Loss of habitat connectivity due to anthropogenic structures is among the greatest threats to freshwater fish populations. Re-establishing river connectivity through fish pass facilities can be an effective and cost-efficient method of enhancing local productivity, yet many are unsuccessful. A good understanding of multispecies swimming performance and behavioural response to hydraulic conditions is therefore needed to improve designs. This thesis aimed to improve knowledge in this field for non-salmonid fish species of conservation concern and economic value. Swimming performance data were collected for juvenile bighead carp (Hypophthalmichthys nobilis), a species threatened by anthropogenic barriers in China, using a range of swim chamber and open channel flume methodologies. Burst swimming performance was relatively weak, especially where multiple high velocity areas had to be passed. In addition, the availability of low velocity areas in a section of open channel flume did not improve endurance, and beyond aerobic swimming speeds these velocity refugia were rarely utilised. Management recommendations for fish pass velocities are presented based on this data. To further explore carp behavioural utilisation of low velocity regions, juvenile common carp (Cyprinus carpio) swimming performance and behaviour were evaluated under various wall roughness treatments. Fish generally maintained position close to smooth walls and small corrugations, yet often moved further from medium and large corrugations and into areas of higher velocity and lower turbulent kinetic energy. Thus, performance was not enhanced by the larger areas of low velocity created by corrugated walls. To assess the influence of accelerating flow on European eel (Anguilla anguilla) behaviour, a constricted flume created a velocity gradient representative of that found at anthropogenic structures and downstream bypass facilities. Of 138 downstream moving silver eels approaching the constriction, 46% reacted by changing orientation and/or a rapid burst of upstream swimming. Furthermore, 36% rejected the constricted channel and returned upstream, delaying downstream passage. The probability of a rejection was increased by a high abundance of the invasive parasite, Anguillicoloides crassus. These findings have potential implications for bypass efficiencies and escapement to sea. Eel swimming performance and behaviour were also evaluated during upstream passage through a culvert. Traditional corner baffles and prototype sloped baffles improved passage success compared to a bare culvert. Although the prototype created higher barrel velocities and turbulence, eel passage success was equal between the two baffle designs. Installation of the sloped baffle is recommended due to high passage efficiency and the potential to reduce the risk of debris accumulation compared to 90° baffles. The data presented in this thesis enhances our understanding of non-salmonid swimming performance and behaviour, and are used to recommend approaches to fish pass design for European eel and Asian carp.

Anabranching fluvial networks recently have become the focus of attention from environmental specialists, especially in the hydraulic field. Anabranching networks can be found in different physical environments; however, the hydraulic and geomorphological natures of such river networks are still not well known leading to on-going discussions on the definition and nature of the networks. Even though, alluvial anabranching networks generally have common features like vegetated islands, low water surface slope and stable channel planform, bedrock-confined anabranching networks also have their own characteristics inherited from the geological and structural controls imposed on the single channels that compose the network complex. This thesis focuses on the provision of a benchmark describing the bulk hydraulic characteristics of a large bedrock-confined, anabranching river network, located within southern Laos. The network can be separated into: (i) the upper river network constituted by two bifurcations and one confluence with an interpolated bathymetry based on soundings of cross-sections along the navigation channels; and, (ii) the downstream river network characterised by a complex anabranching network with five bifurcations and five confluences for which there is no bathymetric survey. The river network as whole is a ‘composite’ – partly bedrock (especially the channel-bed) and partly alluvial-filled and as such it does not accord fully with any prior description or classification of anabranching channel networks (e.g. Huang and Nanson, 1996). To understand the hydraulic nature of the river network, the energy approach in a onedimensional (1D) steady-flow hydraulic model (HEC-RAS) was applied to the network. Significant challenges