Image sourced by betterimagesofai.org
Introduction
The influence of climate change on different aspects of society and ecology is too complex to comprehend from a singular lens, place or point in time. Lahoud poignantly highlights the omnipresence of climate change, stating how it occurs in the periphery of our vision and at the limits of our sensorium and understanding. Conceptualisations, speculations, and interventions around climate change have been consistently evolving. The understanding of climate change moves across registers and scales—spatial (e.g., the so-called local and global) and temporal (e.g., historical, evolutionary, and media time). Existing at different intersections of scales, issues and temporalities, recent research approaches attempt to make the impact of change more tangible1.
Climate data—observed and simulated over time by different communities with different intentions—is gaining prominence as an approach to unpacking and acting upon climate change. The data-driven conceptualisation of climate change was initially undertaken predominantly by research scientists, but there has been a recent increase in interest in its use by stakeholders in other fields2. Organisations and individuals working in public health, social security, and agriculture are beginning to use climate data to inform their decisions, strategies and practices. These transitions are taking place on both global and local scales, producing different kinds of data, emphasising different impacts of climate change. There is an inherent friction between the different data-centric approaches, different positionalities, intentions and biases present in data sets, and analytical methods lead to different conclusions. An instance highlighting this friction is Bill McKibben’s “terrifying new math”, wherein he attempts to encapsulate the inaction surrounding climate change through three simple numbers, one of which is the 2° Celsius global temperature warning, climate change, agrarian distress, and the role of digital labour markets which will further exacerbate the effects of climate change3. The contemporary discussion relies heavily on this data point/statistic. However, heavily relying on this figure for climate action policy and frameworks can lead to the oversimplification of the impact of climate change. In his analysis, Lahoud refers to the 2-degree as a “violent abstraction of a global average”. He further states the number “negates the uneven scale of climate impact and erases the specificity of people from its calculation. A 2-degree average increase globally would allow for a catastrophic 3.5 degrees in many places”. Thus, as data-driven approaches to climate gain prominence, it is pertinent to understand how the complexity of climate change gets intertwined with the complexity of approaches to data.
As the conversation around climate action evolves, so do the approaches, which are becoming more complex and interpolating several data sources and methods of analysis. It has become increasingly important to recognise the role data can play as well as what it is envisioned to do; the spatial-temporal nature of climate change is intertwined with the practices of data collection and deployment, as highlighted by Theodore C Lim, who notes, “different perspectives in time and space, with the help of data and information, can result in very different prioritisation of social and environmental outcomes” (Lim).
This paper is divided into three sections; the first section unpacks how climate action requires multiple stakeholders and how the different stakeholders create different kinds of data. The following section goes on to unpack the approaches to climate data broadly defined as top-down and bottom-up approaches through case studies highlighting the strengths and limitations of the approach. The final section examines data-driven approaches to create accountability for different stakeholders. The conclusion addresses the friction between the various approaches to climate data and the challenges of over-emphasising climate data.
Climate Action different actors, different approaches
Environmental challenges ranging from resource conflicts, conservation and extraction, and pollution are complex socio-ecological problems 4. Solving them involves a constant process of collaboration and negotiation between various stakeholders (Lim). Finding and operationalising adaptation and mitigation strategies requires stakeholders to work together at different stages. Lim suggests that the multistakeholderism of climate action requires a governance model rather than a singular statutory body5.
As the conversation around climate action broadens, engaging with more stakeholders such as non-government organisations, community groups, and individual citizens brings different understandings, insights and approaches to the table6. This permeates into the practices of the data-centric approaches; different perspectives on the role of climate scenarios, the types of information included, and the processes for identifying vulnerable areas have important implications for data production, which is a crucial factor for how the data can be used and where it can be used. As the data-centric approach becomes more prominent, further questions need to be asked about what role data will play, what kind of data will be collected, to what end, and by whom. Different perspectives of the stakeholders raise different sets of challenges and insights, some of which might help paint a coherent picture of the impact of climate change.
Approaches to climate data and its usage can be clubbed into two broad categories: top-down and bottom-up approaches. Both approaches involve various methods, giving insights into different aspects of climate change. To understand what purposes the two approaches are used for, the kind of data they offer and their strengths and limitations, the following section will focus on practices of i) vulnerability assessments(VA) and ii) observation and monitoring, which are undertaken in both approaches. There are several methods of data collection and analysis within the top-down and bottom-up approaches(the notes include methods are referred to in the case studies7).
Top-Down Approach
The top-down approach refers to VAs and observing and monitoring efforts defined within the context of a global, international, or national framework.
