Annex VIII
APDC/01/4

RIVER BASIN MANAGEMENT FOR FLOOD AND DROUGHT PREVENTION AND MITIGATION IN ASIA AND THE PACIFIC^* (Cont.)

4. STATE OF THE ART

4.1 Flood Alleviation Strategies

85. Such strategies have evolved in tandem with those on water management and development, and with a mix of successes and failures. In the less-developed regions of the world, the lack of access to resources and education and low income contribute to a population's vulnerability to floods. They and other hazards need to be addressed systematically by stimulating social and economic advancement, rather than merely implementing alleviation schemes. Emphasis should be placed on a number of important areas, namely, catchment, coastal zone and floodplain management, giving local communities options regarding land development and flood alleviation, strengthening their resilience to disasters, improving their capacities to respond, preserving worthwhile indigenous technology and reducing the impact of humans on the environment.

86. Different strategies are required to tackle different types of floods. Riverine ones, e.g., are a feature of large river basins. Land-use and its control by zoning are important aspects of mitigating them and a key component to the overall integrated watershed management programme within their floodplains. Flash floods can be tackled by prohibiting affected lands from village occupancy. For urban flooding, the provision of adequate waterways and run-off detention storage sites, and prohibition of residential development in highly inundated parts are some of the measures commonly adopted. In developed urban floodplains, where the risk is extremely high, land-use planning and control are vital. Coastal flooding is location-specific and thus requires similar impositions, especially limiting or banning development in areas susceptible to storm surges.

87. Floods and droughts should not be managed separately. Both are natural events that bring too much or too little water. They occur in a continuous cycle, and it is common for the former to follow the latter. Multi-purpose measures are advocated rather than mere prevention and crisis management alone. River basin management may encompass changes in lifestyle patterns and industrial structure, corrections of any environmental imbalance and improvement schemes for rivers and land use. Many major structural flood protection projects have been completed, particularly on lowland rivers. However, raising embankments to reduce damage to floodplains is not the best option, while non-structural protection will grow in importance as part of sustainable river management.

88. The return period should not be used as a design standard to predict floods in any given year or a number of years. Instead, the annual probability of occurrence should be employed to ascertain their severity, and flow and level predictions expressed in probable bands rather than specific values. It is necessary to take into consideration the failure of a particular strategy and to seek solutions to alleviate or mitigate rather than protect or control. This means that any flood management strategy will have to involve a combination of different approaches rather than rely on a single one. Furthermore, reliable flood estimation is a function of the length of the stream flow gauging record, i.e., the return period that is about twice the length of the record. A 100-year flood is usually more severe than its magnitude, which probably can be predicted with greater accuracy. The economic objective is to maximize efficient use of the catchment and not to minimize flood loss. National trends on the latter do not necessarily form the basis for evaluating the success or failure of management strategies adopted by a nation. They can indicate that an efficient management policy can be accompanied by a rise in both flood loss and the cost of flood management.

89. It is important to adopt a catchment-based approach and misleading terms such as floodplain encroachment should be avoided. The question that should be asked is whether the development of a particular area of the floodplain should be encouraged, and if so, what form of management strategy should be adopted? Floodplains have competitive advantages over other areas of land, and therefore have always been centres for human settlement. Also, wetlands that develop within them contain the most valuable ecological resources on this planet.

90. Management of floods and their design are learning processes. It is essential to identify the objectives and assess the options, before designing a project and finally evaluating its economic, environmental and distributional impact (Table 5). Choices are made under uncertain conditions, due to the lack of differentiation between outcomes and/or the probabilities of each outcome. Risky decisions can usually differentiate between the probabilities and the consequences associated with each option.

91. Land use control unfortunately rarely works. The provision of incentives for development elsewhere would probably be more effective than simply attempting to stop settlement on the floodplain. Where land is heavily populated, especially under informal circumstances, imposing constraints are unlikely to be successful. Reforestation does no necessarily reduce flood risk, but forestry practices can be extremely worthwhile. Trees with high evapo-transpiration capability can be planted to lessen inundation and lower annual runoff, but this effect wears off as the forest matures. Deforestation tends to decrease evapo-transpiration and runoff concentration and thus increase annual runoff and flood peaks.

92. Modification of the water regime, a physical mitigation measure, would damage the existing ecosystem, while abandoning or changing past patterns of human intervention could cause environmental damage. Rivers are dynamic and adaptive systems, whose form varies as runoff and sediment loads change over time. They erode, transport and deposit sediment as much as they transport water. Any intervention measure must take into account these characteristics and attempts to stabilize rivers are usually an expensive exercise with little advantage (Table 6).

93. Probably no proven approach exists to manage the rehabilitation of watersheds, as they differ geographically, climatically, economically, socially and politically. Ample participatory methodologies for land resource planning have been developed at the country level, but they must be modified to suit specific situations. Publications, international texts and conventions mentioned in previous sections of this paper provide useful guidelines. There is also a wealth of information and literature on the socio-economic and technical aspects of land and water management (Lal, 1995). However caution is needed, as not all of them are based on solid scientific evidence. Thus a careful review of the exact circumstances and conditions of a particular technique and its effects are recommended, prior to adoption.