arose due to the lack of boundary conditions throughout the model, namely: (i) unknown splitting discharge ratios at each bifurcation; (ii) partly non-survey bathymetry; and, (iii) ungauged downstream boundary condition of one of the channel outlets. To determine the discharge entering each channel, the splitting discharge ratio at each bifurcation was defined originally by the ratio of the cross-sectional area of the first crosssection of each downstream channel and then adjusted based on the Flow Optimization function in HEC-RAS to minimize any rise or drop of the modelled water surface around a junction. For the channels with non-surveyed bathymetry, a SPOT satellite image was processed to construct a pseudo-bathymetry showing a range of elevations, including shallow and deep portions of channels, rather than detailed bed elevations as would be obtained from a measured bathymetry. To define the boundary condition of the ungauged channel outlet, the water surface elevation was interpolated and validated according to predefined assumptions (i.e. the water surface slope along the ungauged channel was interpolated according to the available DEM and cross-sectional width extracted from a SPOT image for low discharge conditions was assumed to be similar to the gauged channels for flooding discharges). In general, the study has helped to develop methods to model the complex river network with data constraints (i.e. the boundary conditions). The findings include: (i) the developed pseudo-bathymetry based on a SPOT image is useful to model a large river network using the energy approach in a 1D hydraulic model in which the cross-sectional area is important in modelling the bulk hydraulic parameters but the influence of the cross-sectional shape is subordinate; (ii) the in-channel hydraulic roughness coefficient at each cross-section may be significantly different from neighbouring values due to the variation in the local bedrock roughness and the roughness of intervening alluvial reaches; and, (iii) the hydraulic roughness of the riparian land cover along the floodplains does not contribute noticeably to the modelled stage along the river network nor to the planform extent of flooding for overbank flooding discharges. Rather, changes in land-cover, and hence the riparian roughness, are registered as small, but measureable, changes in the local velocity over the riparian floodplain and in the average in-channel velocity. Citations: Van, P.D.T., 2009. Hydraulic modelling and flood inundation mapping in a bedrockconfined anabranching network: The Mekong River in the Siphandone wetlands, Laos. Unpublished PhD thesis submitted to the Faculty of Engineering, Science and Mathematics, University of Southampton, England.

Aerosol Optical Thickness (AOT) data have many important applications including atmospheric correction of satellite imagery and monitoring of particulate matter air pollution. Current data products are generally available at a kilometre-scale resolution, but many applications require far higher resolutions. For example, particulate matter concentrations vary on the scale of tens of metres, and thus data products at a similar scale are required to provide accurate assessments of particle densities and allow effective monitoring of air quality and analysis of local air quality effects on health. This thesis describes the development of a novel method which retrieves per-pixel AOT values from high-resolution (30m) satellite data, and this method is the main novel contribution to scientific knowledge of this PhD. This method is designed to work over a wide range of land covers including both bright and dark surfaces - and requires only standard visible bands, making it applicable to a range of data from sensors such as Landsat, DMC, SPOT and Pleiades. The method is based upon an extension of the Haze Optimized Transform (HOT), which was originally designed for estimating the haziness of each pixel in a satellite image, based upon the distance from a `Clear Line' in feature space. In this research, the HOT method is adapted and used to estimate AOT instead. Significant extensions include Monte Carlo estimation of the `Clear Line', object-based correction for land cover, and modelling of the HOT-AOT relationship using radiative transfer models. Validation against ground and satellite measurements, as well as simulated data, shows that 40-50% of the pixels have an error within ±0.1, not much lower than many presently available low-resolution products, with further work likely to improve the accuracy. Two example applications show the potential of this method for per-pixel atmospheric correction and monitoring the spatial pattern of particulate matter pollution. This novel method will enable many new applications of AOT data that were impossible with low-resolution data.

