific, and the global pattern of climatic anomalies that usually accompany this warm water, are referred to as El Niņo events. A major El Niņo event occurred in 1982-3 with severe droughts in Australia, Indonesia, parts of Africa, and India. A schematic of the atmospheric circulation during an El Niņo is provided in Figure 7.1.

During the other extreme of the Oscillation, the eastern Equatorial Pacific is cold (a phenomenon now called 'La Nina') and heavy rainfall and flooding is observed over the areas usually affected by drought during El Niņo events. (See Figure 7.2 for a schematic depiction of a La Nina.) Heavy rainfall in India, Africa and Australia during 1988 was associated with a La Nina event. The dislocations in rainfall distribution associated with El Niņo and La Nina events mean that the areas affected tend to have more variable rainfall than is the case elsewhere.

ENSO is the result of interactions between the tropical oceans (especially the Pacific) and the atmosphere. The detailed form of this interaction is yet to be determined. Major research programmes aimed at modelling it are under way. Models of the Equatorial Pacific Ocean and the atmosphere, apparently capable of reproducing and predicting some aspects of the phenomenon, have been developed (Barrett et al., 1988). A model capable of a realistic simulation of the complete phenomenon, however, has yet to be developed.

There is often confusion about the terminology used in discussions. El Niņo events are just one extreme of the quasi-cyclic ENSO phenomenon. La Nina events are the other extreme. These are illustrated in Figure 7.3 which shows monthly values of the Southern Oscillation Index (SOI). This index, which reflects the behaviour of ENSO, is the standardized difference in pressure between Tahiti and Darwin. El Niņo events occur when the SOI is at large negative values; La Nina events occur at large positive values. The SOI fluctuates quasi-periodically; the nature of this fluctuation is discussed below.

ENSO and south-east Asia

The relationship between Indonesian rainfall and the Southern Oscillation has been documented many times this century (Berlage, 1927; Nicholls, 1981; Quinn et al., 1978; Rasmusson and Carpenter, 1982). Braak (1919,1921-9) and Berlage (1927, 1934) demonstrated that rainfall during the early part of the Indonesian wet season (SeptemberDecember) was significantly related to the Oscillation. Abnormally low atmospheric pressure at Darwin (an indication that ENSO is in its 'cold' La Nina phase) usually signals an early start to the wet. Nicholls (1973) showed that droughts in New Guinea often coincided with El Niņo episodes. Quinn et al. (1978) confirmed the relationship between ENSO and Indonesian rainfall, pointing out that droughts during the 'dry' season of easterly surface winds (May-November) usually coincided with El Niņo events. Figure 7.4 shows a composite of Indonesian rainfall anomalies before, during and after El Niņo episodes. The effect of El Niņo events on Indonesian rainfall is substantial enough to be observable in tree rings (Murphy and Whetton, 1989). Barry (1978) and Wright, Mitchell and Wallace (1985) showed that cloudiness over parts of South-East Asia was reduced during El Niņo events.

The pattern of Indonesian and New Guinea rainfall fluctuations was reexamined, using more extensive data, by Ropelewski and Halpert (1987, 1989). Their earlier work (1987) demonstrated that 80 per cent of the El Niņo events from 1879 to 1982 were accompanied by below average rainfall between June and November. In their 1989 paper, they found that 90 per cent of La Niņa episodes were abnormally wet in Indonesia and New Guinea between July and December. Allen, Brookfield and Byron (1989), examining earlier nineteenth-century data, found that several El Niņo events (notably 1877-8, 1864 and 1804) were associated with Indonesian drought. Kiladis and Diaz (1989) composited temperature and precipitation records during El Niņo and La Niņa events, which showed that most of South-East Asia in the period September-May during El Niņo events was warmer than for the same seasons during La Niņa events. Significant precipitation anomalies were only evident over Indonesia, the Philippines and Singapore. Figure 7.5 is a composite of Singapore rainfall during El Niņo and La Niņa episodes. It illustrates that rainfall is usually higher during La Niņa episodes, as is the case also over much of Indonesia.

