a summary of three regional assessments
Edited by Martin L. Parry, Antonio R. Magalhaes and Nguyen Huu Nih, at the Environmental Change Unit, University of Oxford
This report summarises the major conclusions of three regional studies of the potential impact of climate change undertaken by national governments with the support of the United Nations Environment Programme (UNEP). The three studies were in Brazil, in Indonesia, Malaysia and Thailand and in Vietnam. Full details of the studies' results are given in their respective project reports.
The regional assessments adopted a variety of different approaches to assessing potential impacts. In Brazil the emphasis was on identification of the effects that can occur now as a result of the present variability of climate from year to year and season to season. Impacts from climatic variability are thus taken to be a useful analogue of potential future effects of longer-term climate change. In contrast the Indonesia/Malaysia/Thailand study took the current best estimate of possible future climate, using the results of experiments with General Circulation Models of the earth's atmosphere, and considered its economic implications by modelling effects on yields of agricultural crops, etc. The study in Vietnam adopted an approach somewhere between these two, assuming likely long-term changes in climate and describing (rather than modelling) their potential effects.
The study in Brazil is based on 14 regional assessments in the semi-arid Northeast, the industrialized Southeast and South and at the agricultural frontier of the Midwest. The most substantial effects of present-day variability of climate are due to drought, particularly in the Northeast where drought-induced falls in agricultural production cause hunger and malnutrition in peasant communities. Although such drought impacts are still severe, as for example most recently in 1983 and 1987 in the State of Ceara, there is a long experience of government response to the issue which can illuminate the potential range of measures for adapting to both short-term and long-term changes of climate. These include emergency relief actions such as make-work programmes (frentes de emergencia) and efforts to strengthen the regional capacity to cope with droughts such as major public works (i.e. dams, roads and other aspects of rural infrastructure).
While droughts can have a massive social impact in semi-arid parts of northeast Brazil, dry spells in the Midwest, Southeast and South tend to reduce output and have an economic rather than social effect on the more developed economies of these regions. Ironically, however, their effect is greatest on the people of the Northeast owing to increased rural unemployment of immigrant workers from that region. The effects of floods are greatest in the South and Southeast, in both rural and urban areas. Frosts can cause substantial damage to the coffee and citrus industry here also.
The Brazilian case studies indicate that human actions have frequently led to increased vulnerability of society to climatic variability. The nature of the consequent effects of droughts, dry spells, floods and frosts therefore varies greatly from region to region according to the economic and social circumstances that prevail. Policies must be sensitive to these differences and seek to promote sustainable development as a secure means of reducing social vulnerability to climate.
The study in Indonesia/Malaysia/Thailand employed a hierarchy of models to simulate the possible effects of potential future climate change. Outputs from GCMs were used as inputs to models that simulate rice yield and water supply. Expert judgment was then used to assess the economic effects at the enterprise and regional levels, and policy exercises were devised bringing together scientists and policy makers to explore the range of appropriate responses by national governments.
An equivalent doubling of atmospheric CO2, currently estimated to occur in about 2030 assuming current trends in emissions continue, is projected to lead to an increase in mean annual temperatures in the region of about 3-4deg.C by about 2050. Rainfall may increase in some areas but decrease in others and these patterns of change cannot at present be predicted with any confidence.
Increases in rainfall might occur over parts of Indonesia and be sufficient to compensate for higher rates of evaporation due to higher temperatures thus leading to an extension of the irrigable area for rice cultivation. However, even in these circumstances, water availability for crops might be reduced in the early part of the growing season leading to an overall reduction in yields of the major crops (rice, maize and soybean)
In addition, higher temperatures would tend to shorten the maturation period for these crops and increase their demand for irrigation. Studies in Malaysia thus concluded that yields of rice might decline by between 12 and 22 percent and the demand for rice irrigation increase by about 15 percent.