The past two decades have witnessed significant growth in attention and investment in renewable energy technologies. Replacement of fossil fuels that have long dominated our energy production is favoured as resources are known to be finite and dwindling, leading to increasing prices, as well as the link between their use and global climate change. In 2011, transport accounted for 38% of total national energy consumption with petroleum being the single most used fuel. In terms of how energy use by transport relates to greenhouse gas (GHG) emissions; in 2011 transport consumed around 55.19 million tonnes of oil equivalent and provisional estimates put emissions of carbon dioxide (CO2) at 119 million tonnes (over one quarter of total national CO2 emissions). As a result, biofuels have been increasingly appearing on the agendas of both governments and scientists, and have been picked up by the media and various environmental organisations as a possible means for reducing the GHG contribution from transport. However, the issue is not simple, and there are many who oppose the use of biofuels for various reasons. There are concerns that without a concerted effort to improve the state of knowledge of potential risks and benefits of biofuels, the appropriate long-term development of the technology in the UK may be hindered. The project presented in this thesis was designed to undertake an investigation to identify relevant risks and issues that could inform a risk analysis of the future development, production and use of biofuels in the UK. In the context of biofuels, there is a large and increasing literature in which the associated risks are characterised and assessed scientifically. However, very little research has been done looking at stakeholder opinions, particularly with the public as stakeholders. Increasingly, the media, non-governmental organisations (NGOs) and the public are concerned about environmental issues and large technological developments that affect the environment and themselves. If any of these groups oppose plans and decisions made then it is possible that they can cause significant disruption or halt progress, despite scientific evidence. Through a series of social science methods involving stakeholders, this project has endeavoured to cast light on the broader understanding and perception of biofuels beyond the academic and research communities and their publications. The primary novel contribution of the thesis is in the insights provided into public awareness, attitudes and perceptions of biofuels, which have previously not been studied in any depth. The data collected and issues identified could potentially be very useful in informing a risk analysis exercise. Working in collaboration with the general public, through focus groups and questionnaires revealed widespread, low level awareness and knowledge of biofuels but little in the way of accurate detailed knowledge of impacts and risks. Public concerns were largely focussed on environmental impacts and personal financial impacts, and their views were almost exclusively informed by mass-media sources such as newspapers and television. There was also some evidence of misinformation and awareness of issues that were not considered to be risks by the scientific literature, as well as notable exaggeration of known risks. Public attitudes were deemed to be quite dated, strongly reflecting the view of biofuels presented by the media around 2008, when a number of critical studies were published and shook global confidence in biofuels. With the onset of the global recession, media coverage of biofuels has dropped significantly, and as such, the public have not been exposed to developments in the field. Interviews with expert stakeholders revealed a different picture to the public, and highlighted a completely different perspective – that of threats and risks to the future of biofuels, rather than biofuels as a threat or risk themselves. There was a strong perception amongst the expert stakeholders that the UK and EU governments presented a significant barrier to the potential for biofuels to develop and expand in Europe. Lack of government interest, confidence and action were cited as significant failings that hindered investment necessary to grow a strong biofuel industry. Recommendations for future developments and expansion of biofuels within the EU, should this be deemed appropriate and acceptable, focus on increased government involvement and support to encourage investments that will allow further improvements in the biofuel production process, as well as significant changes in the way scientific information is communicated to the public.

Hexavalent chromium (Cr(VI)) is a highly toxic and soluble substance present in a wide range of industrial effluents. An effective treatment method is the biochemical Cr(VI) reduction to the less toxic trivalent chromium (Cr(III)), and such a transformation has recently been demonstrated in bioelectrochemical systems. However, depending on the pH of the catholyte, a biocatalyst might be required in the cathode and also the process can be very much inhibited by Cr(III) products which tend to form on the electrode surface and deactivate it. Herein is demonstrated how an electrophilic bacterium, that is Shewanella oneidensis MR-1, can be used as a bacterial catalyst in Cr(VI) reducing cathodes of bioelectrochemical systems. Starting with potentiostatically controlled experiments (-500 mV vs. Ag/AgCl for 4 h), the abiotic cathode’s (AC) performance was shown to be affected by the Cr(VI)-reduction products that are deactivating the cathode and are severely inhibiting further Cr(VI) reduction. The presence of metal chelators like lactate delayed this deactivating effect and enhanced the system’s performance to a large extent; in the presence of 30 mM lactate, the AC delivered approximately 3 times more electrons to Cr(VI). In addition, approximately 15 times more electrons were delivered when 5 μM of the electron shuttle riboflavin was also added in the AC. However, the presence of riboflavin did not have any effect in the absence of lactate. The MR-1 biocathode also exhibited an enhanced current production and Cr(VI) reduction, though the pre-treatment conditions were found to be important. When pre-treated in an aerated chamber with a poised electrode at +300 mV vs. Ag/AgCl, the MR-1 biocathode mediated 70% more electrons than the AC with 30 mM lactate, and only 39% more electrons when the electrode was initially poised at -500 mV vs. Ag/AgCl. Cr(VI) reduction was also enhanced, with a decrease in concentration over the 4 h operating period of 9 mg L-1 Cr(VI) in the aerobically pretreated MR-1 chamber, compared to only 1 and 3 mg L-1 in the AC without lactate and in the AC with 30 mM lactate respectively. On the other hand, when pre-treated anaerobically in the presence or absence of Cr(VI), the performance of the MR-1 biocathode was not much different than that of the AC with 30 mM lactate. The positive effect of lactate was further demonstrated in microbial fuel cell (MFC) cathodes, where maximum power densities produced were up to 44 times the power densities reported elsewhere for abiotic cathodes (8.8 mW m-2 vs. 0.2 mW m-2 at pH 7) and at similar levels to the power densities of biotic cathodes at pH 6 and 7. Considerable Cr(VI) reduction was also observed at alkaline pH abiotic cathodes and maximum power densities were 31 times the ones reported elsewhere for biotic cathodes at pH 8 (21.4 mW m-2 vs. 0.7 mW m-2). In MFCs, the presence of MR-1 enhanced the performance of pH 7 cathodes; in the presence of 30 mM lactate, the MR-1 biocathode bioelectrochemically reduced 3 times the amount of Cr(VI) reduced by the AC with the same amount of lactate. Compared to the results in the absence of an electrode, the MFC results suggested that different Cr(VI) reduction pathways could be utilised by MR-1 when the electron donor is in the poised electrode form rather than in the lactate form. In conclusion, effective and continuous Cr(VI) reduction with associated current production were achieved when MR-1 and lactate were both present in the biocathodes.