FIGURE 7.4 El Niņo Composite Precipitation for Seven Indonesian Stations

Singapore rainfall during El Niņo and La Niņa episodes

Rasmusson and Carpenter (1982), analysing surface winds during El Niņo events, found a weakened north-east monsoon circulation over the Philippines around the end of the calendar year during an El Niņo. Wind anomalies at other times through the El Niņo were weak and variable. The weak monsoon probably accounts for the finding by Ropelewski and Halpert ( 1987) that the Philippine precipitation signal associated with an El Niņo was low rainfall between October and May, at the end of the event; whereas in Indonesia, rainfall was most suppressed during June-November.

The evidence available does indicate that ENSO influences the climate of much of South-East Asia. In the following section, the characteristics of climatic variations induced by the phenomenon are discussed. ENSO leads to quite different patterns of climatic anomalies across the areas it affects. These climate patterns should be considered when sustainable development strategies are being developed.

Effects of ENSO on climate

Rainfall Fluctuatians

The best known characteristic of ENSO is the tendency for rainfall anomalies to appear in many areas at about the same time. Thus, droughts in India, North China, Australia and parts of Africa and the Americas tend to occur approximately simultaneously (Ropelewski and Halpert, 1987; Williams, Adamson and Baxter, 1986). The review in the previous section indicates that droughts tend also to occur over at least parts of South-East Asia during El Niņo episodes. At the same time, unusually heavy rainfall occurs in the central and east Pacific. These 'teleconnections', although interesting, are less important to specific regions than some other characteristics of ENSO, but all areas are likely to experience some of the characteristics. These areas should include the parts of South-East Asia affected by the phenomenon, although little work has so far been done to verify this.

High Variability

One feature of rainfall fluctuations in areas influenced by ENSO is a large interannual variability. Conrad (1941) examined the dependence of interannual rainfall variability on the long-term mean annual rainfall. He found that a function relating relative variability (defined as the mean of the absolute deviations of annual rainfalls from the long-term mean, expressed as a percentage of the long-term mean) to the long-term mean precipitation fitted his data closely. The relative variability decreased, in general, as the mean precipitation increased. Over some large areas, however, the relative variability deviated in a consistent way from the global relationship with mean rainfall.

Some of these deviations were due to the influence of the ENSO phenomenon on rainfall. Nicholls (1988a), using Conrad's data, found that the relative variability was typically one-third to one-half higher for stations in areas affected by ENSO compared with stations with the same mean rainfall in areas not so affected.

Nicholls and Wong (1990) confirmed, using post-1950 data and the coefficient of variation as a measure of relative variability, that ENSO does amplify rainfall variability in the areas it affects, relative to elsewhere. This effect was strongest at low latitudes and low rainfalls and so is especially relevant to semi-arid areas in the Tropics and subTropics. The amplification factor is substantial. The variance of annual rainfall in an area strongly affected by the Southern Oscillation might be, depending on latitude and mean rainfall, more than double that in an area with similar mean rainfall not influenced by the Oscillation. Further work needs to be done to examine the variability not just on an annual basis but in the seasons most affected by the phenomenon.

LARGE SPATIAL SCALES

The large-scale nature of ENSO means that wide areas suffer from the same rainfall anomaly at the same time. In Australia, for instance, much of the central and eastern parts is usually in drought during an El Niņo event. The same tendency should also apply in Indonesia. Droughts will be greater in scale than might be expected without the influence of ENSO. These large spatial scales, relative to the droughts in regions not affected, may complicate the measures for providing drought relief and imply that similar pressures on the environment will be exerted across much of the country at the same time.