94. The impact of land use on water resources is clearly visible everywhere. Groups of upstream and downstream stakeholders are few and well organized. Thus the economic impact on them can be quantified. The benefit-sharing incentives to both groups of resource users are high enough to prevent them from resorting to other measures to solve their problems. There is thus justification for political commitment to establish upstream-downstream linkages within a strong institutional and legal framework, which allows for the implementation of benefit-sharing schemes that integrate surface and groundwater resources.

4.2 Surface and Groundwater Management and Accounting

95. It is now accepted that surface water and groundwater are interdependent and need to be considered and utilized together as a renewable resource. Influent and effluent streams, infiltration and interflow, recharge due to river flow and re-emergence of streams at lower stages of the river are manifestations of such interdependence. Closed systems require increasingly efficient and effective management of both surface and groundwater water resources, and the ability to allocate and reallocate water to accommodate changing demands and priorities (Seckler, 1996).

96. One of the major advantages of storing water in underground aquifers is that it can be kept for years, with little or no evaporation loss. It can thus be used as a supplementary source of supply during drought years. It has also the benefit that storage can be near or directly under the point of use and is immediately available through pumping. The tube well revolution that has swept through agriculture capitalizes on these advantages. Another benefit of groundwater is that it is usually purified of biological pollutants as it slowly percolates down to aquifers. Thus it is the best source of drinking water, especially in rural areas of developing countries where water treatment facilities are not available. Increasing storage through a combination of groundwater and large and small surface water facilities holds great promise. Another example is the artificial recharge of groundwater (Kowsar, 2000) developed in Republic of Iran. It not only replenishes the empty aquifers on the debris cones, but also builds soils that can be used for growing food, feed, fibre and fuel-wood. This new approach to soil and water conservation in Iran's deserts has tremendous potential. The critical issue facing many groundwater aquifers today is that the volume of withdrawal exceeds long-term recharge, resulting in rapidly declining groundwater levels in many areas.

97. New and improved water accounting procedures developed at the Integrated Water Management Institute for assessing basin-level use and productivity (Molden, Sakthivadivel and Habib, 2001) offer powerful concepts and tools for identifying ways to improve productivity. Its accounting system maps outs the quantity and productivity for various purposes within a basin. This information is employed to identify the potential for conservation, and the means to increase the volume of managed supplies.

4.3 Water Management Strategies for Agriculture

98. Rain-fed agricultural areas, including those in developing countries, receive less generous and reliable natural water supply, and yields tend to be 25 to 35 per cent of their potential. Given an adequate holding size, such agriculture is able to provide a reasonable livelihood to farmers. Where rainfall is unreliable, improved farming techniques such as land levelling, ridging, etc. that increase water intake can improve the soil's retention capacity and yields. Practices that reduce non-beneficial evaporation such as mulching can have similar effects.

99. Water harvesting³⁴ lies somewhere between rain-fed and irrigated agriculture. It concentrates runoff from a larger land area for use in a smaller one. The objective is to capture as much water as possible and store it in the root zone. If this amount exceeds the quantity that can be stored, it results in recharging the groundwater. Although the area affected by rainfall harvesting is relatively insignificant, it can have a noticeable impact on regional food supplies and incomes.

100. Full irrigation is essential when planting during the dry season in arid, semi-arid and monsoon regions in the humid tropics. Major physical infrastructure is necessary to capture, store, transport and deliver the needed supplies. Such major works are generally managed by public agencies. Full irrigation can also be carried out on a less massive scale, using smaller water sources, wells, pump sets or alternative sources such as domestic wastewater. Farmers have developed and managed these moderate schemes according to their own set of values and rules in the past, and their modus operandi will continue to endure in the future. They adopt technologies that suit their financial and managerial capabilities.

101. Supplementary irrigation often increases yields and reduces the risk of crop failure where, under normal circumstances, reasonable yields can be expected. Farming in rain-fed terrain can be made more productive, cropping seasons extended, production failure caused by short periods of dry weather minimized and productivity expanded by irrigation during critical crop growth stages when water supply is temporarily inadequate. When dangers are reduced, farmers will be encouraged to step up the level of inputs utilized. The overall benefits can thus be very extensive.

102. Much of the existing agricultural land in both rain-fed and irrigated regions still suffer from poor drainage and/or excessive salinity. In Southeast Asia, large tracts experience severe water logging and flooding annually, due to monsoon rains and tropical storms. Irrigation in the absence of adequate drainage has exacerbated natural drainage problems. Its improvement can, however, lead to important production advances by restoring healthy root moisture and encouraging the application of advanced agronomic practices such as high yielding varieties, fertilizers, and mechanization. Drainage has also contributed to agricultural intensification and diversification and made the sector more viable.

103. The recent recognition accorded to the value of natural wetlands has caused its reclamation to come to a virtual standstill. There are significant drawbacks to improved irrigation. Its large-scale use upstream increases the possibility of flooding and consequently damage, as well as shorten the length of time of water in the catchment and the recharge of groundwater. The resulting fall in the latter's water table may have serious consequences on the ecosystem and bio-diversity. Thus high groundwater tables have to be maintained.

5. LIMITATIONS AND DIFFICULTIES

5.1 Complexity

104. The management of hydrological systems is a complex undertaking because of the numerous and simultaneous processes involved, their spatial variability, the spatial heterogeneity of watershed characteristics and scale effects (Kiersch, FAO, 2000). Some of these are intrinsic to the watershed itself, while others are related to hydrometeorological events. A watershed comprises a hydrographic network and a set of hill slope elements that drain water downstream. The characteristic scale of hill slope river flows is usually shorter. Moreover, most threshold effects appear within them. There is a major hiatus effect between these two sub-systems. The watershed appears to be structured, where the hydrographic network is a key point, and this is extremely important for short-term considerations such as flood management.

