This paper presents the first basin-wide assessment of the potential impact of climate change on the hydrology and production of the Ganges system, undertaken as part of the World Bank's Ganges Strategic Basin Assessment. A series of modeling efforts -downscaling of climate projections, water balance calculations, hydrological simulation and economic optimization -inform the assessment. We find that projections of precipitation across the basin, obtained from 16 Intergovernmental Panel on Climate Change-recognized General Circulation Models are highly variable, and lead to considerable differences in predictions of mean flows in the main stem of the Ganges and its tributaries. Despite uncertainties in predicted future flows, they are not, however, outside the range of natural variability in this basin, except perhaps at the tributary or sub-catchment levels. We also find that the hydropower potential associated with a set of 23 large dams in Nepal remains high across climate models, largely because annual flow in the tributary rivers greatly exceeds the storage capacities of these projects even in dry scenarios. The additional storage and smoothing of flows provided by these infrastructures translates into enhanced water availability in the dry season, but the relative value of this water for the purposes of irrigation in the Gangetic plain, and for low flow augmentation to Bangladesh under climate change, is unclear.

This paper summarizes the results of the Ganges Strategic Basin Assessment (SBA), a 3-year, multi-disciplinary effort undertaken by a World Bank team in cooperation with several leading regional research institutions in South Asia. It begins to fill a crucial knowledge gap, providing an initial integrated systems perspective on the major water resources planning issues facing the Ganges basin today, including some of the most important infrastructure options that have been proposed for future development. The SBA developed a set of hydrological and economic models for the Ganges system, using modern data sources and modelling techniques to assess the impact of existing and potential new hydraulic structures on flooding, hydropower, low flows, water quality and irrigation supplies at the basin scale. It also involved repeated exchanges with policy makers and opinion makers in the basin, during which perceptions of the basin could be discussed and examined. The study's findings highlight the scale and complexity of the Ganges basin. In particular, they refute the broadly held view that upstream water storage, such as reservoirs in Nepal, can fully control basin-wide flooding. In addition, the findings suggest that such dams could potentially double low flows in the dry months. The value of doing so, however, is surprisingly unclear and similar storage volumes could likely be attained through better groundwater management. Hydropower development and trade are confirmed to hold real promise (subject to rigorous project level assessment with particular attention to sediment and seismic risks) and, in the near to medium term, create few significant tradeoffs among competing water uses. Significant uncertainties – including climate change – persist, and better data would allow the models and their results to be further refined.

Hydroclimatic variability manifests as abrupt shifts, trends, runs, and recurrent cyclical phenomena, collectively referred to as components. In this article, we tested for the presence and magnitude of each component in the Niger Basin (West Africa), using the 0.5 ∘ by 0.5 ∘ gridded annual rainfall and temperature data produced by the Climate Research Unit (CRU) at the University of East Anglia, UK, for the period 1901-2006. The streamflow data was also analysed for different sub-basins of the Niger Basin. Abrupt shifts were tested using a Bayesian and nonparametric approach. Trends were analysed using the Mann-Kendall trend test. Runs were extracted for dry, neutral, and wet conditions, simulated 1000 times based on the skew normal distribution, and used to investigate various run characteristics. Cyclical behaviour was investigated using continuous wavelet analysis. The results show that an abrupt change point occurred in 1969 in the rainfall and streamflow (but not temperature) time series in all subwatersheds of the Niger Basin. The magnitude of the shift in the mean rainfall varied between 16 and 24%. The temperature time series exhibit strong positive trends in all watersheds. Post change point, the rainfall and streamflow time series show positive, though statistically non-significant trends at = 0.1. In contrast, disregarding the change point, all subwatersheds show significant negative trends at = 0.05 and the maximum run lengths are about 4 years long for both dry and wet conditions. Finally, wavelet analysis showed that both rainfall and streamflow in the Niger Basin fluctuate on cycles that are predominantly 2-4 years long, with a few occurrences in the 6-8 years range. Wavelet activity diminished noticeably when a series appeared strongly dominated by trends. The results provide a more comprehensive view of climatic variability than would be obtained from only one or a few components.

Water demands are outstripping supplies in many parts of Asia and causing conflicts, especially as rapidly urbanizing and industrializing areas compete with more traditional agriculture on alluvial plains near expanding metropolises. The environment is increasingly being seen as a legitimate and important user of water, often in competition with irrigation, domestic, industrial, hydropower and community uses of water. Any future scenario requires an improvement in water productivity, especially in agriculture. Surface and ground water supplies are not well regulated owing primarily to a weak knowledge base, analytical capacity in addition to a number of traditional institutional, governance, political and other factors. Water policy makers have, therefore, to work out strategies for integrated water and environmental management, which rely on a proper knowledge base of the basin hydrological and pollution conditions. Examples from various countries in Asia are elaborated in this paper to demonstrate how spatially distributed evapotranspiration data from remote sensing, in conjunction with other key data, can help to build the knowledge base for integrated basin scale water management. Remote sensing is not a solution, but it provides key data that is difficult to access by conventional data collection methods.

No abstract