1 “Introduction: Why Imagery, Why Now?”
1.1 The promise of imagery for social science and policy
We are living in a fast changing society needing timely, spatially granular, and consistent information to adequately address complex social, economic, climatic, health and policy challenges. Challenges such as widening inequalities, uneven economic development, environmental vulnerability and persistent health disparities represent fundamental questions about places and how they change. Traditional sources, including censuses, household surveys and administrative registers remain indispensable. Yet, they are limited by cost, infrequency, selective coverage and latency in their collection and release. As a result, decision-makers and researchers often lack the evidence needed to monitor rapid change, anticipate emerging challenges, and evaluate the local effects of interventions.
Satellite imagery offers a powerful and underused data solution particularly in social sciences applications. For more than half a century, satellites have orbited the Earth, capturing a detailed record of its surface and atmosphere. Initially designed for environmental monitoring, these data now represent an unparalleled observational archive of human and natural systems. Imagery captures a large spectrum of our natural and built environment, including the physical fabric of places, density of buildings, presence of green and blue spaces, road networks and informal settlements. It does this consistently across the globe and can capture change at multiple scales from local neighbourhood areas to national borders or entire continents. Unlike most traditional forms of social data, which are geographically fragmented and slow to update, imagery is comprehensive, timely and replicable.
An increasing volume of research has already demonstrated the promise of satellite imagery. Night-time lights have been used as proxies for economic activity and inequality across the globe (Henderson, Storeygard, and Weil 2012a), sense changes in urban energy consumption (Rowe, Robinson, and Patias 2022) and delineate labour markets (Soto, Rowe, and Soto 2025). Multispectral imagery has enabled the mapping of urban deprivation (Arribas-Bel, Patino, and Duque 2017b) and the monitoring of land use change in fast-growing regions (Brown et al. 2012a). More recent advances in computer vision and machine learning further enhance our capacity to derive meaningful indicators from raw pixels, opening a new frontier for social research and policy. The challenge (and opportunity) lies in integrating these data into the mainstream to illuminate processes that have long remained in the shadows, to take satellites into the standard toolkit. Key to this challenge is making imagery more accessible and usable.
1.2 Chapter objectives
This chapter seeks to provide readers with an understanding the conceptual foundations that motivate the use of satellite imagery in social science and public policy. Specifically, the chapter aims to:
1. Explain why imagery matters for observing social, environmental and infrastructural inequalities, and why its relevance has grown at this particular technological and policy moment.
2. Identify the unique strengths of imagery, including its spatial reach, temporal depth and consistency, and how these complement traditional data sources such as surveys, administrative registers and censuses.
3. Describe the recent convergence of technological advances that now make imagery feasible, scalable and policy-relevant.
4. Highlight imagery’s emerging role in addressing societal challenges from climate resilience to infrastructure planning and environmental justice.
5. Introduce the barriers and challenges that have historically limited the uptake of imagery in the social sciences, setting the scene for how Imago seeks to overcome them.
After reading this chapter, readers should have a clear understanding of the motivations behind the BoI, the conceptual value of satellite data for social research and the broader landscape in which Imago’s data products and analytical tools are situated.
1.3 What makes satellite imagery unique?
Satellite imagery is unlike most other forms of data used in the social sciences. The uniqueness of satellite imagery comes in the way multiple distinctive attributes compound to provide an observational resource that is comprehensive, consistent, granular and increasingly accessible. These qualities allow imagery to transcend the limitations of conventional social data sources and open new opportunities for analysis and policymaking.
Satellites view the entire Earth, offering data that extend across countries, regions and communities. This comprehensiveness can complement traditional data streams, like surveys, which are often limited by sample size, response bias or geographic reach. They can also enhance administrative registers in territories where governments maintain robust data infrastructures. Satellite imagery generates systematic records of all visible surfaces within its sensor range, covering remote rural areas, informal settlements and conflict-affected regions where social data are especially scarce (Kugler et al. 2019).
Modern satellite platforms capture imagery at a range of spatial resolutions from kilometres to less than one metre per pixel. Coarser imagery (say 30 kilometers) provides consistent monitoring of land cover and climate-related variables at regional and national scales. Very high-resolution imagery (say 3 meters) can reveal building footprints, street patterns and even the configuration of green spaces at the neighbourhood level. This multi-scale capacity makes imagery a flexible resource: it can illuminate both macro-level transformations such as urban sprawl (Brown et al. 2012b) and micro-scale differences in the built environment associated with health or wellbeing (Metzler et al. 2023).
