A climate risk management typology: Integrating approaches to reduce risk

Abstract

Overshoot is increasingly likely to be required to meet the Paris Agreement's 1.5 °C temperature goal, resulting in climate risk management becoming more complex. This is aggravated by an often siloed and fragmentary risk management landscape, making the risks and benefits of potential interactions between interventions less legible to practitioners, policymakers, and academia. Approaches that integrate thinking across all climate risk management interventions are emerging as an increasing priority. We present a qualitative climate risk management typology that aids in exploring the interactions and roles of climate interventions across a variety of climate risk scenarios and contexts. Use of the typology is demonstrated by presenting six portfolios of interventions, which illustrate the utility of different interventions to ultimately reduce loss and damage. This approach builds on existing tools in the literature by recasting them explicitly in terms of risk and extending their range over a variety of qualitatively distinct climate scenarios.

Introduction

Achieving the 1.5 °C temperature goal enshrined in the Paris Agreement requires rapid and deep decarbonisation (IPCC, 2023). The remaining carbon budget for limiting warming to 1.5 °C is rapidly shrinking, with the level of human-induced global warming already at 1.31 °C above pre-industrial levels in 2023 and current rates of warming placing the world on a path to reaching 1.5 °C by the mid-2030s (Forster et al., 2024). Even before the 1.5 °C target is met, natural variability means that individual years are increasingly likely to temporarily exceed 1.5 °C (Hermanson et al., 2022), with 2024 already reaching an average of 1.55 °C ± 0.13 °C (World Meteorological Organization, 2025). Therefore, risks resulting from short-term warming of 1.5 °C will soon start to emerge, often in more vulnerable and exposed socio-ecological systems, before chronic impacts from sustained warming above 1.5 °C are realised over the long-term (IPCC, 2021, 2022b).

Thus, while not inevitable, it seems increasingly likely that limiting warming to 1.5 °C will require an overshoot, in which the peak global temperature initially exceeds 1.5 °C before returning to, or below, this target (Climate Overshoot Commission, 2023; IPCC, 2023; United Nations Environment Programme, 2024). The greater the delay in decarbonising the global economy, the greater the scale and speed required from various climate interventions to ultimately limit long-term warming to 1.5 °C (IPCC, 2023). As a result, interventions that might have especial utility in overshoot scenarios – including greenhouse gas removal, solar radiation modification, drastic emissions reductions, or extreme adaptation – are attracting increasing interest (Browne, 2022; Fialka, 2020; Fuss et al., 2018; Long and Shepherd, 2014; MacMartin et al., 2018; McLaren and Corry, 2021; Morrison et al., 2022; Reuters, 2022; Rogelj et al., 2018; Smith et al., 2024; United Nations Environment Programme, 2023; Warszawski et al., 2021).

These developments are emerging in an existing climate risk management system, which historically has tended towards a siloed or even fragmented approach to reducing risk. Evidence of this is clear in the international climate structures: recall the separate working groups of the IPCC or the differentiated negotiation streams of the UNFCCC (Stocker and Plattner, 2014; Wamsler et al., 2020). Interactions between potential climate interventions can thus seem to be a zero-sum game; for example, greenhouse gas removals and drastic emissions reductions are often positioned in opposition to each other (Brad and Schneider, 2023; Grant et al., 2021; McLaren et al., 2023).

While such trade-offs between interventions can exist, particularly in the context of resource constraints, interventions may also interact in ways that are synergistic (Warren, 2011). Further, since interventions are systemically connected, approaches that do not holistically consider interventions and their interactions may fail to capitalise on potential synergies or, indeed, create more risk as a result of unexpected interactions (Grafakos et al., 2019). We argue that the increasing urgency and scale of the requisite global climate response means that interactions between interventions are increasingly non-trivial and must therefore be carefully considered.

There has been a growing recognition that this exercise might be aided by systemic approaches that synthesise and integrate thinking across these interventions (see, as one example, the IPCC's Synthesis Report) (Grafakos et al., 2019; IPCC, 2023; Wamsler et al., 2020), and we aim in this paper to contribute towards this nascent movement. Accordingly, we present a qualitative tool to construct risk management scenarios that visualise the magnitude of, and potential responses to, a particular risk. To build a scenario requires considering six ‘high-level factors’ that affect the management of the risk. The first three high-level factors set out the magnitude of the risk in question; the former three are interventions, which are actions that aim to decrease the risk. Each high-level factor has a set of archetypal ‘instances’, which are qualitatively distinct ways in which the high-level factor might manifest. ‘Portfolios’ are combinations of interventions that respond to and exist within a particular risk scenario. Presenting all interventions within one diagram, and specifically in terms of risk, is intended to act as a stimulus for risk management discussions, as users are required to assess the various interventions, impacts, dynamics, and interactions, within an intervention portfolio for a given risk management context.