If sea levels were to rise by 10 to 30 cm, the current best estimate by 2030, extensive damage could occur to the fish and prawn industry throughout South-East Asia. In Vietnam, as well as in Indonesia, Malaysia and Thailand, mangrove forests, which are an important breeding ground for fish, could be threatened particularly where these are now backed by a bund that would prevent landward migration. Some low-lying parts of coastal areas could become permanent swamps or lakes due to a rise of the nearcoastal water table and it is possible that the rise in groundwater would be accompanied by the upward movement of subterranean salt resulting in saline damage to rice fields and farmland soils.
An additional complication in South-East Asia is the effect that the El Nino Southern Oscillation (ENSO) phenomenon can have on offshore fisheries in the region. Off the Vietnamese coast in ENSO years there occurs a substantial geographical displacement of areas of upwelling and down welling of waters causing a shift of the productive fishery areas.
A conclusion that emerges from these studies is that, while we cannot yet predict with sufficient precision the nature of likely future changes of climate, we can begin to explore the range of potentially useful measures of response that governments could adopt to mitigate the negative effects of climate change and exploit the more positive ones.
Two types of impact model were used to estimate the consequences of an altered climate in South-East Asia: models of the responses of crop growth, and models of runoff and soil erosion. Models of crop growth were run for three crops: rice, soya and maize. These were CERES or CERES-type models which require daily weather input for precipitation, maximum temperature, minimum temperature and solar radiation. Descriptions of the models are given in the full report.
The SWRRB Water model was used to estimate changes in runoff and soil erosion that might result from changes in climate. Since levels of runoff can affect the accumulation of water in reservoirs, the SWRRB model enabled an estimation to be made of changes in the amount of stored water available for rice Irrigation.
In Indonesia the projected warming under a 2 x CO(2) climate is estimated to lead to an increase in mean annual temperatures of about 3deg.C and a rise in average sea level of about 0.6 m Rainfall could decrease in some regions but might generally increase and, according to the Goddard Institute for Space Studies (GISS) 2 x CO2 experiment, could double in some areas such as south-eastern Indonesia. Such changes in rainfall would be likely to have substantial effects on the amount of water available for irrigation (which might increase and therefore be beneficial), and on rates of soil erosion and soil leaching (which would generally be adverse in their effects on agriculture).
Considering effects on water supply, the estimated increases in rainfall would probably more than compensate for increasing evaporation due to higher temperatures and there would be more water to fill the reservoirs. There could, therefore, be a 30 percent increase in the irrigation area of the Brantas and Citarum Basins in the western part of Java. In the Saddan Basin the area of potential irrigation could increase by 130 percent. However, higher levels of rainfall are likely to increase the rate of soil erosion. Experiments with the SWRRB model for changes in rainfall and temperature projected under the GISS 2xCO2 scenario indicate that erosion rates in the Citarum, Brantas and Saddan watersheds could increase by 14, 18 and about 40 percent respectively. The increase in erosion could result in losses of over 2000 tonnes in soybean production in the upper Citarum River Basin, 2500 tonnes in the Brantas Basin and 2700 tonnes in the Saddan River Basin. Levels of soil fertility would likely be diminished by higher rates of leaching of soluble nutrients, and the study concludes that average fertility levels could decline by 2 to 8 percent.
Rice yields are expected to decrease largely as a result of higher temperatures and, in some instances, reduced water availability. Largest losses would be in early season rice, but overall annual yield losses could be mitigated by increases in late season rice with the consequent average annual yield loss being about 4 percent.
Soybean, an important part of the diet for about three quarters of the Indonesian population, could frequently suffer a yield loss of over 10 percent, largely as a result of lower yields in the early season. If a decrease in insolation occurred, then further reductions in yield could not be avoided, but with appropriate management these potential losses could be compensated by increases in productivity, and overall yields could be expected to increase.
The most severe impacts could be expected on maize, where model experiments with the GISS 2 x CO2 climate scenario indicate reductions in yield of between 25 and 65 percent. Improved management could mitigate these effects and keep the maximum decline to 50 percent.