Approximately 80 of the 170 commercial ports in the UK are either in or near areas protected under the European Habitats Directive. These areas are often also focal points for related industries, urbanisation and recreation. This brings in to question what impact these developments have had on these protected habitats and should future developments be given permission to occur. Pressures on estuaries were found to impact upon the dynamic equilibrium of an estuary through changes in the balance of the sediment budget, energy dispersal and estuary morphology. It was the aim of this thesis to show that through understanding the historical relationship between various pressures and the impact on these components that management of these environments could be improved to protect their environmental interest while allowing sustainable development to occur. Also by understanding the state of equilibrium it could be determined whether historic projects should be included in the cumulative and in-combination impact assessment for new plans or projects under the Conservation of Habitats and Species Regulations 2010. Southampton Water is an industrialised estuary on the South coast of England, the European protected salt marsh and mudflats lining its shores have been declining so this was an ideal case study site to investigate the relationship between pressures on estuaries and their impacts. A geographical information system was used to analyse historic data for Southampton Water between 1783 and 2008 to create 3D model for Southampton Water over time. From this model the volume changes of estuary components, the sediment budget and the change in dynamic equilibrium were quantified for the entire estuary. In novel work the changes were related to historic events which had occurred in the estuary. Importantly the errors associated with the sediment budget and morphological changes were quantified to provide a degree of confidence to these findings. This thesis found that the biggest influences on the sediment budget and equilibrium state of the estuary between 1783 and 2008 were dredging, land claim and Spartina growth/dieback. Land claim alone immobilized 255.4(+/-62) x106m3 of sediment. The estuary has moved from equilibrium state between 1783 and 1894 to one which is sediment starved due to a combination of these pressures.1,525 (+/-280) x103m3/yr of sediment was needed between 2001 and 2008 to balance the budget. Due to continued pressures the basin area is too large for the intertidal area the estuary supports. Error analysis was important in providing confidence to these results. The Main Channel and Test Estuary are still adjusting to historic changes this means that future plans/ projects should take these in to account when assessing the cumulative impact for Habitat Risk Assessments. This thesis showed that future management of the estuary should focus on securing sediment supplies to the estuary and on stabilising the intertidal zone. Subject to data constraints the methods used in this thesis could be applied to any estuary to assess the impacts of drivers and pressures on protected habitats and the management regimes which could be used to ensure their survival.