We provide an illustrative example of six scenarios constructed using this typology, and a method for creating sets of scenarios to support other risk management questions. This climate risk management typology builds on the ‘napkin diagram’ proposed by Long and Shepherd (2014), extending it to accommodate a range of futures and interventions. Crucially, the typology proposed here recasts warming scenarios in terms of risk, instead of temperature. This shift front-lines important discussions on how to characterise the risk (Reisinger et al., 2020), the complex components involved in assessing it, and the values and uncertainties that can often be hidden while focusing on temperature outcomes.

We present example diagrams to illustrate one particular way of thinking holistically and systemically but emphasise that they are a small subset of the much larger space of possible scenarios and portfolios laid out here and made available by the full typology. By combining ‘instances’ from each ‘high-level factor’ when analysing risk management futures, scenarios can be crafted for other values, scales, and subjects as required. In this paper, we focus on climate change, but the typology could be equally well deployed for other risk management challenges, such as those related to biodiversity or public health.

A qualitative tool can be a useful addition, both to help frame quantitative work, and create an agenda in research that helps eliminate blind spots. A tool like the climate portfolios aids in highlighting possibilities and risks but is not designed to make final decisions; in that way, it is a map rather than a navigator (Edenhofer and Minx, 2014). Indeed, the typology can be particularly useful in the setting of collaborative workshopping exercises, particularly those comprising a broad set of stakeholders. Using the typology to frame discussions about risk might aid stakeholders in sharing softer knowledge and coming to common understanding of the interactions between each component in a diagram (Mechler et al., 2019).

Linked to this, no tool is value-free (the role of the mapmaker is not apolitical), and we recognise that the scenarios touch upon a number of contested subjects (Cox et al., 2020; Kopp et al., 2025; McLaren and Corry, 2021; Waller et al., 2020). However, in requiring consideration and explicit representation of the risks and uncertainties involved with a particular choice of intervention (rather than temperature impact alone) when using the typology to build scenarios, we hope that the tool may be useful in underlining the calls for a global moratorium on the deployment of methods that as yet are insufficiently understood, such as solar radiation modification (Biermann et al., 2022), as well as the need for a more integrated global governance system to coordinate workstreams.

The climate risk management typology

The climate risk management typology introduced in this article builds on Long & Shepherd's ‘napkin diagram’ (Long and Shepherd, 2014), see Figure 1 (Shepherd), which was created to illustrate one potential role of various interventions (including emissions reductions, greenhouse gas removal, and solar radiation modification) in limiting the rise of global mean temperatures, and the roles of adaptation in reducing the remaining impacts on loss and damage. The napkin diagram is positioned at a global level and considers a timescale over 200 years (between 2000 and 2200). Its clarity in conveying its message has meant that it has been adopted in the climate community and literature (Belaia et al., 2021a, 2021b; Reynolds, 2019).

Figure 1.

The napkin diagram. The diagram was created by Long and Shepherd to illustrate that climate change should be responded to via a portfolio of approaches, incorporating emissions mitigation, carbon dioxide removal (CDR), solar radiation modification (SRM) adaptation, impacts, and suffering. Reproduced from Shepherd, no date, with permission.

The typology introduced here extends the initial ‘napkin diagram’ in three ways. First, it explicitly considers both overshoot and non-overshoot pathways, which increases the relevance of the ‘napkin diagram’ to the current state of the climate system and action (IPCC, 2023). Second, the typology extends the diagram by explicitly incorporating a range of qualitatively distinct trajectories for mitigation and adaptation responses over time (as laid out in Table 1), instead of the single scenario in the ‘napkin diagram’ (which depicts overshooting then net negative emissions, with temporary solar radiation modification deployment to flatten the underlying overshoot, and adaptation increasing over time before stabilising); the tool introduced here therefore provides greater clarity in the role of all available options and how they might interact. Third, we modify the ‘napkin diagram’ by recasting the primary variable as risk instead of temperature, which has a number of important benefits for integrated risk management approaches.

Table 1.

High-level factors and their instances in the climate risk management typology. This table sets out the climate risk management typology that aids in exploring the interactions and roles of climate interventions across a variety of climate risk scenarios and contexts. The approach to risk management is structured through the use of ‘climate risk management scenarios’, which contain six ‘high-level factors’ that might be useful in understanding the magnitude of, and responding to, the relevant risks. A short description of each high-level factor is set out in this table. Each high-level factor has a set of ‘instances’ attached to it. The instances span the set of baselines, interventions, climate responses, and risk outcomes that we perceive as likely to be strategically relevant in constructing risk management scenarios. The first three high-level factors (‘Reason for Concern’ / ‘Risk Thresholds’; ‘Risk Function’; ‘Baseline’) show the type and magnitude of the risk. The final three high-level factors (‘Mitigation’; ‘Potential Solar Radiation Management Deployment’; ‘Adaptation’) are ‘interventions’, that is, mechanisms for responding to the risk in question. A combination of interventions creates a ‘climate risk management portfolio’.