Additional adverse effects on agriculture are estimated to occur as a result of land losses due to sea-level rise. For example, in the districts of Krawang and Subang 95 percent of the reduction in local rice supply (down 300 000 tonnes) is estimated to occur as a result of inundation of the coastal zone. In the same districts maize output would be reduced by 10 000 tonnes, about half of this due to inundation. Sea-level rise would also be likely to affect fish and prawn production. In the Krawang and Subang Districts the loss is estimated at over 7000 tonnes and 4000 tonnes respectively (valued at over US$0.5m). In the lower Citarum Basin sea-level rise could result in the inundation of about 26 000 ha of ponds and 10 000 ha of crop land. This could result in the loss of 15 000 tonnes of fish, shrimp and prawns and about 940 000 tonnes of rice.
The overall effect would be to reduce potential average income. The estimated reductions of yield would cost the rice farmer US$10.50 to US$17.30 annually, the soybean farmer US$22.0 to US$72.00 and the maize (corn) farmer US$25.50 to US$130.00 annually. It is estimated that the decrease in yield would cause, in the Subang District alone, about 43 000 farm labourers to loose their jobs. In addition more than 81 000 farmers would have to look for other sources of income due to the inundation of their rice fields or prawn and fish farms due to sea-level rise.
In Malaysia the projected warming under the GISS 2 x CO2 climate is equivalent to about a 3-4deg.C increase in mean annual temperature. There is no significant change in the seasonal pattern of rainfall, but increases in rainfall are projected, firstly, for the coastal regions of Sarawak in January and February, and secondly for southwestern Peninsular Malaysia in the intermonsoon period during March, April and May. Absolute levels of air humidity are projected to increase owing to higher rates of evaporation but, since air temperatures are also higher, relative humidity is not expected to alter greatly. No information is available concerning possible changes in the daily, seasonal and annual variability of climate.
Three sectors were chosen for the study of potential impacts from changes in climate, namely agriculture, water resources and coastal. For the agricultural sector, the production of rice, maize, oil palm and rubber were studied.
Potential impacts on rice production were studied in the region of Muda in the coastal plain of Kedah and Perlis, this being the largest rice growing area in Malaysia (126 000 ha producing 700 000 tonnes annually from two crops a year). The CERES Rice Model was run for current climate for 17 years (1968-84) and for the GISS 2 x CO2 climate for 17 years. Under the altered climate the maturation period for rice is shortened by several days, with consequent reductions in yield from 12 to 22 percent and with the largest reductions in the main-season crop. Under the higher temperatures the increased demand for irrigation is 15 percent.
The study concluded that the implied reductions in rice yield would significantly affect levels of farm income. Since farms in the area are already small (averaging 1.4 ha), farmers might be forced to seek alternative income sources with the poorest perhaps relinquishing their land and average farm size increasing as a result. In general, levels of rural poverty might be expected to increase.
Although it is not easy to generalise at the national level from the Muda study, it does appear that there would occur a nationwide increase in the demand for irrigation for rice, that water would be limited as a consequence and that the practice of growing rice two times a year would need to be limited to a much smaller area than today. The effect on national rice output could be substantial. The study concludes that Maiaysia's production, which currently satisfies about 60 per cent of demand, might be reduced to less than 50 percent.
The potential effects of the projected changes in climate on yields of maize were examined in the Serdang region. The crop is not significantly sensitive to small increases in temperature and rainfall, but is more substantially affected by changes in solar radiation. A 10 percent decrease in radiation causes a reduction of total biomass production by as much as 20 percent.
Increases in rainfall in the months of March, April and May could increase oil palm productivity in the alluvial coastal areas of Malavsia provided that solar radiation remains unaltered. But exceptionally high rainfall could limit yields unless soil drainage was improved. Overall, the study concludes that oil palm yields on alluvial soils would not be substantially altered under the projected changes of climate, but these conclusions are not applicable to oil palm plantations in inland regions of sedimentary soils.