World-wide there are substantial quantities of sugar beet pulp, which arises as a residue after the processing of whole beet to extract sugar for refining as a foodstuff or for use in fermentation, in particular for the production of ethanol for the biofuel market. In both cases the resulting pulp residue is still rich in pentose sugars and fibre, and the research considered anaerobic digestion (AD) as a potential technology for the conversion of this material into renewable energy in the form of biogas. To determine the best operational conditions for biogas production both mesophilic and thermophilic digestion options were considered. Both were tested using 4-litre working volume mixed digesters operated with semi-continuous feed over a minimum of three hydraulic retention times (HRT). The first long term trial used mesophilic temperatures (37 oC±0.5 oC) at applied organic loading rates (OLR) from 2-5 g volatile solids (VS) l-1 day-1. This resulted in a specific methane yield of ~0.31 l CH4 g-1 VS day-1 with a biogas methane content of 51.05%. VS destruction was ~90% at all loadings, and increasing the loading resulted in an increase in volumetric biogas and methane production without significant loss in specific yields. The major limitation found was not in the biochemical conversion but in dewatering of the digestate, the characteristics of which were assessed using capillary suction time (CST) and frozen image centrifugation (FIC). At the higher loading there was also the appearance of a stable foam which made the digesters difficult to operate as this could block the gas outlet, leading to pressure increases and the loss of digestate by ‘blow out’. In the same digesters at mesophilic temperatures antifoam was tested to assess if this could offer a solution by suppressing foam formation. In practice this required unusually high doses of the reagent and, in continued use, these appeared to have an inhibitory effect on the digestion process. Dilution of the feedstock to the digester was also tested but showed no beneficial effects on dewaterability or foaming. As a post-treatment alternative cellulolytic enzymes were added to the digestate, but had no effect on improving dewaterability. Trace element (TE) supplementation to the digesters was, however, shown to eliminate the occurrence of foaming and also gave a slight improvement in dewaterability. TE supplementation reduced the polymer dose required for dewatering as determined by the CST test, and eliminated polymer dosing when dewatering was by centrifugation. Digestate dewaterability could also be improved in a post-digestion one- and two-stage chemical treatment with the use of chemical coagulants/flocculants alone or combined. The second long-term trial compared mesophilic (37 oC±0.5 oC) and thermophilic (55 oC±0.5 oC) digestion over 3 HRT using duplicate digesters fed at OLR of 4 and 5 g VS l-1 day-1. The digesters were operated without water addition. The thermophilic digesters gave higher biogas and methane productivity and were also able to operate stably at the higher OLR, whereas the mesophilic digesters showed signs of instability. Digestate dewaterability was assessed by the CST and FIC tests and the likelihood of stable foam forming was assessed using a foaming potential test. The results showed thermophilic operation performed better even at the higher loading and gave a digestate with superior dewatering characteristics and with very little foaming potential. Using a combination of CST tests, filtration tests, Frozen Image Centrifugation, SEM and grading centrifugation it was concluded that the poor dewaterability seen in mesophilic digestate was due to the presence of extracellular polymer substance (EPS) leading to blinding of the filter by fine particulate materials. The carbon, energy and nutrient (CEN) footprint was estimated for mesophilic and thermophilic digestion in which the process was coupled with combined heat and power (CHP) and biogas upgrading to biomethane. The results showed that the energy input for thermophilic digestion was higher than for mesophilic although this could be compensated for by the increased specific methane yield at the higher loadings modelled. There was also no significant difference in the emissions savings or in the quantities of nutrients recycled in the digestate. The model indicated that the use of CHP gave a higher net energy yield compared to biogas upgrading, but this of course is dependent on there being an economic use for the heat produced. Keywords: anaerobic digestion, sugar beet pulp, biogas, trace elements; digestate, dewatering, foam formation, carbon footprint, energy footprint, nutrient footprint.

As a consequence of climate change, global coastal communities are increasingly at risk from sea level rise and increased storm intensities. Therefore, to inform coastal zone management coastal vulnerability assessments with respect to present and predicted climate change scenarios is important. Most of the literature concentrates on physical, and to a lesser extent socio-economic aspects but no comparable studies detailing coastal vulnerability from both physical and economic vulnerability were found. To fill this important research gap, the current study developed a combined coastal vulnerability (physical + economic) index by integrating both a Physical Coastal Vulnerability Index (PCVI) and an Economic Coastal Vulnerability Index (ECVI). All indices were applied to eleven case study sites across the country and based on assessments, the Combined Coastal Vulnerability Index (CCVI) was validated. Subsequently, coastal areas were ranked according to their PCVI, ECVI and CCVI values. PCVI results showed that Great Yarmouth and Happisburgh have high vulnerability, contrasted against an Aberystwyth frontage that was least vulnerable. ECVI assessments showed that both Great Yarmouth and Skegness have high economic vulnerability while Spurn Head had low economic vulnerability. In total, the economic costs related to case study site vulnerability was assessed at £22.36 billion. Combined coastal vulnerability results showed that Great Yarmouth is highly vulnerable with the highest aggregated score (25) followed by Aberystwyth (21). Llanelli (16) and Lynmouth (16) were least vulnerable with respect to site CCVI. This research makes a contribution to knowledge, not just for the UK but on a global level. Each location has a unique set of conditions and economic needs, and was found to be functions of physical and economic pressures, e.g. number of properties, coastal erosion and population. Finding the most effective and sustainable solution is important and one that includes knowledge of environmental impact and socio-economic consequences. The three indices (PCVI, ECVI and CCVI) are justified as tools for planners and policy makers for developing management strategies to improve coastal resilience under scenarios of sea-level rise and climate change.