First, this risk framing accounts for uncertainty in the temperature-risk coupling (Seneviratne et al., 2018) and also increases the relevance for decision-makers for whom risk outcomes are ultimately what matter most (United Nations Framework Convention on Climate Change, 2015). Second, this focus on risk enables consideration of risks relating to a variety of systems and geographical scales (by choosing a locally defined risk indicator, for example), and the co-dependence between local impact and global and local actions to be concisely expressed. Third, this risk framing adds utility by integrating the outputs of the three working groups of the IPCC; the typology represents at a high level the risks generated by changes in climate-related hazards (WGI), the coupling between climate hazards and risk (WGII), and the relationships between potential risk reductions from mitigation (WGI and WGIII), adaptation (WGII), and solar radiation modification, and the consequent risk of loss and damage (WGII).

Multiple types of risk emerge from climate change, and their exploration is supported by this typology. We take as our starting point the IPCC's understanding of risk: the ‘potential for adverse consequences for human or ecological systems’, determined by the interactions between hazards, vulnerability and exposure (IPCC, 2023). We define physical climate risk as the emergent risks due to the state of the physical climate system after all mitigation activities that physically alter its composition or state have been factored in, but before the resulting risks to society have been reduced through adaptation. The specific physical climate risks that would be incorporated are dependent on the choice of risk variable in a scenario. We define societal climate risks as the remaining risks faced by society after adaptation measures have reduced the exposure and vulnerability to hazards from the physical climate, a definition therefore synonymous with the risk of loss and damage (Roberts and Pelling, 2018). An intervention is defined here as an activity explicitly taken to minimise physical or societal climate risk and is therefore additional to activities that might have occurred anyway without any risk-limiting intention but may reduce risk as a side effect. For example, carbon dioxide removed from the atmosphere through active reforestation efforts would be included, whereas removals resulting from carbon fertilisation (i.e. the increased carbon uptake of plants that comes along with increased atmospheric CO₂ concentrations) would not be (Nolan et al., 2024).

High-level factors: The building blocks of a risk management scenario

A selection of six climate risk management portfolios is presented in the next section; these were constructed using the climate risk management typology that is detailed in this section. The portfolios were chosen to illustrate, at a high level, the potential role of various interventions in addressing risks resulting from climate change. However, the utility of the typology is that it allows for the construction of arbitrary sets of portfolios that support risk management for a much wider range of scenarios. Here the typology is focused on climate change but could be directed at other risk management contexts.

In this section, we present the typology of six high-level factors that can be used to build any generalised climate scenario as follows and as is presented in Figure 2. First, the type of risk is chosen for the vertical axis, and potential threshold levels of risk are identified and plotted as horizontal lines. Further, since the figure is cast in terms of risk, a risk function is required to convert the various amounts of a given intervention into risk outcomes for the given risk type. Next, the baseline trajectory is selected in order to give a baseline of what the outcome might have been in the absence of any interventions. The mitigation wedge is then included, based on trajectories of emissions, reductions, and removals. The deployment (if at all) of solar radiation modification is inserted between the mitigation and adaptation wedges. Lastly, the adaptation wedge is inserted, dependent on perceived ability and capacity of adaptation measures to decrease the level of vulnerability and exposure to physical risk over time. The remainder of the graph under the adaptation wedge refers to the risk of unavoidable loss and damage accruing from the risk type that could not be avoided or adapted to.

Figure 2.

Constructing a climate risk management scenario. Here we set out the process for creating a climate risk management scenario through six steps. Each step corresponds to a high-level factor (explained further in the main text and in Table 1). Choices must be made at each step about which ‘instance’ of the high-level factor will apply (instances shown here as graphs). To illustrate how this coheres to a scenario, one example representation of a climate risk scenario (Figure 3’s ‘Plan C’ scenario) is included on the right of each step; further climate scenarios are included in the main text. To construct a climate risk management scenario, a specific risk is chosen for the vertical axis (risk represented here as the choice of axis), and any useful thresholds for this risk can be depicted with horizontal lines (represented here as three lines in purple, red, and orange) (step 1). A risk function must be selected to convert the adoption or absence of interventions into quantity of risk increase or decrease (step 2). Following this, a baseline trajectory must be selected to use as a benchmark against which interventions can be measured (here the linearly increasing instance of the baseline is chosen and is represented in the graph by the black line. Note how this effects the magnitude of the risk, effectively setting up a risk ‘envelope’ that is relevant for the scenario) (step 3). From this, interventions are inserted to decrease the risk within that envelope (steps 4–6). First, mitigation interventions are inserted (blue wedges), followed by solar radiation management where relevant (teal, here not inserted); and finally, adaptation inventions (here in yellow).