105. Hill slope characteristics play a major role in a watershed, as their patterns fit in with the hydrographic network. The watershed is small enough to give them importance in the entire temporal scale. Rainfall variability also complicates matters, as it interacts with the hydrographic, geometric and dynamic organization, and with the hills lope pattern.

106. Farmers or resource users tend to encounter some kind of environmental, economic and/or social risk. These are summarized in Table 8 for Central Asia. An outcome-oriented approach to classifying risk is to examine its impact on those affected. The dangers to individual households (individual risk), or all households in a given area (covariate risk) have fundamentally different consequences and thus call for different responses (Swift, 1999).

107. Managing disasters and minimizing agricultural vulnerability (FAO, 2001) refer to actions taken prior to, during, and after a disaster. Measures to mitigate the agricultural impact of storm-related catastrophes are required for all stages of an emergency. Greater attention needs to be placed on pre-disaster efforts that strengthen the resilience of agricultural production and rural livelihood.³⁶ As countries, communities and population sub-groups are prone to various types of natural calamities simultaneously (e.g. storm and flood) or subsequently (e.g. drought and flood), the remedies to arrest them must be integrated with those that address other kinds of disasters.

108. Traditionally, two very different categories of "actors" are involved in developing management approaches, those with a "resource focus" that leads to the development of a catchment on an "integrated water resource management (IWRM)" scale, and those with a "people focus" that emphasize "sustainable livelihoods" on a village or household scale. Each approach has its limitations. The former may not be taking into account impact at the village or household level, especially when they affect the poorest in society, while the latter, in the absence of any resource regulation, may lead to resource constraints and depletion in the catchment (Calder, 2000).

5.2 Problem of Scale

109. Recent literature (Kiersch, FAO, 2000) reviewed the impact of land use changes on downstream hydrology in the catchment area. Two conclusions are drawn from it; the impact of certain types of land use can extend beyond the field or plot to affect downstream users; and, in general, land-use impact on water resources are only measured for basins of up to a few hundred square kilometres. The size of the river basin is critical for effective management. Table 9 gives the optimal scale of meso-basins within well-defined state or national jurisdictions as 100-500 km². Benefit-sharing arrangements are unnecessary when the land use impact does not extend beyond the farm plot. Clearly, the relevant scale(s) will differ with regard to the type of impact. Also, environmental conditions (climatic, topographic, socio-economic, etc.) may determine the scale of impact.

110. This confirms ESCAP's statement (1997) that when the run-off behaviour of a watershed is within the hydrologically small category, the implementation of appropriate forms of land use can be particularly effective in mitigating floods. On the other hand, if a catchment is considered hydrologically large, the effectiveness of such measures to minimize major flood disasters is severely limited, especially at the lower reaches of the basin and on the floodplain.

111. This is not meant to suggest that land-use management on the upper watershed should not be undertaken. Such management is, in fact, valuable for direct flood mitigation and may have a range of other advantages, including preserving the integrity and productivity of the catchment soil, its vegetation, wildlife habitat, quality of its ecosystem and environment, besides improving the standard of living of its community. It may not reduce major flood peaks substantially further downstream, but it may reduce catchment and stream bank erosion and transport of sediment to lower reaches. It would be necessary, if only to maintain or improve water quality throughout the entire river system.

112. Land use impacts on hydrologic regimes and sediment transport are most obvious on small spatial scales. At the same time, the number of water users who might readily benefit or suffer from this change in land use, increases with the size of the watershed. Due to the decreasing magnitude of its impact, the respective costs and benefits will be small. Effects on water quality, such as salinity and pesticide pollution, however, are equally relevant in medium-to large-scale river basins because of nutrient influx. These may affect downstream uses, including those providing drinking water to industries, fisheries and other agricultural uses.

113. Large projects in upland areas such as afforestation programmes, should be implemented to improve the ecological conditions and livelihood opportunities in mountainous terrain. Further efforts are needed to clarify or eradicate misconceptions regarding land-water linkages to facilitate political discussions on international river systems and assist mountain communities develop or maintain sustainable land use strategies and practices. A number of massive reforestation programmes are unfortunately still based on the erroneous idea that flooding problems in the plains are caused by deforestation in upland watersheds.

114. Problems of scale have intriguing consequences. The solutions depend on the scale at which one envisages a problem. For instance, its relationship to the consequences of over extracting groundwater has substantial environmental, economic and other implications (FAO, 2001). Focusing solely on macro food security concerns could divert attention from groundwater issues that are important in their own right. Viewing over-exploitation of groundwater from a global food security perspective would mean focusing attention on world and national efforts to manage and control the use of this resource base. If, on the other hand, it is regarded more as a regional concern, then the efforts made to accommodate the scarcity by moving out of agriculture, could be considered economically viable and justifiable.

115. The temporal scale of land use impact (Kiersch, 2000) is important as it determines the perceptions of such an impact as well as its economic cost. Two aspects of this are crucial; firstly, the length of time it takes for a particular land use to have an impact downstream and, secondly, in negative cases, the timeframe for remedial measures to work. The temporal scale of such impact varies widely, depending on the consequences, which may range from less than a year, as in the case of bacterial contamination, to hundreds of years as in the case of salinity. Similarly, the time scale for recovery from adverse impacts is very diverse, depending on their nature. However, in most cases, the time span needed to restore an aquatic system following an adverse impact is much longer than the period taken for an impact to appear.

