Water security risk assessment in a fluvio-tidal setting through an integrated system dynamics model

Abstract

The southwest and southcentral coastal regions of Bangladesh are beset by high bio-physical vulnerability in the form of multiple hazards (i.e., cyclonic storm surge, fluvio-tidal flood, river-bank erosion, waterlogging/ drainage congestion, and salinity intrusion), impacting the primary livelihood activity (i.e., agriculture) via frequent direct damage and economic loss and irrigation water constraints, highly constrained irrigation opportunities because of presence of salinity in surface water as well as groundwater, water quality constraints, principally salinity, for both drinking water and sanitation, unsatisfactory implementation and/or effectiveness of water related laws, acts, rules, policies or plans, and high socio-economic vulnerability (i.e., high incidence of poverty, low level of education, poor access to resources, dilapidated housings, etc.). These different vulnerability dimensions produce water insecurity outcomes of multiple dimensions. This study was taken in the backdrop of these multi-dimensional water security aspects and was motivated by the importance of risk-based planning and the experiences of traditional planning exercises not being able to always provide desirable outcomes. The objectives of this research was to: (a) explore and identify the key factors and drivers (natural, physical, socio-economic, and institutional) that influence water security and livelihoods of the coastal communities; (b) develop an integrated model considering the natural, physical, socio-economic, and institutional processes of the coastal system; (iii) assess water security risk using the integrated model; and (iv) evaluate the options/interventions (physical and policy interventions) under different scenarios primarily for assessing trade-offs and minimizing risks. The terms ‘water security’ and ‘water security risk’ have been used interchangeably in this thesis, reflecting different dimensions or themes of water security, with the former focused mostly on the water security dimensions in terms of physical availability (i.e., availability of irrigation water or safe drinking water and improved sanitation), while the latter specifically focused on determining water related risks arising from coastal hazards and disasters. This allowed integration of information from different dimensions and comparison of their relative importance and priorities. A risk-based approach has been considered in the study for assessing water related risks, with two types of framings used for water insecurity assessment: (i) quantitative assessment (objective; statistical), wherein risk is the likelihood of a potentially harmful event and its negative consequence (damage) on lives, properties, and ecosystems; and (ii) qualitative (intangible; subjective) assessment, in which risk is not a real event in itself but is subjectively attributed to perceived potential threats. The definition of water security is extended to include the concept of “tolerable water security risk” or tolerable level of risk, which reflects a perceived tolerable level of potential losses that a society or community can absorb or consider acceptable, given the bio-physical, socio-economic, and natural-cultural conditions). The physical (quantitative) risk concept has been utilized to estimate impacts of cyclonic storm surges (in the form of storm surge induced loss and damage to agricultural crops, fisheries, housing and transportation infrastructure, and reduction of crop yield due to salinization) of different intensities and impacts due to drainage congestion (impacts on agriculture) under business-as-usual (BAU) and climate change (CC) scenarios. The ‘perceived risk’ concept was used to estimate the tolerable expected annual damage (TEAD), which reflected people’s perception about the frequency of cyclone occurrence and/or drainage congestion (pluvial flood) with corresponding damage and their recovery of assets and savings in subsequent non-disaster (or hazard) years. For assessing risk due to salinity intrusion and riverbank erosion, a qualitative risk framework was employed drawing from IPCC 5th assessment framework, which sees risk as a product of hazard, exposure, and vulnerability. This research has given emphasis on integrated framework or decision support tool for water security risk assessment, for which the system dynamics (SD) modeling approach has been. The SD models can address the complexity of a social-ecological system including non-linear processes, can simulate the dynamic interactions among physical, socio-economic, and institutional system components, and importantly also simulate and integrate the quantitative and qualitative components of risk to facilitate integrated assessment. Deterministic, stochastic, and empirical equations have been used to develop the mathematical relationships among different elements of the system. The SD model simulations have been carried out considering Business as Usual (BAU) scenario (for cyclone: no changes in cyclone intensity and mean sea level for drainage congestion: no changes in rainfall and mean sea level; for agricultural water security: increasing Boro crop land with corresponding increase in food demand for the growing population) and Climate Change (CC) scenario (for cyclone: an 8% increase in cyclone intensity. i.e., wind speed, and 1-meter sea-level rise by the year 2100; for drainage congestion: a 17.19% increase of monsoon rainfall by 2050s and 1-meter rise of mean sea level by 2100; for agricultural water security: decreasing Boro crop land because of sea-level rise inducedaggravation of soil salinity). This research focused on the southwest and southcentral coastal regions, with 6 polders, 4 in southwest region (Polder 29, Polder 32, Polder 17/1, and Polder 35/1) and 2 in southcentral region (Polder 43/2F and Polder 46), selected for deep-dive analysis. The polders represent areas from both exposed and interior coasts, with distinct heterogeneity in natural, physical, and socio-economic characteristics. Indeed, the research clearly showed differences in water security status in terms of the different dimensions, with varying degrees of severity or importance of different water security aspects across the polders. The location of polders, i.e., the distance from the coastline is an important factor determining the impact of storm surges, with the “exposed” polders more vulnerable to storm