For each of the high-level factors, a stylised set of qualitatively distinct ‘instances’ are identified (see Table 1), which together span the set of baselines, interventions, climate responses, and risk outcomes that we perceive as likely to be strategically relevant. The high-level factors and varying instances are set out in Table 1 and explored further below.

Specific archetypal instances of the high-level factors can be combined to create a climate risk management scenario, which illustrates the portfolio of measures that are necessary to ultimately decrease the risk of loss and damage. From the combination of instances in Table 1, there are approximately one thousand potential qualitatively distinct combinations, though not all are possible, likely, or deemed to be relevant.

The climate risk management scenarios vary depending on the qualitative structure of interventions required for limiting risk to a certain level, rather than quantitative differences in the amount of activity relating to each intervention within scenarios that use the same selected set of instances from the high-level factors. Potential interactions between different types of intervention are accounted for by the interventions being combined into a portfolio; this allows potential concerns over resource constraints to be considered, for example.

While the typology can be applied to globally averaged risks and related metrics, it can also be applied in a way that explicitly addresses local contexts and geographical nuance by carefully defining the risk considered on the vertical axis. For example, the drought risk to a city could be used as the risk variable on the vertical risk, with incremental, systemic, and existential thresholds relevant to the particular urban area. Mitigation and solar radiation modification deployment could be global activities in the scenario, while the locally specific realisation of physical climate risk, adaptation responses to the given risk, and resulting remaining loss and damage would apply specifically to the city's locality. This helpfully expresses co-dependence between global and local actions and unifies them in one visualisation for risk managers to demonstrate how global activities would link with locally realised impacts.

Risk type and thresholds

When building out a scenario using the typology, the preliminary step is to identify the extent or type of risk considered. Conceptually this underpins the entire exercise; visually, it is reflected as the variable on the vertical axis. Any risk could be chosen: for example, the burning embers employed by the IPCC to visualise key Reasons for Concern (as presented in IPCC AR6 WG2 SPM Figure SPM.3) (IPCC, 2022b; Marbaix et al., 2025; Zommers et al., 2020). Note that beyond global concerns (such as the global risk of extreme weather events occurring, risk to the stability of large-scale components of the Earth system, or global biodiversity loss), more localised risks may also be used (such as a city's risk of flooding due to sea-level rise over time, risks of local crop loss due to droughts, or the climate-induced risk of local increase of vector-borne diseases), which broadens the utility of the typology across a range of scales and risk-related problems.

Given a particular risk variable, it can be useful to set stylised risk levels as a visual mechanism for incorporating into risk management any concerns relating to the degradations or thresholds of risk. These correspond to certain levels of risk on the vertical axis and can therefore be depicted as horizontal lines on the figure (e.g. Figure 3). For our purposes, we have categorised risk thresholds into three degradations: incremental; systemic; and existential. However, the boundaries between each risk threshold are not certain or set, so care should be taken while choosing and communicating them. Examples of threshold setting could include the planetary boundaries framework, or estimated thresholds for tipping points that may exist in physical or socio-economic systems (Armstrong McKay et al., 2022; Malhi et al., 2009; Steffen et al., 2015).

Figure 3.

An illustrative set of six climate risk management scenarios. To construct a scenario, a specific risk is chosen for the vertical axis, and any useful thresholds for this risk can be depicted with horizontal lines. The black line refers to the choice of baseline trajectory, and the wedges depict the risk reductions relative to this baseline resulting from various interventions, including mitigation (blue wedges), potential solar radiation modification (green wedge), adaptation (yellow wedge), and the remaining unavoidable loss and damage (pink wedge). Arbitrary sets of scenarios can be constructed using the typology introduced in this paper, depending on the specific risk being considered, and the sorts of integrated risk management scenarios that are being explored. Note that the variable on the vertical axis is risk, and therefore risk reduction, resulting from interventions is visualised — not the amount of an intervention used, or the temperature impact resulting from these interventions. The set of six scenarios presented here provide an example use of the typology, depicting the roles different interventions may play; as such the risk type, risk function, and thresholds are kept the same across each depicted scenario.