Assuming an increase in mean annual temperature by 2deg.C and in rainfall by 10 percent, the east coast of Peninsular Malaysia may become too wet for rubber cultivation, with rainfall interfering excessively with tapping; and regions which are currently near the dry margins of current rubber production may become too dry. Overall, the study concludes that potential yield levels might, as a national average, be reduced by about 15 percent, but that improved clones could more than compensate for this by enabling a 25 to 50 percent yield increase. At present total rubber production is valued at M$3.6 billion. The potential effect due to climate-related yield reductions could amount to a quarter of this. The impacts are likely to be most strongly felt by rubber smallholders. There is already a trend toward smallholders selling out to larger estates, with consequent outmigration to urban areas. Climate change would be likely to aggravate this problem.
The possible effects of climate change on water resources was studied in the Kelantan River Basin in the northeast corner of Peninsular Malaysia. Under the higher levels of rainfall projected under the GISS 2 x CO2 scenario and assuming current patterns of rainfall variability, the frequency of occurrences of peak discharge increases by 9 percent, implying more flood damage and a 5 percent increase in the size of the flood-affected population. In contrast, higher rates of evaporation would lead to approximately a one-third increase in water deficit in the dry season, resulting in shortage of water for irrigation, reduction of rice yields and a reduction of the cultivable area for rice.
The effects of sea-level rise could also be severe. Much of Malaysia is characterised by low lying coastal plain. In these areas topographical surveys have indicated that a 1 m rise in mean sea level would, on average, lead to a landward retreat of shoreline of about 2.5 km. This implies substantial losses of agriculturally productive land, of mangrove forests and of the fisheries associated with mangroves.
In Thailand the warming under the GISS 2 x CO2 climate is equivalent to a 3deg.C to 6deg.C increase in current mean annual temperature, a projection that is broadly in agreement with other GCMs. There are, however, substantial differences between GCMs concerning changes in precipitation, which vary widely from normal but generally show a reduction under the GISS 2 x CO2 scenario. Northern Thailand may be drier in most of the months except in July which is currently a dry period and this would appear to benefit cropping. However, August and September would experience only between 73 percent and 89 percent of present rainfall. Other GCMs however do not indicate such a reduction in rainfall and it is important to emphasise this uncertainty. Under the GISS 2 x CO2 scenario winters are also drier but as very little rain is normally expected during that time of year the adverse implications may be less.
Two particular aspects of the Thai economy were studied with respect to potential impacts from these projected changes in climate: effects on rice production in Ayuthaya Province and effects of sea-level rise in Suratthani Province.
The CERES model was run for a 25-year set of daily climate variables (1964-1988). Model outputs for the current climate substantially exceeded observed values for transplanted rice and were lower than expected for yields of direct seeded rice. It was not possible, however, to conduct an adequate validation of the model and to re-tune it to observed data for Thailand. As a consequence, the results should be treated with caution.
The results indicate that under a change of climate projected for a doubling of CO2 main crop rice cultivation in Ayuthaya Province would increase in the order of 8 percent. These benefits would however be, in most cases, quite marginal because they are substantially less than the existing year-to-year variation. The modelled yields were also characterised by marginally greater yield variations. Off season rice, planted from mid-December to early February, exhibits a 5 percent increase in average yield under the GISS 2 x CO2 climate with concurrent increases in variation of 3 40 percent. However, little value can be placed on these results because of lack of model validation. Indeed, the results are not consistent with those for Chiang Mai which were validated against observed data, and which indicate a decrease in rice yield of about 5 percent under the GISS 2 x CO2 scenario.
Thailand has approximately 2940km of coastline, much of which contains important economic activities such as shrimp farming and rice farming. The study considered the potential impact of a 0.5 m and 1 m rise of sea levels in the Suratthani Province in southern Thailand. This region is characterised by a sand dune line which may mark an ancient shoreline and has a consistent elevation about 1m above present sea level. It was therefore used as an indicative boundary to the area potentially affected by a 1m sea-level rise. The suggestion is that 7400 ha (37 percent) of the study area would be affected by inundation under a 1 m sea-level rise. About 4200 ha of productive agricultural land and large numbers of shrimp ponds would be lost.