116. Time-scales are of great value for observing and understanding phenomena associated with changes in land use, particularly vegetation cover and water management. Deforestation is less extensive and complete than is generally imagined. Hydrologists and climatalogists (Reynolds and Thomson, 1988) are rarely conversant with the social and economic reasons that govern this activity, and with the complete range of resultant vegetation types, their uses, or the degree of impoverishment. These two problems could be remedied by investigations that give a historic perspective of changing land use in areas selected for their hydrological and climatological significance. This is crucial as forest succession and age impinge on hydrology. For example, the contrast that can be observed from the pronounced changes in water yield following the first year of deforestation, and the subsequent, often marginal effects (rarely equal to 20 per cent of the original value), when a new vegetative cover is established. Forest clearance brings on an immediate, short-term increase in discharge. Subsequent tree growth produces well-known and important alterations to surface hydrology. For natural tropical forests, rapid growth rates mean that any analogous effects would operate over greatly reduced time-scales.

117. The heterogeneity of important hydrological parameters and processes make extrapolation difficult (Reynolds and Thomson, 1988). The number of classes of soil, vegetation, or land use often increases with scale. Perceived spatial heterogeneity is thus intimately linked with scale. For instance, local soil moisture measurements cannot be applied to the entire catchment because of the small-scale variability of soil and a poor understanding of runoff characteristics. Thus simplistic models for both these processes have to be adopted. Although widely used, they are poor predictors of macro-catchment behaviour. Ideally, appropriately scaled models should be utilized for extrapolating to a larger scale. Unfortunately, the spatial and temporal variability of soils and hydrology cause risks to be very much conditioned locally (Burke, 2000), that any generalization masks the root causes of degradation. These have largely to do with the failure to recognize the broad hydro-environmental dangers that are associated with the modification of landscapes.

5.3 Problems of Uncertainty

118. Data on water balance and productivity for irrigation schemes at basin, system and farm levels are scarce and when reported, the method of derivation is not described. Such inadequacy makes it difficult to analyse its productivity (FAO, 2001). The situation as shown below is no better for groundwater resources³⁸ (FAO, 2001).

119. The available groundwater database contains a wide range of inherent limitations, including inadequate records and limited yearly monitoring. This constrains any analysis that can be conducted on the magnitude and extent of water level fluctuations. Experience in locations such as the San Luis Valley in southern Colorado indicates that substantial uncertainty³⁹ remains in water balance estimates despite decades of detailed monitoring and the application of appropriate models following intense discussions on water management.

120. Water balance estimates shed little light on whether users in a given region are likely to or actually face problems of scarcity. Available basic data on levels and fluctuations are generally inaccessible. Groundwater information is politically sensitive and thus subject to pressure. Consequently, objective evaluation on the extent of groundwater depletion is lacking.

121. A wealth of resource-based information exists in most countries, but access to or use of it is often difficult. It is held by different organizations and, in some cases, by various departments or individuals within them. Spatial and non-spatial data is stored in a wide range of formats; in maps, remote-sensing images, tables, texts, yearbooks, research papers and computer discs. The spatial and temporal scales of data collected differ enormously and its quality is also extremely variable.

122. Many erroneous explanations of phenomena arise because the models used only give partial understanding of the systems involved (Ives and Isserlé, 1989). Although the physical elements underlying the models are accurate, they have little predictive value. Phenomena also depend on complex interactions between the water resource and social and economic forces that are not fully understood. Even when there is comprehension, interactions between non-linear systems often produce unpredictable and counterintuitive results. In addition, the data available for validation is frequently insufficient.

123. An example is the Ordos Plateau in the Peoples Republic of China where conflicting interpretations cannot be resolved. The plateau is a major area for sandy deposits from the Yellow River. Serious desertification and erosion add coarse, sandy deposits to the silt discharge of this river (United Nations University, 1995). It is estimated that the Ordos accounts for 10 per cent of such discharge and transports one-third of the coarse sand for the Yellow River. In short, environmental changes in this region affect deposits in the Yellow River, thereby endangering its lower areas. Disputes over the causes of such changes throughout history are apparent. Some scientists argue that natural conditions have played a dominant role, while human activities made only a slight impact. Others believe otherwise.

124. The above observations are not intended to discredit modelling efforts. They are useful tools for explaining the physical data and interpreting the results of management changes. When properly developed and calibrated with adequate data, they can provide valuable predictions. The basic scientific information necessary to structure models is, however, lacking in analyses carried out in India and many developing countries.

125. Natural resource managers have to deal with strong economic and political forces. They must often make decisions without complete knowledge of complex natural phenomena. Therefore their strategies can be challenged from time to time, as inadequate knowledge prevents accurate analyses of the effects certain activities can have on watersheds. Thus more emphasis must be placed on monitoring such impact. Table 10 shows the degree of uncertainty in which managers or decision-makers operate, especially when compelled by necessity to make crucial decisions.

126. Very few managers can, in reality, follow the guidelines recommended for disaster preparedness and mitigation to arrive at decisions in a risk-ridden situation. They are lucky if they can act under unsure conditions. Managing uncertainty and, whenever possible, either reducing it or taking it into account by avoiding strategies that cannot adapt to change or unpredictable forces seem particularly "risky" to begin with.