Risk function: Coupling between interventions and risk reductions

After identifying the type and extent of the risk, the second step is to account for the coupling between interventions and risk reductions through the risk function. This coupling incorporates complexity in the relationship between a given intervention and the resulting risk reduction, which requires careful consideration of the uncertainties, dynamics, and potential non-linearities in the system connected with the specific risk under consideration. This coupling acts as a risk function that converts the amount of an intervention carried out to the risk reductions resulting from those activities. As such, to work out the risk-reduction wedge shapes for each intervention, you can either: (i) work forwards from a proposed amount of an intervention, using the risk function to calculate the resulting level of risk reduction over time to plot in the scenario, or (ii) work backwards from a risk-reduction wedge shape required in a scenario, using the inverted risk function to calculate the amount of the intervention required to give those risk reductions. This can be done through the use of mathematical formulas or, at a higher-level, as a qualitatively estimated change in shape of the relevant wedge.

For activities in the portfolio (including the baseline activities, mitigation interventions, and solar radiation modification interventions) where risk and risk reductions are mediated through the physical climate system via the emission, reduction, and removal, of various climate pollutants, the coupling includes components relating to both (i) how the specific risk is coupled to the state of the climate (the ‘risk response’), and (ii) how the state of the climate is, in turn, coupled with emissions and other anthropogenic forcings (the ‘physical climate response’).

While the relationship between cumulative emissions and global temperature has been roughly linear to date (MacDougall, 2016; MacDougall and Friedlingstein, 2015; Matthews et al., 2009), explicitly including the ‘physical climate response’ enables scenarios to be built that incorporate concerns that non-linearities that may exist in the relationship between forcing and the state of the climate (Huang et al., 2022; Nicholls et al., 2020). For many risks types, the ‘risk response’ that connects the state of the climate to impacts can be challenging to identify and quantify to a high level of precision and confidence due to, for example, non-linearities and regional variations (IPCC, 2022b).

The risk reductions resulting from adaptation interventions are more direct (but not necessarily linear), as the nature of these activities and their impacts tend to be more local, specific, and not mediated via the global climate system (IPCC, 2022b; Shepherd and Sobel, 2020).

Three stylised examples of coupling are included in the typology in Table 1. In the first (‘incremental’), risk is roughly monotonically coupled to emissions; in this case risk increases as a function of anthropogenic forcing but stops increasing approximately when anthropogenic forcing stops. In the second (‘runaway’), risk may initially increase ‘incrementally’ but past a certain point it continues to increase even after anthropogenic forcing ceases. An example of this would be if the chosen risk type featured tipping behaviours (like AMOC) (Armstrong McKay et al., 2022; Lenton et al., 2023; Rahmstorf, 2024; Wunderling et al., 2024), where runaway changes in the system can continue long after forcing stops once a threshold is crossed, often irreversibly; a softer example is risks with a long response time (like sea-level rises), where the system takes a long time to equilibrate to a given level of anthropogenic forcing, even in the absence of self-reinforcing tipping dynamics (Ditlevsen and Ditlevsen, 2023). In the final response ('saturate’), risk increases as a function of emissions but, after a certain point, ceases to increase, or may even decrease, as the risk may saturate beyond a certain point. An example of this would be risk from flooding in a coastal location resulting from sea-level rise. Initially the risk of intermittent flooding may increase in line with sea-level rise. However, if the sea level has risen past a certain point, the settlement may become permanently flooded, essentially saturating the risk.

Baseline

The risk reductions resulting from the various interventions must be measured against a baseline, so each climate risk management scenario requires the selection of a baseline trajectory. This is depicted in the figure as the top line, i.e. the envelope, of the risk profiles. In the absence of mitigation or adaptation, the baseline shape corresponds to the risk of loss and damage and is therefore dependent on a range of speculative factors, including, for example, time lags between cause and effect, the physical climate response, population changes, and technological shifts over time, amongst others (Newman and Noy, 2023). An example of a baseline would be a business-as-usual scenario.

There are four qualitatively distinct baselines for CO₂-forcing-equivalent emissions (Jenkins et al., 2018; Wigley, 1998) included in Table 1: (i) accelerating; (ii) linearly increasing; (iii) decelerating; and (iv) overshoot. The shape of the risk envelope for a scenario is obtained by first choosing a specific baseline and then applying the risk function discussed in the section above. As discussed above, the risk function is used to convert the quantity of a high-level factor (which might be expressed in, e.g. carbon dioxide emissions) into a comparable risk quantity.

Mitigation interventions

Decreasing climate risk relative to the baseline trajectory can be achieved through the deployment of the ‘intervention’ high-level factors: that is, mitigation, solar radiation modification and adaptation. The ‘mitigation’ high-level factor represents the pathway that anthropogenic forcing takes over time. Mitigation wedges in the risk figure therefore correspond to interventions consisting of both emissions reductions and greenhouse gas removals, separated with a line through the wedge (reductions above and removals below). Greenhouse gas removal is chosen as the more general term over carbon dioxide removal to incorporate all contributors to global warming beyond just carbon dioxide. The top line of the mitigation wedges relates to a baseline risk trajectory, and the bottom line (in the absence of solar radiation modification deployment) corresponds to physical climate risk.