An integrated survey was made of the potential effects of sea-level rise on coastlines of Thailand, Maiaysia and Indonesia. In general all three countries have a high proportion of coastal plains both sandy and swampy that would be physically and ecologically sensitive to sea-level rise The issue may have particular importance because some of the coasts of South-East Asia are at present characterised by land subsidence which may be contributing to a more frequent occurrence of flooding, for example in the Bangkok region and in the coastal suburbs of Jakarta.
Where the coastal zone is characterised by steep and cliffed coasts they will tend to be undercut to form basal cliffs and slumping will become more frequent on the vegetative slopes. Receding cliffs are already a characteristic of the more exposed shores of promontories and islands especially those washed by ocean swell and waves generated by the SW monsoon as on the Andaman sea coast of Thailand and in south-west Sumatra.
On sandy beach coasts sea-level rise would tend to initiate beach erosion, or accelerate it where it is already taking place. In general, submergence will result initially in the deepening of near shore water so that larger waves break upon the shore, thus increasing erosion. It should be emphasised that it is not possible to predict the location of a sandy coastline in the next century if sea level has risen 1m but it is likely that most sandy coastlines will have retreated and that they will be eroding. Beach resorts and tourist facilities that have been developed extensively on low lying coasts in South-East Asia will be threatened (e.g. on Bali in Indonesia, at Port Dickson, Penang and Kuala Trengganu in Malaysia and at Pattaya and Rayong in Thailand). Structural work such as concrete sea walls and boulder ramparts will be necessary to protect developed seaside land, but such structures usually result in wave reflection which depletes the beaches that were the original tourist attraction. Artificial beach renourishment is an expensive alternative (about $3m/km) and may only be feasible in intensively urban resort areas such as Pattaya in Thailand, the north coast of Penang in Malaysia and the resort beaches on Bali.
There are large areas of swampy lowland on the coasts of South-East Asia, especially on the shores of the deltas built where large rivers have delivered vast amounts of silt and clay to prograde the coast. These are very extensive on the north-east coast of Sumatra and the south coast of Irian Jaya. Sedimentation from rivers is still prograding deltaic areas. Such deposition will accelerate if rainfall and runoff from the river catchments is augmented as a consequence of global warming. However, a rising sea level will tend to curb the growth of deltas and if the rate of submergence is greater than the rate of deposition, their shorelines will be cut back. Examples of this can already be seen on parts of subsiding delta coastlines which are receding because of a diminished fluvial sediment supply to the river mouth. On the north coast of Java some river mouths have changed naturally, during episodes of flooding, to a new outlet for subsequent delta growth. The outcome has been rapid erosion of the abandoned delta lobes.
The natural vegetation associated with the coastal lowlands of South-East Asia is mangrove swamp, backed by marshes and areas of freshwater forest, but in many areas this vegetation has been profoundly modified by human activities, notably drainage and land reclamation. In the three countries studied mangroves occupy a total of about 40 000 km2. They grow in the upper part of the inter-tidal area, usually near mean tide line. Mangroves have become extensive on these coasts during the past 6000 years when sea level has remained constant. Prior to this there was a phase of rising sea level beginning about 18 000 years ago. The projected future sea-level rise could reverse this sequence, reducing and removing mangroves from the more exposed areas and confining them to inlets and estuaries where continuing muddy sedimentation, keeping pace with the rising sea, might allow them to survive. Under natural conditions mangroves are backed by low lying, estuarine and alluvial land which could be displaced if the rising sea drove the mangrove zone landward. However, over much of South-East Asia the hinterland has been reclaimed for agriculture, usually rice farming or plantations producing rubber, palm oil or coconut; and embankments (bunds) have been built at the inner margin of mangroves. Where there is a bund at the rear of the mangroves delimiting the present high tide limit, attempts would need to be made to maintain it as sea-level rises and to enlarge it to prevent wave overtopping and marine flooding. If this happens the retreating mangrove fringe will not be able to colonize the developed hinterland. It will become narrower and in many places will disappear altogether as the inter-tidal zone narrows and steepens. The coastline will thus become more artificial.