5.4 Limits to Flood or Drought Alleviation Strategies

127. Box 7 describes what is now classically known as the turn of the water screw. It explains how responses to perceived water scarcity are becoming increasingly complex. The inability of societies to adapt to shortages can be termed social water scarcity (FAO, 2000). The definition of drought is relative to demand, and is therefore ambiguously related to the notion of scarcity. However, the idea of successive turns of the screw means that, after a time, mitigation and recovery measures, and resilience are insufficient to address forthcoming scarcity, which may take the from of droughts of increasing frequency and/or occurrence. The danger is that, at a time when society must be ready to change drastically (the third turn of the screw), it is still "comfortably installed" in what appears to be a sound adaptation phase.

128. It is recommended that disaster mitigation strategies, particularly for drought, should incorporate long-term adaptation measures, and not merely designed to cope with the phenomenon. ESCAP⁴¹ (1999) suggested that long-term strategic planning based on prospective studies are particularly relevant.

5.5 Myths

129. Decisions on land use and water resource are often still based on perceived wisdom (myth) rather than on established scientific reality (Calder, FAO, 2000). This is especially the case with the relationship between deforestation, reforestation and afforestation. Some of the "myths" (listed below) have recently been reviewed at a FAO e-mail conference on land-water linkages.⁴²

130. Forests increase rainfall. There is some kind of relationship between land use controls and precipitation, but it is much less intense than imagined (forest by grass or secondary forest), although it cannot be totally dismissed. The distribution of forests is a consequence of climate and soil conditions and not the reverse.

131. Forests increase runoff. They generally decrease runoff (evaporate- and transpire more) compared with areas under shorter vegetation. There is a need to know precisely the soil type/forest cover combinations for greater accuracy.

132. Forests regulate and increase dry season flows Such need not necessarily be the case, as competing processes may result in either increased or decreased dry season flows. Effects on them are very likely to be location specific.

133. Forests reduce erosion. Man's influence is limited in areas of high natural sediment that are caused by steep terrain. Large leaves generate large splash erosion. Disturbed forests and forest plantations may result in either increased or reduced erosion. The latter does not guarantee the same erosion benefits as natural forests. Forest cover alone is inadequate; forest quality is perhaps a more important factor.

134. Forests reduce floods. Deforestation of the Himalayas is made responsible for floods in Bangladesh and northern India. However, hydrological studies have shown that there is little connection between land use and storm flows. Theoretically, this applies for small floods, but is not significant for large storms, reduced runoff and stork response. Scale approach is the key to an understanding of highland-lowland issues. Tree planting on hill-slopes, watershed activities and extensive soil conservation measures are valuable for the hill farmer, but potentially disastrous for foreign aid agencies and national governments to undertake such activities with the conviction that they can solve the problems on the plains.

135. Agroforestry systems increase productivity. Single cropping is better for low rainfall areas (less than 800 mm). Practices such as multiple cropping and agro-forestry that increase productivity, if successful, are likely to have downstream effects due to rising water usage. Evidence suggests that there is no difference to productivity with mixed cropping and the planting of trees. They generally lead to a reduction in biomass of any associated crop, and at best, no change in total biomass production can be expected. Recent research suggests that any decrease in crop yields due to tree competition will automatically increase the frequency of poor yield years and threaten food security: There is thus little gain in productivity by mixing trees with agricultural crops, and if there were, it would mean increasing the use of water.

136. Reforestation of dry land is useful to combat drought:In saline-prone dry land, replanting trees and shrubs in the catchment's recharge area increases evaporation and reduces recharge and groundwater levels. This is good for salinity control. The removal of forests lessens evaporation and saline water tables rise to produce saline seeps. Nevertheless, tree and shrub planting as part of watershed programmes is frequently advocated as the key to regenerating springs and river flows.

137. Shifting cultivators are the main initiators of erosion in (previously) forested hilly areas. This is generally incorrect, although a popular belief that leads to the identification of agroforestry as a solution to erosion problems.

138. The report (Calder, FAO, 2000) concludes that there is a need for greater cooperation among agencies responsible for making policy decisions, so that erroneous opinions do not interfere with the selection of the best approaches to improve water resources. If the objective of reforestation or forestation is to improve water resources, it is unlikely to be achieved.⁴³

5.6 Limits in Valuation Methods

139. There is a need to consider the costs and benefits of a well-preserved watershed, especially in terms of environmental goods and service⁴⁴ (Eschevarria, 2000). There is also a growing importance to effectively value water resource practices and policies that promote its efficient use. Appropriate methods are available to estimate the impact in terms of water quantity. However, they are only as good as the evaluation of real impacts/effects. Information gaps and contradictory evidence, time scale and intergenerational issues, the diversity of stakeholders and distribution of benefits and costs in watersheds make their valuation particularly uncertain in land and water management.