Table 1 includes three archetypal emission mitigation trajectories. In the ‘increasing indefinitely’ instance, mitigation interventions fall short of halting emissions, which results in cumulative global emissions following an indefinite upwards trajectory. In ‘net zero’, net global emissions decrease to, and then remain at, net zero. ‘Overshoot (net negative)’ sees global emissions decrease to net zero and continue into net negative territory.

It is important to note that the diagrams are expressed in terms of risk, not temperature or quantity of emissions reductions and removals over time. Therefore, the wedge width for mitigation is proportional to the risk reductions at a given time resulting from historical activities relative to the baseline, which is a function of the emissions difference between the baseline emissions trajectory and the absolute level of emissions after mitigation in a given scenario. The specific wedge shape for mitigation interventions in a scenario is obtained using the risk function on the difference between the mitigation trajectory and baseline. For example, given a baseline of constantly increasing emissions, a mitigation state of net zero would represent a constantly increasing emissions difference. This would lead to an increasing temperature difference, and therefore an increasing risk difference. In particular, if the ‘incremental’ instance is chosen for the coupling between climate risk and anthropogenic forcing, a state of net zero would have a constant (horizontal) base of the mitigation wedge, and an ever-increasing top line for the baseline.

The shape of this emission trajectory wedge aligns with the research of all three working groups of the IPCC, with WGIII providing the assessment of mitigation levels in terms of emissions, and WGI and WGII calculating the temperatures from emission levels and indicating risk outcomes from temperature for the particular risk being considered (IPCC, 2022a).

Potential solar radiation modification deployment

Like the ‘napkin diagram’, the typology in Table 1 incorporates potential risk reduction from solar radiation modification interventions. This high-level factor captures the temporal length of the solar radiation modification deployment, of which there are three instances: (i) ‘none’; (ii) ‘temporary’; and (iii) ‘sustained’. In the temporary instance, a short-term period of solar radiation modification deployment is used to temporarily reduce temperatures in a given climate scenario. For example, it could be used to limit peak risk in an overshoot scenario, and then once warming-induced risk decreases to the desired level, solar radiation modification deployment would be ceased. In the sustained instance, continued deployment is used to keep temperatures, and therefore temperature-related risks, below certain thresholds on an ongoing basis. Where there is deployment, the level of physical climate risk lies at the bottom of the solar radiation modification wedge, as solar radiation modification activities affect the physical state of the climate.

Since the climate risk management scenarios are cast in terms of risk and not temperature, the shape of the solar radiation modification wedge in the case of a long-term deployment is more complex than the case of long-term sustained carbon dioxide removals. Whereas the risk reductions of one ‘unit’ of carbon reductions and removals are ideally long lasting and therefore cumulative, the risk reductions of one ‘unit’ of solar radiation modification are short-term and therefore dependent on the current and recent levels of solar radiation modification deployment. Therefore, termination of solar radiation modification deployment would cause the solar radiation modification wedge to quickly diminish, leading to the physical climate risk level rapidly rising to the base of the mitigation wedge, potentially driving responses more severe than if the physical climate risk level had more gradually risen to that level – this sudden change in the physical climate system resulting from termination of solar radiation modification is sometimes called ‘termination shock’. Therefore, while the risk reductions from solar radiation modification's physical effects may keep increasing as the annual amount of solar radiation modification increases, the overall risk reductions may be attenuated owing the non-zero risk of termination shock; even without the probability of termination changing over time (which depends on governance), the magnitude of the impacts from termination would increase, and so the overall risk from termination shock could correspondingly increase. This tension means that the overall shape of the solar radiation modification risk reduction is non-trivial, contains much uncertainty, and should be considered with great care. Indeed, the requirement to consider the full gamut of risks associated with a particular course of action, rather than just an estimate of the likely temperature response, is a core benefit of using a risk-framed typology such as this. As with the baseline activities and mitigation interventions, obtaining the wedge shape of risk reduction for potential deployment of solar radiation modification must also incorporate use of the risk function to couple the deployment of solar radiation modification to the specific risk under consideration.

Adaptation interventions

Regardless of the speed and extent of mitigation interventions, the current level of carbon dioxide in the atmosphere means that some amount of warming is already occurring. Some form of adaptation interventions will thus be required. The risk reductions from adaptation interventions incorporate changes to learning over time as a result of improvements in both skill and capacity. The top line of the adaptation wedge corresponds to the level of physical climate risk, while the bottom line refers to the societal climate risk, or loss and damage. The size of the wedge refers to the actual risk reductions that occur as a result of the activities in question; it does not refer to the amount of adaptation required.