Satellite missions provide regular and predictable revisits ranging from daily (e.g. PlanetScope) to a few days or weeks (e.g. Sentinel-2 and Landsat). This cadence allows researchers and policymakers to generate consistent time series and monitor rapid changes in near real-time. For example, vegetation indices derived from multispectral imagery can track seasonal dynamics in urban greenness, while radar imagery can detect flood extents immediately after a storm. Such timeliness is difficult to achieve with traditional data sources, which are often updated only every few years. There is often a trade-off in satellite data between resolution and frequency, but technological advances such as nanosat constellations1 are constantly pushing outwards the frontier at which this trade-off takes place.
Satellite programmes, such as Landsat, have been operating since the 1970s, creating an unparalleled archive of the Earth’s surface. These historical datasets allow researchers to reconstruct long-term patterns of land cover, urbanisation and environmental change, offering insights into trajectories that no survey or census could capture. The ability to ‘’look back in time’’ makes imagery especially valuable for understanding the cumulative effects of policies, economic shifts and climate change across decades (People and Pixels 1998a). It also allows to fill the gaps that the lack of temporal consistency of some traditional social datasets introduce.
Unlike survey data, which rely on voluntary participation and may exclude hard-to-reach groups, satellite imagery does not require human consent or response to be generated. This makes it relatively immune to self-selection bias and more representative across space. While imagery has its own challenges (e.g. cloud cover and sensor noise), its systematic and global character ensures a level of consistency that complements traditional data sources (Sherbinin et al. 2002).
Imagery is not limited to visible light. Satellites measure across the electromagnetic spectrum, producing data on heat, vegetation health, surface water and air pollutants. Synthetic Aperture Radar (SAR) captures structural features through cloud and darkness; and hyperspectral sensors enable fine-grained detection of material properties. This diversity of sensors allows for the derivation of novel indicators, for example, from rooftop solar potential to urban heat island intensity that can directly inform pressing policy agendas in sustainability, prosperity and wellbeing (Deilami and Kamruzzaman 2017).
A unique source of social data
Together, the above identified attributes position satellite imagery as a uniquely powerful form of smart data. Its comprehensiveness, resolution, timeliness, retrospective depth and sensor diversity create a multidimensional evidence base that is unmatched by conventional sources. What makes imagery transformative is not just its technical sophistication, but its capacity to bridge long-standing evidence gaps in the social sciences and policymaking, providing the foundations for more informed and timely decisions.
Urban greenspace is increasingly recognised as a determinant of health and wellbeing. Yet official statistics on its distribution are often fragmented, inconsistent or outdated. Satellite imagery provides a systematic alternative. Using freely available Sentinel-2 multispectral data, researchers can derive the Normalised Difference Vegetation Index (NDVI) to estimate vegetation cover at fine spatial scales.
By linking these imagery-derived measures to administrative health records, studies in the UK and elsewhere have shown systematic disparities in access to urban greenery, with deprived communities often having less exposure (Geary et al. 2023; Kim et al. 2023). For policymakers, this information is critical. It identifies “green deserts” within cities, supports the design of equitable planning policies and provides indicators to monitor the effectiveness of urban greening initiatives.
The key advantage is comprehensiveness and comparability. Unlike local audits or surveys, satellite imagery captures greenspace consistently across entire cities and regions, enabling cross-neighbourhood benchmarking and long-term monitoring. This makes it a unique input into strategies for healthy and sustainable urban development.
1.4 Why now? The timely confluence of technological advances and shifting societal needs
For decades, the potential of satellite imagery to transform the social sciences and inform public policy has been recognised (People and Pixels 1998b). Yet until recently, this potential remained largely aspirational. The barriers were formidable: imagery was expensive, technically complex and difficult to process at scale; expertise was confined to environmental sciences and engineering, and the social science community lacked both the tools and the training to make effective use of complex and unstructured imagery data; in addition, the state-of-the-art technology did not allow the detail required to identify features and processes of interest in social contexts. Recently, key developments have transformed the landscape. We are witnessing a convergence of technological, institutional and societal changes that, together, create the right conditions for mainstreaming imagery in social research and policy.