These approaches typically define essential variables that link to broad societal benefits and specific agency or operational missions. Furthermore, they often consist of large-scale programs for creating climate model projections to assess physical and ecological impacts; multiple projections are used to assess ranges of uncertainty for future states.
Technological advancements, especially in remote sensing and cloud computing, are allowing researchers to better monitor changes to ecology and the impact of natural disasters cost-effectively and more frequently than on-ground approaches and surveys are able to. The increasing accuracy and spatial resolution can detect fine-scale changes indicative of ecological changes. Variables such as clouds, aerosols, GHGs, land use and land cover changes (LULCC), glacier ablation, water and energy flux, and sea level change, which have become vital to our understanding of climate change, are monitored and dissected through remote sensing8. The approach is not without limitations, especially with regard to the accessibility of the technology and the kind of data it provides in its deployment.
For instance, the Global Forest Watch (GFW), an online platform run by the World Resources Institute, draws on Google software and academic expertise to display national and sub-national data layers on forest cover, forest loss and areas licensed for agriculture plantations and mining concessions. Such mapping
Monitoring: It involves tracking a particular variable or phenomenon over time to identify trends that require some type of action—for example, for adaptation or resource management (Eicken et al.).
Observing: Recording data for a particular variable or process, often in the context of scientific research programs devoted to understanding and predicting the behaviour of a particular system (Eicken et al.).
Monitoring efforts are often conditioned by the parameters of the technology itself and the situated politics of resource use, value, and governance. The lack of technical capacity is a big hurdle for stakeholders situated in the Global South or low-resource areas in accessing platforms such as GFW. In conversation with Yangon conservationists, Goldstein shows how the resolution is too low to guide on-the-ground interventions. Several organisations in Myanmar are even dismissive of the platform as they observed that GFW’s objectives to map forest change and concession boundaries do not align with the group’s objectives to report on community issues. Other issues highlighted concern unstable or slow internet connections, making it difficult to access the data and local environmental organisations’ lack of capacity to upgrade existing platforms.
Bottom-up Approach
Focusing on efforts defined and undertaken at the local scale, bottom-up approaches aim to align with the desired outcomes for the local communities9. It is a combination of attributes relating to local initiatives and establishing observations, and community involvement in data acquisition and analysis.
Vulnerability assessments at the local level provide critical insights into location-specific vulnerabilities 10. They primarily have two dimensions: first, the close involvement of local stakeholders; and second, instead of using global model scenarios far into the future, the assessments should examine vulnerabilities to current climate variability and extremes, as well as the current adaptation strategies, policies and measures, based on experience at different scales.
The limitation of bottom-up approaches is that scalability, practices and strategies are highly contextual to different communities and geographies.
Furthermore, the weaker linkages between local and international sources of climate information and tools are unable to calibrate to the demands at the community level, often causing confusion and anxiety. This also leads to reducing the knowledge produced by local communities to simplistic projections that fail to capture the complexities and uncertainties specific to certain regions, which in turn create issues during the operationalisation of interventions.
The vulnerability assessment in Sinazongwe in the Southern Province of Zambia brings to light the strengths and shortcomings of bottom-up approaches. With an arid climate and uneven rainfall, Sinazongwe, is a disaster-prone district with three consecutive years of poor rainfall. This has led to severe food insecurity and triggered individuals to sell valuable assets, such as livestock, exacerbated vulnerability and people’s coping capacity. There was also an observed increase in cases of HIV/AIDS. The vulnerability community assessment highlighted how the issues were closely interlinked; drought directly affects water quantity and quality and can also result in poverty. Indirectly, food shortages related to droughts may force women into commercial sex, thus increasing the risk of HIV/AIDS. The VCA also noted that disasters increase the district’s dependency on external assistance to mitigate disaster impacts.
The community approach was not only limited to highlighting the challenges to the community, but also came up with actionable adaptation strategies. A number of drought risk measures were put in place, such as water harvesting schemes, irrigation schemes, and sustainable agriculture practices, as well as the introduction of environmental conservation awareness in schools and communities (leading to reforestation efforts). Among key capacities to deal with disasters, communities also emphasised alternative livelihoods rather than specific risk mitigation measures.
The issues with the insights that emerge from Zambia are two-fold. First, the assessments do not reflect explicit attention to trends beyond what is already experienced by the communities. Second, the challenges that emerge with scaling the results from vulnerable community assessments to inform risk reduction or adaptation efforts beyond the communities where the assessment has taken place. Since VCA’s are often done with the pre-requisite of the site and location being highly vulnerable—due to limited capacity and funding— they do not provide representative samples that can inform broader risk reduction policies and programs.