In Table 1, we present three stylistic instances of adaptation as a function of physical climate risk over time: (i) increasing, (ii) increasing and then decreasing, and (iii) constant. In the first instance (‘increasing’), the community in question learns over time and is able to increase its adaptation levels, even in a world where physical climate risk remains constant or increases. In the second instance (‘increasing then decreasing’), the community is initially able to adapt to low levels of climate risk, but over time is able to adapt less and less well, perhaps because exposure increases (e.g. due to a growing population), or owing to the existence of limits to adaptation as physical climate risk increases (IPCC, 2022b). In the final instance (‘constant’), these factors balance out, and the risk reductions from adaptation as a function of physical climate risk remain constant over time; note this constancy need not correspond to a constant amount of adaptation activities.

Six example portfolios of climate interventions

To illustrate the value of the typology in exploring climate risk management scenarios, we present in Figure 3 a case study of six scenarios that narratively highlight the potential utility of various interventions. These are presented at a global scale but can be customised to other spatial levels as necessary. The temporal scale here is considered to be over centuries, recognising that risks in the climate system have long timelines associated with them (IPCC, 2021).

All six of the example climate risk management scenarios here share the same risk type (here generalised climate risk), and risk thresholds (generalised thresholds: ‘incremental’, ‘systemic’, ‘existential’), and the same risk function that couples a given intervention to its resulting risk reductions. The horizontal risk levels on the figures incorporate the concern that risk thresholds exist that once crossed may have qualitatively different impacts; since it may be of strategic interest to limit the physical climate risk below such thresholds, these levels are stylistically used as guiderails for the intended goal for different interventions. In addition, all six example scenarios here share the same baseline (‘linearly increasing’), in order to focus discussions on different portfolios of interventions instead of speculative differences in baseline scenarios (Rogelj et al., 2023). Differences in these example scenarios therefore lie in the profiles of mitigation, solar radiation modification and adaptation interventions chosen, and the resulting levels of physical climate risk (after mitigation and solar radiation modification), and loss and damage risk (after adaptation).

Depending on the question one is interested in exploring, one could equally well choose different high-level factors to keep the same and/or different between each scenario. For example, one could construct scenarios that keep the mitigation and adaptation instances of the same for all scenarios (following Plan A below, for example), while changing the risk function that couples the intervention to risk type; this might be useful to explore how robust the outcomes of a particular set of climate interventions is to the range of possible climate responses.

Plan A is a stylised manifestation of a ‘no overshoot’ scenario. It incorporates early and rapid emissions reductions with levels of greenhouse gas removal limited to the necessary quantity to reach and sustain a stabilisation of risk over time. The simplest way that this might be achieved is if net zero emissions are reached and sustained, without net negative emissions after the point of net zero, and in an incremental physical climate response where risk increases with temperature and halts when it stabilises, and temperature is linearly proportional to cumulative emissions. In this stylised case, physical climate risk is stabilised at an ‘incremental’ level. The requirements of adaptation in this portfolio are therefore relatively low in comparison to other scenarios (but given the currently committed levels of warming and consequent climate impacts, any stabilisation at or beyond this state would crucially still entail high costs, whether monetary or otherwise). Thus, we chose the ‘increasing’ instance of adaptation, which results in a decrease over time in the risk of loss and damage.

In Plan B, slower initial ambition leads to temporarily overshooting the ‘incremental’ physical climate risk level, but a greater reliance on greenhouse gas removal enables net negative emissions and a consequent decrease in physical climate risk over time. In this example, the amount of risk reduction from adaptation activities is chosen from the typology to remain constant after the initial rise of physical climate risk (i.e. the wedge thickness is constant over time after the initial ramp up). This might be due to the more complex and less well-understood changes in the physical climate emerging from the both the unprecedented decreasing atmospheric carbon (from net negative emissions) and the fact that the physical climate risk state is constantly changing throughout the scenario. These factors might prevent learning that could increase the benefits from adaptation over time as the environment keeps changing. Despite the increase in the probability of crossing risk thresholds, loss and damage remains comparatively low due to the peak and decline in physical climate risk resulting from the mitigation efforts. Nevertheless, the risk of significant loss and damage still peaks above the incremental level, and in absolute terms would remain substantial.

In Plan C, mitigation goes far enough to both stabilise physical climate risk, and to begin decreasing it in an overshoot scenario; however, mitigation is not fast enough to prevent the peak physical climate risk level from first reaching the systemic risk level. Therefore, solar radiation modification is temporarily deployed to limit peak physical risk to a level that is more likely to be addressable through adaptation. Despite this, and unlike with the stabilisation portfolio in Plan A, we choose the ‘constant’ instance for adaptation. While the physical climate risk might have been stabilised due to the global temperatures being stabilised by solar radiation modification, the climate system will not be in a static equilibrium state as the varying levels of solar radiation modification changes (potentially unpredictably) the atmospheric composition and behaviour throughout deployment, and these changes may prevent adaptative capabilities from increasing over time. Further factors in the large-scale deployment of solar radiation modification could lead to further unintended side effects such as trans-border conflict over solar radiation modification deployment, which may also non-trivially affect the risk of loss and damage.