5.7 Limits in Institutions

140. When confronted with the problems faced by South Asia, where solutions require actions that often span across state, national and international boundaries, existing approaches to water management (such as those adopted for the South India Water Vision) run the risk of being inadequate and slow in implementation or adoption for various reasons. Some form of cross-border collaboration (such as allocation, and joint planning and enforcement of control structures) is necessary but difficult to achieve politically. Besides, it entails major environmental costs (such as water logging associated with embankments and the necessary diversion of in-stream flows to actively recharge groundwater on a large scale). Basic scientific data on basin and aquifer hydrology that is currently unavailable, but vital in Asia, has taken western nations numerous decades to develop. In addition, some form of cooperative action by countries to conserve water and reduce demand may take many years to attain but albeit essential. Major institutional restructuring of legal systems (water rights) and organizations (dealing with irrigation, public health and engineering, etc.) is required.

141. This is another variation of the time scale problem. Strategies must achieve results to avoid further degradation or distress, or at least to maintain it at bearable or reversible levels. Reforms in water policy and legislation, and the creation of new institutions take time to institute. In the interim, institutional analysis for the design flood and drought plans must continue (Helweg, 2000). Present institutions, however imperfect, will have to be retained to assist in future improvements (Green, Parker and Tunstall, 2000).

142. The larger the watershed, the more difficult it is to keep the interest of the general public. Upland dwellers find themselves too far removed to connect with catastrophes downstream. Organizationally, basin work for large rivers is the responsibility of governments. A fundamental reason for the lack of understanding of interactions between surface and subsurface hydrology is the fact that surface hydrologists tend to concentrate on land and surface water linkages, while hydro-geologists tend to focus on land and groundwater linkages. As these studies are usually carried out independently, a number of important interactions are omitted.

143. The sharing of the costs and benefits of watershed management is critical to its viability and efficient management. This is not confined to only upstream and downstream beneficiaries. There is a need for equal distribution within a community and within households. This entails long-term planning based on sound environmental principles that are immune to economic and political change. However, all production and development strategies have social, political and environmental implications. Strong participation may be a key element for the sustainability of river basin institutions, which are subject to general political events. Many water quality problems seem to become obvious only after a period of time. Due to this time lag, it is difficult or even impossible to retain the attention of the public for long. In addition, many formidable variables prevent effective management and protection of watersheds that deliver clean water to users.

144. The Haihe Basin (Burke, 2000) is subject to extreme hydrological cycles and inter-annual variability. This is characteristic of southern China's climate where flooding has brought massive human and economic burdens. They resulted in major capital investment in structural flood control at the state and local levels, but no similar effort was made to co-ordinate basin management during times of drought. Instead, reliance on groundwater was considered sufficient, without any attempts made to smooth out peak demands and shortfalls. The resulting impact on grain production is significant, but difficult to capture, since the effects are felt both in space and time across the basin. Between 1949 and 1990, the average area hit by droughts in the Huaihe River basin was 2.5 million ha, while the severely affected region amounted to 1.3 million ha, or 20.1 and 10.6 per cent respectively of the cultivated area (12.4 million ha).

145. Response to drought entails tackling issues regarding water allocation, its priorities, restrictions and utilization. Normally, the order of priority is domestic supply, followed by environmental demand, and then services/industries' requirements. Imposing restrictions on access or use is not as popular as protecting a user from flood damage. Priority rights can only be established after overall rights are in place.

146. In spite of growing concerns over groundwater, its extraction and past catastrophes, the creation of an institution capable of managing it has as yet to be realized. Global water management literature tends to emphasize the importance of complete organizational restructuring in the water sector and the development of "comprehensive" and "integrated" strategies. Calls to integrated water rights, markets or other frameworks that allocate limited supplies, and the formation of basin or aquifer management authorities with substantial regulatory powers are common. The focus is heavily on water resource management per se. Integrating such institutions, according to some authors, may be too narrow an approach to effectively solve even water problems.

6. CONCLUSIONS AND RECOMMENDATIONS

147. The development of integrated water management systems capable of addressing problems at a basin or aquifer scale is a long-term undertaking. They are unlikely to fulfil the immediate needs of populations living in drought and flood prone areas, and develop adaptive strategies for emerging water problems in South Asia. The latter require the role of institutions that are both flexible and systematic when identifying the main forces of change at a particular time, and to propose appropriate responses that may well be outside the spheres of management.

148. Managing hydrological risk involves not only coping with extreme events caused by climatic variability, floods and low river flows, but also dealing with the day-to-day flow changes, extractions, releases and pollution loads. Efficient management is required with increasing water scarcity and closure of river basins. The advent of large-scale surface water storage structures, mechanized boreholes, cheap fertilizers and pesticides in the mid-20th century may have given irrigated agriculture a false sense of security. The rigidity of resource management as manifested in many recent irrigation projects reveals an oversight of the inherent variability of natural systems. Such schemes can be sustained by spreading risk equitably and transparently among resource regulators, managers and users, but it involves greater flexibility in responding to management and taking into consideration natural parameters and socio-economic settings. This flexibility is attainable by modernizing institutions and distribution systems. A clear understanding of the risks and underlying mechanisms of land and water resource management is also warranted to make the case for the equitable and transparent spread of hydrological risks and associated costs.

149. The adoption of disaster alleviation measures and correct planning techniques during early stages of design are instrumental in ensuring that they not only improve the resilience and capacity of people to cope with short-term hydrological risks, but also contribute towards the preparation for growing water scarcity and other acute situations. A comprehensive knowledge of water resource systems is essential to move away from a situation of ignorance or at best uncertainty to one of calculated risk.