The global economic system has been undergoing fundamental changes since the 1980s. Many emerging countries drastically increased their openness to trade and foreign investments. Formerly socialist countries entered a transition towards a market-based model and deepened their integration into the global economy. As a result, the geography of trade, investment flows, and innovation is becoming more multipolar. This thesis seeks to improve our understanding of the links between these macro-level shifts and the geography of innovation, spatial patterns of economic deprivation, as well as firm-level outcomes in emerging countries. This thesis is structured into an introductory chapter and four analytical papers. The introductory chapter outlines three themes corresponding to the areas to which this thesis makes a contribution: (1) the interplay of the local and the global dimension in shaping regional patterns of knowledge creation, (2) the link between the relative weight of the private sector and spatial patterns of economic deprivation, and (3) the role of global production networks and the changing geography of trade in shaping regional patterns of innovative performance and heterogeneous firm-level outcomes. The first paper examines the geography of innovation in Russia, adopting a perspective that combines Soviet-era legacies, contemporaneous regional conditions, and global linkages. The results shed light on multinational enterprises’ (MNEs) role as key agents providing Russian regions with knowledge from distant places. The findings simultaneously point to the importance of path dependencies in regional patterns of knowledge generation. The second paper investigates the link between regional innovative performance in Brazil, Mexico, and Colombia and investments of MNEs categorized by entry mode and business function. The analysis suggests that the relationship between global linkages established by MNEs and regional knowledge creation is jointly shaped by the heterogeneity of MNEs’ investments and the heterogeneity of region-specific conditions in Latin American economies at different stages of technological development. The third paper focuses on Vietnam, a country that has seen some provinces act as pioneers and others as laggards in the journey towards an outward-oriented marketbased economy. The link between the private sector’s weight in the economy and economic deprivation is a topic of considerable policy interest, but its subnational dimension remains underexplored. The analysis considers the relationship between provincial differences in the change of private firms’ formal employment share and changes in the geography of economic deprivation. The findings reveal that increases in private firms’ employment share are associated with reductions in poverty. MNEs appear to be a key driver of this association. Finally, the fourth paper concerns Vietnam’s growing trade with China. It looks at the link between imports from China and firm-level outcomes in Vietnam’s manufacturing sector. The results show that, contrary to previous findings for advanced economies, exposure to imports from China is positively linked with firm-level employment. Information on trade in intermediates suggests that inputs imported from China may support Vietnam’s export growth. The findings cast light on the necessity to consider the role of global production networks and trade in intermediates when assessing the developmental implications of changing trade patterns.

This thesis explores two distinct but related aspects of the relationship between the spatial location of people within a country and national economic growth. The first three essays set out to establish whether a country’s urban structure impacts economic growth at the aggregate level. Each paper explores a different aspect of the location of people, including the level of urban concentration and the size of a country’s cities. The analyses rely on cross-country regressions and employ different estimation techniques, including fixed effects, system GMM and instrumental variables. Particular attention is paid to whether the relationship differs between developed and developing countries and how contextual factors, such as a country’s economic structure and government capacity, shape the relationship. Across the three papers, the results consistently suggest that concentration and larger cities can be growth promoting at the country level; however only in developed countries or if certain contextual factors are met. The findings add to a growing body of empirical literature which questions the universal validity of the benefits of agglomeration for economic growth. Furthermore, they specifically address a gap in the empirical literature which so far had failed to link city size to aggregate growth as opposed to city-level productivity. The fourth essay moves to the policy level and analyses Special Economic Zones (SEZs) – a policy which isfrequently employed to influence the location of economic activity and people. Specifically, it studies the performance drivers of SEZs. A scarcity of data has limited quantitative research on this topic so far. The analysis relies on a novel dataset, which resorts to nightlights as SEZ performance proxy and covers zone and policy characteristics for SEZs in 22 countries. The findings partially confirm, but also refute the dominant knowledge on the viability of SEZs. While larger zones tend to perform better, growth is difficult to sustain over time and particularly hard to achieve for high-technology focused zones. Other factors commonly assumed to matter, such as the nature of the zone operator, the incentive package and programme set-up, seem to be highly context dependent. Furthermore, contextual factors, such as proximity to markets and a pre-existing industrial base, influence zone performance.