1.4.1 Proliferation and democratisation of satellites
The number of satellites orbiting the Earth has grown exponentially. After decades of roughly 150 satellite launches per year (1957–2012), annual launches surged to about 600 in 2019, 1,200 in 2020, and 2,470 in 2022 (United Nations Office for Outer Space Affairs (UNOOSA) 2023). Public programmes such as the European Space Agency’s Sentinel missions and NASA’s Landsat archive provide high-quality data freely and openly, while private constellations like Planet or Maxar deliver near-daily high-resolution imagery. Launch costs have plummeted, fuelled by commercial providers and advances in satellite miniaturisation (Sweeting 2018). The result is not only more satellites but more diverse sensors, including optical, radar, hyperspectral and thermal offering unparalleled coverage of the Earth’s surface. What was once the preserve of specialised agencies is now accessible to researchers, policymakers and even the general public.
1.4.2 Advances in resolution and sensor capability
Alongside this expansion of satellite availability has come a dramatic improvement in the quality of imagery. Spatial resolution has increased from kilometres to sub-metre detail; temporal resolution has improved to daily or even multiple daily revisits; and spectral resolution has grown into hyper-spectral imagery that enables the measurement of heat, vegetation stress, air pollutants and urban morphology. These advances expand the analytical frontier: for example, tracking micro-greenspaces (Ramdani 2024), monitoring the impacts of heatwaves on vulnerable populations (Buscail, Upegui, and Viel 2012; Johnson, Wilson, and Luber 2009) or estimating contamination in agricultural soils (Yao et al. 2024). Such detailed information is key for understanding the intersection of environmental exposures, health and social inequality at scales that matter for policy.
1.4.3 Revolution in computation and artificial intelligence
Raw pixels alone are not enough. The primary value of imagery relates to the transformation of raw pixels into meaningful indicators. This is what Imago terms the “pixel-to-metric” challenge. Until recently, this process was limited by computational bottlenecks. Advances in machine learning and computer vision, combined with the rise of cloud computing and high-performance infrastructures, have fundamentally altered this landscape. Modern neural networks can extract building footprints, classify land cover or estimate deprivation with remarkable accuracy (Arribas-Bel, Patino, and Duque 2017a). Cloud platforms such as Google Earth Engine make it possible to analyse terabytes of imagery without local supercomputers (Gorelick et al. 2017). And the open source revolution that has powered many of these advances has democratised access to knowledge that used to required extremely advanced domain training. These factors have all contributed to lowering the barrier to entry for social researchers and policymakers.
1.4.4 Easier access and integration with existing data ecosystems
Equally transformative are the changes in data access. Imagery has also benefitted from the rise of data science over the last two decades, and is increasingly delivered through portals, APIs and interoperable formats (Jacobsen et al. 2020). This accessibility means that imagery can be linked to household surveys, administrative records or longitudinal cohort studies, allowing researchers to integrate contextual measures of environment, housing or infrastructure into existing datasets. Such integration bridges the long-standing gap between social data on individuals and contextual data on places, creating powerful opportunities for spatially-explicit analysis and evidence-based policymaking. Yet, while an increasing number of satellite datasets are more accessible, their sheer volume and complex, unstructured nature remain a major challenge for most social scientists and policy makers to use and analyse imagery. This is a key barrier that Imago will tackle.
1.4.5 Growing societal demand for timely, granular evidence
Technological advances alone would not be relevant if demand was absent. The societal context has shifted dramatically. Some of the most pressing challenges, including the climate emergency, health inequalities, housing crises and uneven regional development, all require data that are timely, spatially detailed and robust. Policymakers seek indicators that can capture the dynamics of local communities, monitor change in near real time and evaluate the impacts of interventions. Imagery is uniquely placed to meet this demand, offering consistent coverage at scales from a national to neighbourhood scale. In this sense, the supply of new imagery technologies is met with an urgent demand for better evidence in sustainability, prosperity and wellbeing.
1.5 The case for social science and policy use
Every second of human existence is now recorded from space. The challenge is no longer availability. It is use. Despite a world of data being generated, a persistent gap remains between what satellites capture and what reaches policy makers’ hands. Pioneers have already shown how imagery can move the needle. Yet broader impact requires removing the frictions that keep this data out of reach. Imago bridges this gap.