LONG TIME SCALES

In most parts of the world, it is assumed that the duration of droughts or wet periods are random variables. This is not the case where ENSO is experienced. Droughts, and pluvial or wet periods, tend to last about 12 months or so in these areas (Ropelewski and Halpert, 1987). El Niņo and La Nina events also last about 12 months (Rasmusson and Carpenter, 1982) and this sets the time-scale of the rainfall fluctuations. This long life cycle is clear in Figure 7.3 which shows monthly SOI values from 1970 to 1989. The El Niņo and La Nina events are marked. There is a clear tendency for them to last about 12 months.

Figure 7.6 shows the monthly rainfall anomalies at Jakarta for 1976-80. The monthly values of the SOI are also plotted on the figure. There was a strong El Niņo in 1977-8, represented by the negative SOI values. (SOI values are unavailable for the first six months of 1977.) Jakarta rainfall was consistently below average from May 1977 to February 1978. This extended period of drought is typical of the situation during an El Niņo. Similarly, during a La Nina, substantial rains tend to occur for about 12 months in the areas around the western edge of the Pacific that are affected by ENSO.

Phase-locking to Annual Cycle

These extended periods of drought or heavy rain do not occur randomly over time, in relation to the annual cycle, as demonstrated by Figures 7.3-7.6 and the results of Ropelewski and Halpert (1987) and many others. In fact, the ENSO phenomenon, and rainfall fluctuations associated with it, are phase-locked with the annual cycle (that is, they tend to recur at the same time of the year). The heavy rainfall of a La Nina tends to start early in the calendar year and finish early in the following year. The dry periods associated with El Niņo events usually occupy a similar time period (see, for example, Figure 7.4). This means that if an extensive drought or wet period is well established by the middle of the calendar year, it is unlikely to 'break' until at least early the following year. The 1982-3 Indonesian drought provides another example. The drought started about April 1982 and lasted until at least January 1983. Such phase-locking has been found in most other variables associated with ENSO (Rasmusson and Carpenter, 1982).

BIENNIAL CYCLE

This phase-locking is related to a biennial cycle which is a fundamental element of ENSO variability (Rasmusson, Wang and Ropelewski, 1990). There is also a lower-frequency variation, but it is the biennial mode which captures the major features associated with El Niņo and La Nina episodes. The biennial cycle is observed over the Equatorial Pacific and Indian Oceans and is itself phase-locked with the annual cycle. It varies in amplitude from cycle to cycle and sometimes changes phase. Nicholls (1979, 1984) discussed how ocean-atmosphere interaction around Indonesia, modulated by the seasonal cycle, could result in a biennial cycle phase-locked to the annual cycle.

FIGURE 7.6 Monthly Rainfall Anomalies at Jakarta, 1976 1980

The biennial mode means that El Niņo events will often be preceded and/or followed by La Niņas and vice-versa. In terms of rainfall, this means that year-to-year changes in rainfall can be extreme. Change from El Niņo-related drought to La Nina and pluvial conditions can be rapid, and it usually occurs early in the calendar year. An El Niņo-related drought in 1925 in Indonesia, for instance, was followed by three months (January-March 1926) of about double the average rainfall in Jakarta. The descent into drought can also be rapid. In Jakarta, the 1902 El Niņo was accompanied by a drought which started after three months (December-February) of well above average rainfall. Each of the next 10 months received well below average rainfall.

EXTREME EVENTS

The amplified climatic variability induced by ENSO and the temporal patterns of the variations leads to different distributions of extreme climatic events, compared with other areas with a more benign climate. Obviously, droughts and floods will be more severe in the ENSO-affected areas, including, presumably, South-East Asia. The biennial nature of ENSO, however, means that floods can quickly follow droughts. In Australia, such rapid changes from drought to flood have resulted in unanticipated failures of earth dams (Ingles, 1990). In other areas, such 'rare' events as heavy rains are considered to occur randomly in time, so the likelihood of 'pairing' a drought and a flood within a short time is considered highly unlikely. In Australia, and probably in many areas affected by the biennial nature of ENSO, this is not a reasonable assumption. These areas need to identify the frequency distributions of extreme events and consider the possibility that pairs or even clusters of 'rare' events might not be as rare as is the case in regions with more benign climates. Statistical models for extreme events developed in areas not affected by ENSO should be treated with some suspicion.