Considerations for a data-driven approach
One of the outcomes of using new sources and approaches to data collection and analysis, as well as increasing accessibility for data-driven climate governance, is strong systems of accountability. Sara Hughes’s work on how data-driven climate governance has implications for the notion of accountability for city governments can also be extended to national and global stakeholders. Her accountability framework focuses on three rationales for prioritising and strengthening data-driven decision-making: standardisation, transparency and capacity building. The rationales can be pivotal in shaping and influencing the efficacy of initiatives as they help ensure political feasibility and legitimacy, focusing on the engagement of multiple stakeholders’ nuanced approaches to data governance. Practices such as sharing international reporting frameworks, efforts to increase the usability and accessibility of public data, and building city government capacity—in our case, local capacity— for processing data as well as embedding social justice considerations into data-driven policymaking become ways of rethinking and amalgamating top-down and bottom-up approaches. Cadasta’s partnership with the Karen Environmental and Social Action Network (KESAN) in Southeast Myanmar within the Indo-Burma Biodiversity
Hotspot highlights how capacity building for different stakeholders helps bridge different kinds of data on social and spatial vulnerability to climate action. The project focused on operationalising data for policies and interventions as well as creating better environmental governance mechanisms. Rooting their approach to indigenous knowledge and operationalising modern technology to promote ecological stewardship led to better adoption and deployment. By providing support in setting up land demarcation methods and submitting data directly to online platforms, Cadasta helped communities self-document agricultural plots and community forests. KESAN successfully mapped over 3.5 million hectares of land using participatory methods, leading to the designation of forest reserves and community forests and the issuance of over 107,000 self-governed land title documents. These land titles are pivotal for local Karen communities to defend their land rights and engage in community-led conservation efforts, thus showing how remote communities can leverage technology for land protection 11.
Developing approaches to accountability in data-driven governance needs to emphasise the importance of publicly available data. Only with accessible data can public watchdogs and civil societies hold different stakeholders accountable for their actions. Data as a transparency mechanism thus becomes even more critical to building and growing trust. Returning to Hughes’s rationales, they also show the friction between top-down and bottom-up approaches. With standardisation, for instance, the trade-off emerges between uniform measurements at the different scales for comparing and benchmarking performance and being able to capture and highlight local nuances. Similarly, transparency in conforming to different global and national standards can lead to limiting and redirecting resources and distort regional understanding of climate change impacts. Thus, the question of rethinking data flows among different actors and scales to what end we need data and where cooperation and collaboration among different actors becomes more critical.
Conclusion and rethinking data for the environment
One of the implications of data-driven environmental governance is the assumption of data as a neutral, objective resource for decision-making.
Confronted with the limitations of existing “knowledge infrastructures” on climate change and conservation, data technologies ranging from drones to dashboards are envisaged to give more direct access to “raw” information and make “invisible problems not only visible but solvable” because the results of management can be more precisely and rapidly measured. The Global Forest Watch also similarly articulates the issue, stating, “It’s hard to manage what you can’t measure”. Unpacking the logistical and physical infrastructure enabling these approaches in light of who has ownership and control over climate infrastructures and the resources of knowledge produced highlights the inherent power dynamics. The continuous cultivation and uneven distribution of relevant technical expertise may create an asymmetric relationship between those who collect, store and mine large quantities of data and those whom data
collection targets. Despite the rhetoric of ethereal “cloud” data infrastructures, data-driven environmental governance does not unfold in abstract space. For instance, many conservation NGOs and tech start-ups require data storage and analytical capacities beyond their own, so they turn to energy-intensive server farms in data centres. These globally distributed facilities rely on access to cheap water and energy and are networked by international undersea cables. Thus, it becomes critical to understand the impact of a data-centric approach on the environment and how the approach has an impact on ecologies, especially in the Global South.
Data collection cannot be considered the primary reason behind the gaps, nor the sole reason for the inaction around climate change. Issues surrounding limitations in data accessibility and the absence of data-sharing structures throughout the value chain become significant factors. What requires more focus is who collects data, how it is used, who else can access and use it, but most importantly, for what end is it going to be used for: prediction or intervention. Equitable data-driven environmental governance requires facilitative structures and practices that allow data to be used and shared, prioritising transparency and capacity. Without such practices and approaches, data-driven decision-making can be disempowering when data is inaccessible or conceals spatial and demographic variation and inequalities.
Addressing the friction between top-down and bottom-up approaches furthers the question of rethinking conceptions surrounding data governance. Creating a hybrid approach that combines top-down and bottom-up approaches can be one such way. Using the outputs from top-down approaches to feed into the bottom-up approaches, thus increasing the skill of bottom-up approaches.