Extensive areas of mangroves have also been converted to ponds for the production of fish or prawns. The simplest ponds (traditional for many centuries on the north coast of Java) are located in banked areas with sluices to prevent the gravitational inflow and outflow of sea water and the entry of fish and prawn fry. In recent decades fish and prawn ponds have become more elaborate, especially in Thailand and Malaysia, with pumping systems to maintain a sea-water supply and the use of aerators, breeding techniques and fertilisers to generate high productivity from intensive aquaculture. Where the mangrove area has been converted to aquaculture, the sea-level rise will threaten to breach the enclosing banks and submerge the fish and prawn ponds. If these ponds are to be maintained the enclosing walls and the floors will have to be raised to match the levels of the rising sea. Alternatively a protective sea wall may be built and pumping systems introduced to control the inflow and outflow of sea water.
In addition to these direct effects, marine submergence of coastal areas in South-East Asia will raise the water table so that some low lying parts of coastal plains will become permanent swamps or lakes. The salinity of these will depend upon the interaction between sea water incursion and any increase in rainfall and freshwater runoff. It is possible that the rise in groundwater will be accompanied by the upward movement of subterranean salt, resulting in saline damage to rice fields and farmland soils. It may be tempting in such conditions to convert these areas into brackish-water, fish and prawn ponds to replace those threatened or lost in the mangrove areas.
Coral reefs occupy about 150 000 km(2) within the East Asian seas, and there are fringing reefs around many headlands and high islands. A slowly rising sea level will stimulate the revival of coral growth on reef flats and these may maintain their level relative to the rising sea. Upward growth of existing corals is in the range of 4-7mm/year and thus may be able to keep place with rising sea levels. However, as indicated above, many coral reefs are already under various kinds of ecological stress, and some of the less vigorous may fail to revive, being permanently submerged as sea-level rises. Fringing reefs are less likely to survive than outlying reefs because of increasing turbidity in coastal waters as larger waves erode the beaches and the land behind them.
It will be evident that some environmental changes are already in progress on the coasts of South-East Asia, and that substantial modifications, both natural and man-made, would have occurred on these coasts during the coming century even if there were no global warming and sea-level rise. Coastal erosion is already extensive and likely to continue, and coastal environments will be changed by further urban and industrial development. The combination of such pressures, together with possible future sea-level rise due to global warming suggests a number of possible strategies:
In the short term, over the next few decades, the wisest response to the predicted sea-level rise is likely to be a reorganization of coastal land use planning in low lying coastal areas, delimiting these areas in relation to predicted submergence and erosion. For example, it is unwise to develop new resorts within 200m of the present high tide line on beach-rich terrain unless plans allow for abandonment or relocation during the coming century. Aquaculture could be restructured towards intensive production from relatively small and concentrated areas which can be protected from submergence and erosion as sea-level rises and can be adapted to new tidal levels
The study included two workshops (in Malaysia and Indonesia) to consider the policy implications of the reported results. The workshops were designed to inform policy makers about the magnitude and characteristics of potential future climate change, to consider the range of possible response strategies to mitigate adverse impacts and to outline the need for future research.
The workshops were conducted as policy exercises, bringing together policymakers and their scientific advisors at the national government level and scientists who had worked on the study to generate the impact assessments. Five major types of policy response were considered at these exercises: economic (changes in existing tax structure, subsidies, pricing systems, etc.), technological (breeding new varieties, constructing dams and coastal protection structures), institutional (enhanced or distorted market mechanisms, formal government regulations, legal instruments), research needs (information required for formulating adequate response strategies), and monitoring (characteristic signs of change, both biological and socio-economic, that could provide the necessary early warning to ensure timely action).