1.5.1 Persistent gaps in evidence
Across social research and policy, an enduring challenge is the death of timely, reliable and spatially detailed data. Inequalities in health, prosperity and wellbeing often manifest at local scales: between neighbourhoods, across urban–rural divides or within regions. Yet, the data streams traditionally used to study these questions rarely provide the necessary resolution or frequency. Censuses are comprehensive but infrequent. Household surveys are costly and often geographically limited. Administrative data can be inconsistent or inaccessible due to privacy and governance restrictions. These gaps are especially acute in areas where policy demand is greatest, such as monitoring the uneven impacts of climate hazards, evaluating local housing markets or designing interventions to address health disparities.
Satellite imagery directly addresses these shortcomings. Its global, repeated coverage provides a spatial and temporal granularity rarely achievable with traditional data, offering opportunities to fill evidence gaps that constrain research and policymaking. Just as importantly, this value is entirely additive rather than substitutive. Satellite data complement traditional social data, help stretch their value and provide a multiplier effect that increases the value proposition of traditional data.
1.5.2 Applications across domains
Satellite imagery has already transformed empirical practice across a set of domains that sit at the core of contemporary social science and public policy. These advances reflect demonstrated, peer-reviewed applications where imagery has enabled new measurement strategies, sharpened causal inference, extended spatial and temporal coverage and provided actionable intelligence for governments and organisations. Below, we outline four domains where satellites have already moved the needle, illustrating both breadth and depth of impact.
Economic development and inequality
Satellite imagery has become a foundational tool in measuring economic activity, particularly in contexts where official statistics are incomplete, delayed or unreliable. Night-time lights (NTL), captured by the DMSP-OLS and VIIRS sensors, are now widely used as proxies for local economic performance, enabling estimation of GDP, income distribution and regional inequality (Henderson, Storeygard, and Weil 2012b; Pinkovskiy and Sala-i-Martin 2016, 2014). NTL-based approaches have supported policy evaluations ranging from infrastructure investments to post-conflict recovery.
More recently, advances in high-resolution optical imagery and machine learning have enabled prediction of poverty at unprecedented spatial detail. Deep learning models trained on multispectral and street-level imagery can accurately map consumption, housing quality and asset deprivation (Jean et al. 2016; Yeh et al. 2020). These methods have been integrated into development planning, supporting organisations such as the World Bank, UNDP and national statistical offices in targeting social assistance programmes and identifying underserved communities (Smythe and Blumenstock 2022a; Barriga-Cabanillas et al. 2024; Corral, Henderson, and Segovia 2025).
Imagery also supports monitoring agricultural productivity, land degradation and environmental shocks which represent key determinants of rural livelihoods. For instance, satellite-derived vegetation indices, such as the Normalized Difference Vegetation Index (NDVI) and Enhanced Vegetation Index (EVI) have been used to assess drought severity, crop yields and food security risk (Funk et al. 2015). These indicators underpin early warning systems across Africa and Asia, demonstrating the critical role of imagery in resource allocation and crisis preparedness.
Urbanisation, housing, and infrastructure
Satellite imagery enables detailed characterisation of the built environment, including settlement extent, building footprints, street morphology, and urban density. These indicators are critical for understanding patterns of urban growth, housing supply constraints and infrastructure needs. Research using Landsat, Sentinel-2, and commercial imagery has documented urban expansion, informal settlement growth, and land-use change across rapidly developing regions (Seto, Güneralp, and Hutyra 2012; Sorichetta et al. 2015).
High-resolution imagery and computer vision techniques now allow automated extraction of building outlines (Li et al. 2021; Ayala et al. 2021), road networks (Liu et al. 2024) and roof types (Ölçer, Ölçer, and Sümer 2023). These datasets support municipal planning by identifying infrastructure deficits, monitoring construction activity and quantifying exposure to hazards. Imagery also supports housing studies, for example, by detecting informal settlements (Kuffer et al. 2017; Mahabir et al. 2018), estimating built-up density or evaluating the distribution of impervious surfaces linked to stormwater runoff.
For policymakers, these indicators offer a consistent, scalable method to track progress toward Sustainable Development Goals (SDGs), especially SDG 11 (“Sustainable Cities and Communities”) (Persello et al. 2022; Q. Zhao and Yu 2025). They also enable comparative assessment across cities and countries (Guo 2024; Xi et al. 2023). Such assessment are not feasible with administrative records alone.
Environment, exposure and population health
Satellite-derived environmental indicators have become central to understanding spatial inequalities in health exposures. Multispectral imagery provides measures of greenspace, such as NDVI, tree canopy cover and blue space proximity, which are linked to mental and physical health, physical activity and cardiovascular outcomes (Rugel et al. 2017; Twohig-Bennett and Jones 2018). These measures have consistently revealed patterned inequities across socio-economic and racial lines.