In Plan D, the ‘Danger’ scenario, the trade-offs between solar radiation modification and mitigation-based on emissions reductions and removals are illustrated. Here, the risk reduction from mitigation relative to the baseline initially increases but after a certain period the reductions slow down, with emissions never reaching net zero and therefore never stabilising physical climate risk. Therefore, global-scale long-term solar radiation modification deployment must be constant and increasing to constrain physical climate risk and prevent it from indefinitely increasing.

In Plan E, the ‘Emergency’ scenario, while mitigation efforts are pursued, the overall portfolio falls short of the actions required to constrain warming and therefore physical climate risk, leading to significant and permanent exceedance of risk boundaries. Adaptation efforts are also pursued, but similarly fall short of the scale and effectiveness to constrain loss and damage, which therefore increases indefinitely.

In Plan F, the ‘Fatal’ scenario, the limits to adaptation are highlighted (IPCC, 2022b). No additional mitigation efforts are pursued beyond what might have occurred in the absence of climate action (e.g. changing global emissions due to disruptions from escalating impacts of climate change). While adaptation is attempted beyond what might have occurred passively, we choose the ‘increasing then decreasing’ adaptation instance as the ever-increasing physical climate risk eventually surpasses humanity's ability to respond.

Customisation and further use of the typology

The typology presented in this article can be customised to represent other risks at varying spatial and temporal scales. An example is a coastal city government's consideration of the risks associated with sea-level rise. The city might use the typology to consider the issue at a high level: business-as-usual emissions (the selected baseline) will result in sea-level rise, which might have impacts on infrastructure and livelihoods. Global and local emission reductions, as well as local adaptation measures, can decrease the risks that sea-level rise poses to the city. By visualising the risk-reduction interventions together, the city would be able to identify whether local mitigation and adaptation interventions are working in tandem to decrease risk or are inadvertently in tension. The planned relocation of infrastructure as part of adaptation efforts may, for example, have an inadvertent impact of increasing travel distance to the new site, increasing local emissions.

The exercise might go a step further by including quantitative elements, like using emission projections to estimate the level of sea-level rise and thus to calculate the cost of adaptation measures. These estimations could be combined to arrive at an estimate of the cost of the unavoidable loss and damage associated with sea-level rise in the city. From this estimation, the city could, for example, better understand how the costs might be distributed.

The typology could also be used in conjunction with other risk tools and frameworks. For instance, as mentioned above, the burning embers tool (used by, e.g. the IPCC) can be used as the chosen risk. The tool could also, for example, be used in exploratory approaches commonly used by the decision making under deep uncertainty community (Workman et al., 2024), or other exploratory approaches like the adaptation pathways approaches (Werners et al., 2021), by providing structure around which to generate a variety of scenarios.

Conclusion

The current state of climate action indicates that overshooting the 1.5 °C temperature target is becoming increasingly likely. As such, a wider variety of interventions may need to be deployed, likely at increasing scale and under intensifying time pressure. Careful risk management will thus be required to avoid the inadvertent creation of additional risk from unintended side effects, yet risks accruing from the interactions between interventions can be difficult to gauge since approaches are often siloed.

In response, the climate risk management typology and example portfolios presented here provide a tool to holistically integrate climate interventions via their impacts on risk. Using this typology to create the portfolios brings together a variety of disciplines, and all IPCC working groups, in one figure; accordingly, the typology can contribute towards a common framework that supports better communication between practitioners and researchers. In particular, the six portfolios can act as a shorthand for the role of various climate interventions and the need to consider the overall risk reductions, and potential amplifications, resulting from each intervention. Further, the typology allows for arbitrary scenarios to be created for management of a variety of risks, across a range of timescales, localities, and sets of values. Importantly, it underlines the multifaceted approach to decreasing risk, which requires an integrative approach from decision-makers. This typology creates the structure for decision-makers to do so.

It is important to emphasise that the interventions included in the typology and portfolios are selected in order to prioritise the understanding of potential linkages between them, rather than as an equally weighted endorsement of each, recognising the contested nature of some of these interventions in a variety of contexts. Indeed, one goal of integrating all options is to help in recognising trade-offs, more effectively use constrained resources, maximise synergies, and consider how the full gamut of implications and uncertainties relating to each intervention ultimately affects risk.