Winds and Temperatures

Rainfall is not the only aspect of the climate influenced by ENSO. Kiladis and Diaz (1989) reported increased temperatures throughout South-East Asia during El Niņo events, relative to La Niņa episodes. Figure 7.7 shows composite mean monthly temperatures in Singapore during El Niņo and La Nina years (and the preceding and following years). Long-term mean temperatures are also shown. Cooler temperatures are evident from about May in the year before the El Niņo through to about April of the El Niņo year. From December onwards, temperatures are higher in the El Niņo years and the following year, relative to the long-term mean and the La Nina years. A similar pattern could be expected throughout much of the region according to the results presented. Figure 7.7 also indicates the phase-locking to the annual cycle that is evident in temperature as well as rainfall anomalies.

The lack of cloud cover over parts of South-East Asia, such as Indonesia, during El Niņo events suggests that increased radiational cooling at these times might lead to decreased minimum temperatures. Allen (1989) and Allen, Brookfield and Byron (1989) reported a tendency for frosts in the New Guinea highlands to be more severe and widespread during El Niņo. The 1982 El Niņo, for instance, led to frosts throughout the Papua New Guinea and Irian Jaya highlands.

There are also clear variations in wind between El Niņo and La Niņa events, especially close to the Equator (see Figures 7.1 and 7.2). As mentioned earlier, the northeast monsoon over the Philippines during the northern summer tends to be weaker during an El Niņo (Rasmusson and Carpenter, 1982). The south-east monsoon over western Indonesia and Malaysia tends to be stronger during an El Niņo episode. These variations in the basic flow, in conjunction with the number of islands of the region, will result in many local variations in wind speed and direction between El Niņo and La Nina years. In turn, these local variations will lead to local differences in precipitation or temperature that may be very different to the broad-scale anomalies associated with El Niņo; for instance, Nicholls (1973) found that rainfall at Lae (on the coast) tended to be high during El Niņo events, even though the opposite was the case over much of the rest of New Guinea. The increase in rainfall at Lae was attributed to the interaction between the changes in the prevailing winds associated with El Niņo and the local topography.

Predictability of Climatic Fluctuations

The biennial cycle underlying the ENSO phenomenon and the phase-locking of this cycle to the annual cycle provide some regularity to the phenomenon and to climatic variables associated with it. This regularity allows a degree of predictability. The phase-locking means that ENSO, or an index of the phenomenon (for example, the SOI), will tend to change phase around March-May and only rarely at other times of the year. Thus, if the SOI is strongly positive (La Nina) during the middle of the year, it will probably stay in that phase until early the following year. So climatic variations normally associated with this phase of ENSO, and which occur towards the end of the calendar year, may be predictable simply by monitoring the SOI earlier in the year.

FIGURE 7.7 Mean Monthly Temperatures in Singapore during El Niņo and La Niņa Years

The work of Braak (1919, 1921-9) and Berlage (1927, 1934) indicated that early wetseason rainfall for Indonesia could be predicted using a simple index of the Southern Oscillation. Nicholls (1981) confirmed, using data from 1883 to 1965, that Indonesian rainfall in the early part of the wet season could be predicted with an index of ENSO, in this case atmospheric pressure at Darwin. Kiladis and Diaz (1989) found a relationship between the Oscillation and South-East Asian temperature from September to May (see Figure 7.7) which indicates that this rainfall could also be predicted using the Oscillation.

So while ENSO amplifies the climatic variability in the areas it affects, it also allows some degree of predictability of the year-to-year variations. This predictability can, if it is properly utilized, offset the greater variability, in effect, restoring the variability to levels similar to that experienced elsewhere. The use of the seasonal climate predictions made possible by ENSO should play a major role in sustainable development strategies in the areas affected by the phenomenon.


Contents - Previous - Next