Thermal sensors, such as MODIS and ECOSTRESS, allow mapping of land-surface temperature, revealing disparities in exposure to extreme heat which is an increasingly important determinant of morbidity and mortality (Lei Zhao et al. 2014). Imagery-derived air quality indicator, such as aerosol optical depth have been used to estimate PM2.5 concentrations, especially in regions lacking ground-based monitoring networks (Donkelaar et al. 2019). These data underpin global burden-of-disease estimates and guide national air pollution regulations. By linking environmental indicators with health records, imagery can support epidemiological analyses, inform environmental justice legislation and enable evaluation of interventions such as urban greening, cooling strategies or pollution controls (James et al. 2016; Weigand et al. 2019; Massaro et al. 2023).
Disaster response and climate adaptation
Satellite imagery has become a core component of disaster detection, monitoring and response (Akhyar et al. 2024a). Optical and radar sensors enable rapid assessment of flood extent, wildfire boundaries, landslides, earthquakes, and storm damage, especially in inaccessible areas (Mateo-Garcia et al. 2021; Shafapourtehrany et al. 2023a; Akhyar et al. 2024b). RADAR imagery (e.g. Sentinel-1) is particularly valuable for penetrating cloud cover during extreme weather events (Tebaldini et al. 2022).
These imagery products are routinely used by humanitarian agencies, civil protection authorities and insurers to inform crisis logistics, direct relief resources, and estimate recovery needs (Shafapourtehrany et al. 2023b; Cao and Choe 2023; Schmidt et al. 2025). Time-series imagery enables monitoring of post-disaster reconstruction, shoreline retreat, glacier melt and drought evolution, providing essential evidence for climate resilience and adaptation planning over decades (Friedrich et al. 2024; Le Cozannet et al. 2014; Bolch et al. 2012). Imagery also supports anticipatory action, such as monitoring vegetation dryness to predict wildfire risk (Pickell et al. 2017; Watt et al. 2025), tracking urban heat island dynamics (Imhoff et al. 2010; Lili Zhao, Fan, and Hong 2025)or identifying neighbourhoods with high flood susceptibility based on impervious surface extent (Ahmad, Shao, and Javed 2023; Tokarczyk et al. 2015).
1.5.3 Complementarity and integration
The greatest potential of imagery lies in data integration with other data sources. Imagery-derived indicators can be linked with household or cohort surveys to provide rich contextual measures of local environment, infrastructure and housing conditions. For example, linking greenspace indices to health records can illuminate associations between neighbourhood environments and mental wellbeing (Sarkar, Webster, and Gallacher 2018; Su et al. 2019; Araújo et al. 2024). Similarly, combining imagery-based poverty maps with demographic data can support more equitable allocation of resources (Jean et al. 2016; Smythe and Blumenstock 2022b).
This integrative capacity allows imagery to act as a bridge between individual-level data and the broader characteristics of places, enabling multilevel analyses that capture the interaction between people and their environments (Singleton et al. 2022; Cohen-Cline, Turkheimer, and Duncan 2015). It also aligns with the growing demand in policymaking for place-based evidence that reflects the lived experience of communities rather than national averages (OECD 2025).
1.5.4 Towards useful, usable and used imagery
The case for social research and policy use can be summarised as a matter of timing and translation. The technological advances described in Section 1.4 imply that, for the first time, imagery is poised to become a routine part of the evidence base. But realising this potential requires making imagery usable, useful, and used:
Useful, by co-producing data products with stakeholders to ensure relevance to research and policy questions.
Usable, by lowering technical barriers through open platforms, user-friendly interfaces, and interoperable data formats.
Used, by embedding imagery into established data ecosystems, training communities of practice, and demonstrating impact through visible policy applications.
By addressing these three pillars, imagery can evolve from a promising niche resource to a cornerstone of evidence-based social science and policymaking. The case is not only academic, it is practical and aligned with the growing demand for timely, localised and equitable data infrastructures.
Currently, informal settlements house over a billion people globally, with this number expected to increase substantially over the next 30 years (United Nations, 2023). Yet, they are often absent from official statistics and maps. This invisibility perpetuates exclusion, making it difficult for governments and international organisations to target investments in housing, sanitation, and health.