Furthermore, this creates a continual process through which both top-down and bottom-up approaches inform each other conceptually and practically, generating hybrid methods and information which can be useful in the short and long term. We must also evaluate data collection and analysis methods as dialectic processes where different approaches provide different types of insights, creating space for many types of data to be produced for stakeholders with different data needs and helping identify priority areas for further investigation and assessment. For these scenarios to play out, the role of knowledge brokers is central to knowledge synthesis and communication to inform practical actions 12.
Combining the several vantage points in understanding climate change, Emily Scott highlights the multiscalar, multitemporal, multidimensional, and multidisciplinary. Creating a network of actors working together to identify and address complex societal challenges becomes pivotal for a multi-disciplinary approach to work. Providing and analysing data of interest to different stakeholders, directly or indirectly, can help in expanding networks of data-based governance. Conway has provided a different approach which tries to reimagine the sectoral approach as a ‘systems of receptors rather than conventional sectors’, allowing for a more holistic approach to multidisciplinary methodology. This will help bring approaches and research tools that increase transferability and draw correspondences between indigenous and local knowledge’s qualitative and interpretative nature into categories that allow generalisation. Finding ways of enabling knowledge integration, combining inputs from multi-site place-based research and addressing scalability to ensure that place-based information effectively upscale to climate research becomes more prevalent than going after the obscure “data gap”.
Reconceptualising climate data, thus, moving beyond unpacking the problems and predictions on global scales to finding contextual solutions and practices, requires furthering our understanding of climate data flows and infrastructure.
Doing so will require unpacking further questions surrounding responsibility and ethics in the approaches. Frictions between different datasets, issues with how the data is produced and analysed, and the environmental impact of the processing itself are just a few of the broad issue areas to be investigated further. Additionally, methods and practices in different climate spaces will need to take a step back in order to come to resolutions which allow data flows and governance mechanisms to evolve as a patchwork, weaving together different perspectives and practices in negotiating with the multitude of challenges that arise with climate change.
1 See Surie, A., & Sharma, L. V. (2019) and Singh, C., Tebboth, M., Spear, D., Ansah, P., & Mensah, A. (2019) for insights on unpacking new methods such as life history approaches at the intersection of migration,
2 See Garica et al. on understanding how natural sciences focus much more on climate model helpful in assessing global magnitude and what pathways do more locally grounded research open for understanding climate change
3 Levinson (2013): The “three simple numbers” — the 2° Celsius that world leaders have agreed global temperatures can rise but which scientists call “a prescription for long-term disaster”; the 565 gigatons of carbon dioxide scientists estimate we can pour into the atmosphere before irreversibly damaging life systems; and the 2,795 gigatons already contained in the fossil fuel pipelines
4Rose and Chang (2023), socio ecological problem requires social ecological data— ‘society and their habits, traditions, and beliefs’ (Cambridge University Press n.d.). They feature structured and statistical cultural, ethnic, religious or demographic information, as well as (more subjective) data on cultural practices, day-to-day activities and routines, and beliefs—which requires Combining different types of data sources and stakeholders allows for triangulating information and complementing analyses with different perspectives.
5 See Anguelovski 2011 and Okereke, Bulkeley, Schroeder; for better understanding of climate governance at different scales ranging from transnational climate governance and urban climate governance
6 Although the piece does not delve into citizen science some insightful work see Albagli (2022)
7 Vulnerability assessments(VAs): Vulnerability assessments—systematic examinations of who is vulnerable, to what and why—are a widely used instrument comprising a broad group of tools with varying characteristics and goals (Næss).
8 See Yang, J., Gong, P., Fu, R. et al. (2013) for more insights on remote sensing technologies promise users easy and quick access to process large geospatial datasets. The website highlights how different stakeholders, ranging from journalists, policymakers, conservation organisations, and companies, can us the platform. Platforms such as GFW support functions via a suite of techniques that allow users to move seamlessly between the global scale of environmental visualisation and the local scale of political action.
9 Reyes-García V, Álvarez-Fernández S, Benyei P, García-del-Amo D, Junqueira AB, et al. (2023) highlighting the importance of the involvement of local communities in data collection and analysis
10 See Næss (2006) work on Norway
11 Other work by Cadasta and Forensic Architecture which also focuses on collaborating with community for climate justice
12 Browning (2019) A case of the need of data sciences in understanding climate change “he information sciences and the tech industry have an unfortunate history of sometimes jumping in to “solve” longstanding social problems without thoroughly researching them or consulting the affected communities. We must be wary of technological utopianism and of assuming our own experiences are universal, and we have to acknowledge that information and data science can help address climate change, but only in collaboration with environmental science helping us understand the scope and nature of the problem and specific communities collaboratively participating in the design of solution”