150. River-basin management strategies including those for floods and droughts can be facilitated by the adoption of proper accounting techniques and the continued development of reliable hydro-meteorological, agro-meteorological and groundwater monitoring services. They provide the necessary data not only for daily management and forecasting, but also for modelling purposes and strategic planning. The ability to cope with disasters is dependent on the flow of good hydrometeorological information from data collection agencies and institutions that conduct analysis, and finally on the capability of public bodies, authorities and communities responsible for implementing protection and mitigation measures.

151. Investments in information collection, analysis and dissemination is as critical as establishing a strong institutional framework in which vital tasks are clearly mandated. When the resource base is stretched to the limit, and during times of crisis, disputes and arguments over areas of responsibility can result in loss of livelihoods and economic opportunities. It should also be recognized that hydrological risk is reflected in its impact on the financial, economic and public health/safety aspects of a country.

152. In Asia, research and development are essential on how to manage water under monsoon conditions, especially in the tropics and semi-tropics where major river basins are closing fast. The productivity of water, including floodwater as presently used, must also be assessed to determine the extent which increasing demand for irrigated agricultural production can be met by stepping up water productivity, and when growing requirements need increased consumption. Any untapped discharges from basins that are open or semi-closed must be exploited and plans made to effectively capture and put this water to use.

153. Whenever feasible, combinations of small and large storage, and surface water and groundwater recharge are found to give the best systems. The lack of information on land use at local and sub-national levels also poses a constraint to adequate mitigation strategies for the agricultural sector. This needs to be remedied. Research must also be conducted on operational catchment hydrology, e.g. land-water linkage modelling, rural groundwater and conjunctive use management, relevant institutions, and on the medium- to long-term impact of strategies implemented to cope with water-related disasters on natural resources. The compatibility of this approach with integrated land and water management should be assessed.

BIBLIOGRAPHY

Amarasinghe, U.A., Mutuwatta, L. and Sakthivadivel, R. 1999. Water Scarcity Variations within a Country: A Case Study of Sri Lanka. Research Report 32. ISBN 92-9090-383-X. IIMI.

Research Centre for Natural Resources and Animal Husbandry. 2000. Aquifer Management: A New Approach to Soil and Water Conservation in the Deserts of Iran. Shiraz.

Burke, J. 2000. Land and Water Systems: Managing Hydrological Risk. Natural Resources. JNRF 24:2.

Calder, I. 2000. Land-Water Linkages in Rural Watersheds. FAO Electronic Workshop, Background Paper 1: Land Use Impact on Water Resources.

De Graaf, J. 2000. Land-Water Linkages in Rural Watersheds. FAO Electronic Workshop, Background Paper 5: Downstream Effects of Land Degradation and Soil and Water Conservation.

ESCAP. 1997. Guidelines and Manual on Land-Use Planning and Practices in Watershed Management and Disaster Reduction. United Nations, New York.

ESCAP. 1999. Regional Cooperation in the Twenty-First Century on Flood Control and Management in Asia and the Pacific. United Nations, New York.

Eschavarria, M. 2000. Land-Water Linkages in Rural Watersheds. FAO Electronic Workshop, Discussion Paper 4: Valuations of Water-Related Services to Downstream Users in Rural Watersheds: Determining Values for the Use and Protection of Water Resources.

FAO. 1976. A Framework for Land Evaluation. FAO Soils Bulletin No. 32. Rome.

FAO. 1977. Guidelines for Watershed Management. FAO Conservation Guide 1. Rome.

FAO. 1985. Watershed Management Field Manual - Vegetative and Soil Treatment Measures. FAO Conservation Guide 13/1. Rome.

FAO. 1986. FAO Watershed Management Field Manual: Gully Control. FAO Conservation Guide 13/2. Rome.

FAO. 1986. Strategies, Approaches and Systems in Integrated Watershed Management. FAO Conservation Guide 14. Rome.

FAO. 1987. Guidelines for Economic Appraisal of Watershed Management Projects. FAO Conservation Guide 16. Rome.

FAO. 1996. Rome Declaration on World Food Security and World Food Summit Plan of Action. Rome.

FAO. 2001. Committee on Agriculture. Reducing Agricultural Vulnerability to Storm-Related Disasters. Sixteenth Session. Rome.

FAO. 2000. New Dimensions in Water Security-Water Society and Ecosystem Services in the 21st Century. AGL/MISC/25/2000.

FAO. Water Security: Adaptive Strategies for Addressing Emerging Water Problems in South Asia (draft).

Gommes, R. 1992. The Role of Agro-Meteorology in the Alleviation of Natural Disasters. FAO.

Gommes, R., Bakun, A. and Farmer, G. 1998. An El Niño Primer, SD Dimensions.

Green, C.H., Parker, D.G. and Tunstall, S.M. 2000. World Commission on Dams, Thematic Reviews Option Assessment. In: Assessment of Flood Control and Management Options.

Gleick, P.H. 2000. The Changing Water Paradigm, a Look at Twenty-First Century Water Resources Development. Water International 25 (1) 33-39.

Helweg, O.J. 2000. Water for a Growing Population, Water Supply and Groundwater Issues in Developing Countries. Water International 25 (1) 33-39.

Government of India. Ministry of Environment and Forestry. 2000. National Report on Implementation of United Nations Convention to Combat Desertification. New Delhi.

Ives, J.D. and Messerli, B. 1989. The Himalayan Dilemma. United Nations University Press. New York.