surges in terms of area of inundation (Polder 46, Polder 35/1, and Polder 32), while the ‘interior’ polders (Polder 17/1 and Polder 29) are very lowly vulnerable both in BAU and CC scenarios, because of their relatively long distance from the coastline. Geospatial variation is also important, as differences of land slopes allow higher tide propagation and hence likelihood of more inundation area on the southwest coast than on the south-central coast. To what extent the polders are maintained, i.e., whether the embankment height is at design level or degraded to a lower height due to lack of maintenance, makes a huge impact on storm surge related damage and loss. Keeping the embankment heights at the ‘design level’ will keep the interior polders safe against most of the moderate to big cyclonic events. However, the exposed polders are likely to suffer more in the CC scenarios with higher inundation areas and higher expected annual damage (EAD) than tolerable expected annual damage (TEAD), implying that height enhancement of embankments may be required for these polders close to the sea. However, the SD simulation considered uniform lowering of embankment height while considering the case of existing embankment height, meaning that the inundation area has likely been overestimated compared to reality. This further reinforces the need for regular maintenance of the polders with an effort to maintain the design embankment height.Raising the plinth levels of a house can be a good adaptation measure to protect a home from tidal floods, fluvio-tidal floods as well as storm surges. The SD simulations showed that the EAD can be reduced considerably by raising the plinth level by 1 m, especially in Polders 17/1, Polder 29, and Polder 35/1. Also, the lowering of height of polder embankments much below the design height in many places due to lack of maintenance has contributed to fluvio-tidal flooding inside the polders, especially for Polders 32, Polder 35/1, and Polder 46. As tidal flood and fluvio-tidal flood magnitudes are lower than storm surge flood, in some cases, full loss and damage could be avoided by raising the house plinth level. Lack of maintenance of water control structures (like sluice gates and flap gates) and silting up of internal canal/khal of polders have created drainage congestion problem for some polders. Also, rainfall intensity has increased in recent years, exacerbating drainage congestion. Among the six selected polders, Polder 29 and Polder 17/1 are highly vulnerable, with the EAD due to drainage congestion higher than the TEAD perceived by the local people.Two interventions could be used: re-excavation of the internal canals and pumping out water from within the polders. Polder 17/1 and Polder 29 are the two most vulnerable polders, in which re-excavation of canals and the functioning of water control structures can reduce the EAD due to drainage congestion but is not enough to reduce it to the tolerable level (i.e., TEAD). Pumping water out of the polders together with canal re-excavation can reduce the EAD below the TEAD. The polders near the coastline are at higher salinity risk than the interior polders. The polders located in the southwest zone are at more salinity risk than polders located in the south-central zone because of less freshwater inflow from the freshwater zone in the upstream. In addition, lack of proper operation and maintenance of water control structures like sluice gates is also responsible for increasing salinity inside the polder. During cyclones, storm surge (saltwater) enters the polder by overtopping or breaching the embankment that causes increasing salinity risk.Salinity is a major problem as far as household water security (access to safe drinking water and access to improved sanitation and hygiene) is concerned. However, there is variation in the occurrence of salinity among and within the polders. Salinity is a major problem in Polder 32 and Polder 35/1 both in surface water and groundwater, with concentrations above the drinking water standard in many places. Agricultural water insecurity is an outcome of the interactions among irrigation water demand, changing demand with population growth, climate parameters (dependable rainfall, temperature, etc.), source of irrigation (i.e., surface water and principally groundwater, constrained by high levels of salinity), health of soil (i.e., soil salinity, which impacts yield of Boro), and water control structures (i.e., drainage cum irrigation canals). Because of less salinity constraints in groundwater, the cultivated lands in Polder 29 and Polder 17/1 in the southwest region are higher than other polders and so are irrigation water demands (about 16 Mm3 for Polder 29 and 11.4 Mm3 for Polder 17/1), 80% of which are met by groundwater and 20% by surface water. Enhancing Boro crop production is more relevant for these polders, which can be done through introduction of Salt Tolerant Varieties (STVs), while for other polders with limited increase in irrigation water requirement in the future, re-excavation of existing canals/khals will facilitate irrigation. There is room for improvement in the context of implementation and enhancing effectiveness of acts, laws, rules, plans, and policies. Strong and effective WMOs are important for sustainable management of water security. Formal WMOs exist in polders where water management projects are related to institutional water management. The WMOs are mostly project based, formulated during the project, and have better performance throughout the project, while the performance gradually declines with lower functionality once the project is completed. Lack of good governance such as political influence and/or power practice, and lack of funding and transparency are the main causes of lowering of performance of WMOs. In sum, the coastal zone needs collective interventions rather than a single intervention to secure water security for people and for enhancing and sustaining their life and livelihoods. This will require an appreciation of the complexity and spatial heterogeneity in the coastal zone, and across the polders in particular. Given that polders are affected by different dimensions of water insecurity of different degrees of severity, an integrated polder management system, with multiple synergistic interventions following a systems approach, will help achieve the desired goals. The system dynamics modeling approach used in this study is a good example of tools that can assist the process.