Satellite imagery offers a way forward. High-resolution optical data, combined with machine learning classifiers, can detect the dense, irregular patterns characteristic of informal housing. Studies in sub-Saharan Africa and South Asia have demonstrated that imagery-based maps of settlement extent align closely with ground surveys, but can be produced faster, at lower cost, and with full coverage (Kuffer et al., 2016; Mahabir et al., 2018).
For policymakers, this capacity is transformative. Imagery can reveal previously unmapped communities, track their expansion over time, and help allocate resources more equitably. By integrating imagery with household surveys or administrative records, it becomes possible to link population characteristics with environmental exposures, providing a fuller picture of vulnerability and need.
Importantly, this is also relevant in advanced economies such as the UK, where fine-grained spatial data can identify pockets of deprivation, housing precarity, or poor living conditions. Deriving granular insights from satellite data can enhance local authorities’ ability to design targeted, place-based interventions, aligning with the UK’s broader Levelling Up agenda.
This example illustrates the central case for imagery in social science and policy. It addresses critical data gaps in contexts where conventional sources are absent, unreliable or prohibitively expensive, enabling more inclusive and responsive decision-making.
1.6 Barriers and challenges
Despite its significant potential, the mainstream adoption of satellite imagery within the social sciences and public policy remains limited. These constraints stem from three broad and interconnected classes of friction: (1) challenges in processing imagery; (2) difficulties in translating pixels into meaningful indicators; and, (3) issues relating to the responsible use of high-resolution data.
Processing raw satellite data is non-trivial. Imagery is vast in volume, often encompassing terabytes of data for national or multi-year analyses and is stored in specialised formats that require dedicated software, libraries and computational environments. Many datasets—such as multispectral time series, thermal imagery or radar products—demand sophisticated pre-processing steps, including atmospheric correction, cloud masking, radiometric calibration or speckle filtering. These workflows vary across sensors and platforms, and they can easily exceed the capacity of standard desktop computing environments. As a result, users typically need access to cloud platforms, high-performance computing clusters or advanced geospatial tooling. For social scientists, who seldom have these resources readily available, the technical overhead represents a substantial barrier to entry.
Even when imagery can be processed, a second barrier arises in the form of a translation gap. Raw pixels are not interpretable by policymakers or practitioners. What is needed are validated, meaningful indicators, such as measures of building density, impervious surface extent, greenspace exposure or land-surface temperature, that correspond to socially and policy-relevant constructs. Producing these metrics requires expertise in computer vision, classification methods, model validation and spatial harmonisation, as well as domain-specialist knowledge. The intellectual task is substantial; that is, linking the physical properties measured by satellites to the conceptual categories used in social science and policy. This challenge is compounded by the limited presence of remote sensing and spatial data science training in most social science curricula. As a result, many potential users lack the methodological literacy required to engage with imagery-derived indicators, even when data products exist.
A final set of concerns relates to the responsible use of satellite imagery. Very high-resolution datasets can reveal fine-grained details of the built environment, raising ethical and governance issues—particularly when imagery is combined with sensitive administrative or demographic data. Although raw imagery does not contain personally identifiable information in a traditional sense, there are legitimate concerns regarding re-identification risks, the potential surveillance implications of frequent revisits and the use of high-resolution data to infer characteristics about vulnerable populations or communities. Ensuring responsible practice requires clear protocols for data access, model development and output release, robust privacy-preserving techniques, and governance frameworks that align with legal and ethical norms. These considerations are especially important when imagery is used to inform decisions about resource allocation, risk management or land use regulation.
Taken together, these technical, translational and governance frictions explain why, despite decades of data availability, imagery has not yet become a routine evidence source in social science and public policy. The barriers are not merely operational. They are structural. Without shared infrastructure, common data standards, interdisciplinary training and mechanisms for responsible use, imagery remains siloed within specialist communities. Overcoming these barriers requires sustained investment in computing infrastructure, in the development of analysis-ready data products and in the capacity of researchers and practitioners to work confidently with imagery-derived indicators.
Imago has been designed explicitly to confront these limitations. By producing analysis-ready data products, lowering technical barriers to processing, embedding interpretability through well-defined metrics and establishing best-practice protocols for responsible use, Imago seeks to bridge the gap between what imagery can offer and what researchers and policymakers can realistically use. Through this effort, the programme aims to move satellite imagery from a niche analytical tool to a mainstream component of the evidence infrastructure for social science and policy.
Fleets of shoe-box sized satellites that can obtain ever higher resolution at very high frequencies.↩︎