Keller, A., Keller, J. and Seckler, D. 1996. Integrated Water Resource Systems: Theory and Policy Implications. Research Report 3. ISBN 92-9090-326-0. IIMI.

Keller, A., Sakthivadivel, R. and Seckler, D. 2000. Water Scarcity and the Role of Storage in Development. Research Report 39. ISBN 92-9090-399-6. IIMI.

Kiersch, B. 2000. Land-Water Linkages in Rural Watersheds. FAO Electronic Workshop, Discussion Paper 1: Land Use Impacts on Water Resources, A Literature Review.

Kiersch, B. 2000. Land-Water Linkages in Rural Watersheds. FAO Electronic Workshop, Discussion Paper 2: Instruments and Mechanisms for Upstream-Downstream Linkages: A Literature Review.

Lal, R. 1995. Sustainable Management of Soil Resources in the Humid Tropics. United Nations University Press. New York.

Le, H.T. and Facon, T. 2000. The FAO-ESCAP Pilot Project on National Water Visions. Synthesis of Experience in Southeast Asia. FAO-ESCAP.

World Water Council. 2000. A Vision of Water for Food and Rural Development.

Molden, D., Sakthivadivel, R. and Habib, Z. 2001. Publications & Research Output. Basin-Level Use and Productivity of Water: Examples from South Asia. Research Report 49. ISBN 92-9090-425-9. IWMI.

Mainguet, M. 1995. Man-Induced Desertification? Based on a Presentation Made. UN University Lectures: 12. Tokyo.

Manshard, W. and Morgan, W.B. 1988. Agricultural Expansion and Pioneer Settlements in the Humid Tropics. United Nations University. New York.

Molle, F. 2001. Water Pricing in Thailand; Theory and Practice. DORIS-Delta Research Paper 7. Kasetsart University, Bangkok.

RAP Publication. 1999. Flood Management and Mitigation in the Mekong River Basin. In: Proceeding of the Regional Workshop, Vientiane. 19-21 March 1998. FAO, Bangkok.

Reynolds, E.R. and Thompson, F.B. 1988. Forests, Climate, and Hydrology. United Nations University.

Riddell, J. and Palmer, D. 1999. S D Dimensions: The Importance of Coordinated Land Administration in the Next Millennium. Land Tenure Service. FAO.

Seckler, D. 1996. The New Era of Water Resources Management: from "Dry" to "Wet" Water Savings. Research Report 1. ISBN 92-9090-325-2. IIMI.

Swift, J. 1999. Pastoral Institutions and Approaches to Risk Management and Poverty Alleviation in Central Asian Countries in Transition. University of Sussex and FAO Rural Development Division.

United Nations University. 1995. The Relationship Between Drought and Famine. United Nations University. New York.

UNEP. 1999. Climate Change Information Sheets.

UNEP. 2001. Intergovernmental Panel on Climate Change. Working Group I. Third Assessment Report. Summary for Policymakers.

UNEP. 2001. Intergovernmental Panel on Climate Change. Working Group II. Third Assessment Report. Draft.

UNEP. 2001. Intergovernmental Panel on Climate Change. Working Group. Final Draft Summary for Policymakers. Climate Change 2001: Impacts, Adaptation and Vulnerability.

UNEP. 2000. The Regional Impacts of Climate Change.

White, W.R. 2000. Water in Rivers: Flooding. A Contribution to the World Water Vision. International Association of Hydraulic Engineering and Research. Wallingford.

World Commission on Dams. 2000. Dams and Development: A New Framework for Decision-Making.

^* Researched and prepared by Thierry Facon, RAP Regional Water Management Officer.

³¹ Green, Parker and Tunstall, 2000.

³² Green, Parker and Tunstall, 2000.

³³ Keller, Sakthivadivel and Seckler, 2000.

³⁴ RAP has developed a training programme on water harvesting (Klaus Siegert, RAPG contact person: Kluas [email protected]).

³⁵ Swift, 1999.

³⁶ Disaster mitigation and preparedness activities also play an important role at the relief and rehabilitation stage, by strengthening the resilience of agricultural systems and rural communities against future disasters through appropriate resettlement measures. However, time constraints during the post-disaster period may cause poor planning and co-ordination of recovery activities and lead to the adoption of schemes that are not sustainable in the long-run. These aspects need to be considered when deciding the appropriate placement of disaster interventions (OECD, 1994: Disaster Mitigation Guidelines).

³⁷ Kiersch, 2000.

³⁸ India's case, but the situation is similar for most countries.

³⁹ In the SLV case, a gap of about 30 per cent remains in water balance estimates despite detailed monitoring and numerous modelling efforts made over the past three to four decades.

⁴⁰ Green, Parker and Tunstall, 2000.

⁴¹ Ti and AFacon, 2001.

⁴² www.fao.org/ag/agl/watershed/

⁴³ A programme of pine reforestation in Sri Lanka on the Mahaweli catchments was meant to regulate flows and reduce erosion. In reality, both annual and seasonal flows were reduced. This project is having the reverse effect. It increased erosion. Fortunately there was little sedimentation in the reservoir. Hill slope sediment is deposited on the lower slopes and floodplains to the benefit of paddy farmers.

⁴⁴ Environmental goods and services are the conditions and processes through which natural ecosystems and their species sustain human life. They maintain biodiversity and produce such goods as seafood, forage, timber, biomass fuels, natural fibre, and many pharmaceutical and industrial products and their precursors.