2.1 Model description

This paper illustrates the SCoMaE method for the example of model simulations performed with version 2.9 of the University of Victoria Earth System Climate Model (UVic ESCM), an Earth system model of intermediate complexity (Eby et al., 2013). It includes schemes for ocean physics based on the Modular Ocean Model Version 2 (MOM2) (Pacanowski, 1995), ocean biogeochemistry (Keller et al., 2012), and a terrestrial component including soil and vegetation dynamics (Meissner et al., 2003). It is coupled to a thermodynamic sea ice model (Bitz et al., 2001) with elastic visco-plastic rheology (Hunke and Dukowicz, 1997). The atmosphere is represented by a two-dimensional atmospheric energy moisture balance model (Fanning and Weaver, 1996). All model components have a common horizontal resolution of 3.6^∘ longitude and 1.8^∘ latitude and the oceanic component has a vertical resolution of 19 levels, with vertical thickness varying from 50 m near the surface to 500 m in the deep ocean. Wind velocities used to calculate the advection of atmospheric heat and moisture as well as the air–sea-ice fluxes of surface momentum and heat and water fluxes, are prescribed as monthly climatological wind fields from NCAR/NCEP reanalysis data (Eby et al., 2013). Wind anomalies, which are determined from surface pressure anomalies with respect to preindustrial surface air temperature, are added to the prescribed wind fields.

A list of the globally aggregated output variables is given in Table A1.

2.2 Spin-up and scenario forcing

For the default model simulation, the UVic ESCM was spun-up with preindustrial (year 1765) seasonal forcing for over 10 000 years. All simulations were integrated from 850 until 2005 using historical fossil-fuel emissions and land-use changes, as well as radiative forcing from solar variability and volcanic activity following Eby et al. (2013). Following Keller et al. (2014), continental ice sheets were held constant to facilitate the experimental setting and analyses. Warming from black carbon, indirect ozone effects, and cooling from indirect sulfate aerosol effects were not included. From 2005 onward until 2100 the Representative Concentration Pathway (RCP) 4.5 and 8.5 scenarios from Meinshausen et al. (2011) were implemented as an intermediate and high-CO₂ emissions driven scenario, respectively.

For the sensitivity analysis performed with the UVic ESCM, different model input parameters and parameterizations were perturbed, and for some of them it was necessary to do a new model spin-up to reach steady-state conditions again; apart from this the forcing was the same for all simulations.

Figure 1Illustration of the correlation matrix construction for the example case study, and the model output variables surface air temperature (SAT) and Northern Hemisphere (NH) sea ice area. In the first step, temporal differences of the simulations are calculated between 2005–2015 and 2090–2100. Second, changes in the variables induced by the parameter perturbations are correlated. Last, this correlation information is used as one of many entries in the correlation matrix.

Download

Table 1List of perturbed model input parameters.

Download Print Version | Download XLSX

2.3 Parameter perturbations

In the following sections, the single-parameter perturbation experiments, which are used in the example and shown in Fig. 1, are explained in detail. We chose these parameters to explore the sensitivity of the UVic ESCM to uncertainties in terrestrial and marine biological productivity with respect to temperature and CO₂, since these processes will influence the future carbon cycle. In addition, we perturbed ocean vertical diffusivity, since this is a physical process influencing marine carbon uptake. All parameters were perturbed within physically meaningful ranges, which were evaluated based on their agreement with the time series of the historical global mean air temperature (Fig. S5). See Table 1 for a quick overview of the simulations.

3.1 Step 1: calculate the correlation matrix

Throughout this study, a variable is defined as a model output or observational time series, whereas we refer to it as an indicator if a variable was selected to represent a certain aspect of the considered system. To obtain a comprehensive, nonredundant set of indicators to describe a given system, the first step is to construct a correlation matrix, i.e., a matrix including the correlation information of all the relevant Earth system variables to each other. The construction of the correlation matrix strongly depends on the research question and needs to be adjusted accordingly. The selection of which variables are the relevant variables for the given research question and hence should be included in the matrix, as well as the choice of how the correlations should be calculated is very important for the outcome of the study. In the same way, it is important to consider a reasonable signal-to-noise ratio within the data set chosen. Correlations could for example be calculated between time series of variables or their derivatives, absolute temporal changes, or spatial patterns. Alternatively, output from ensemble simulations could be used to calculate correlations between changes in variables due to the different ensemble members. The matrix is then evaluated based on the significance information of these correlations (see Step 2). Note that for this preselection of the possibly relevant variables to answer the given question, as well as for the construction of the correlation information in the matrix, a certain level of expert judgement is needed.

To illustrate the construction of the matrix based on our example simulations, we show how the correlation between changes in global mean “surface air temperature” (A_sat) and “Northern Hemisphere sea ice area” (O_iceareaN) in the Representative Concentration Pathway (RCP) 8.5 emission scenario (Meinshausen et al., 2011) due to the parameter perturbations translates to the corresponding correlation matrix entry (Fig. 1). In our example we want to study the correlations between changes in model output variables, induced by varying poorly constrained model input parameters concerning the carbon cycle. In the following we will refer to these as “correlation of variable changes”.

Assuming that the signal of interest is of a similar kind as the state differences between the start and the end of a climate change simulation, we start by calculating the temporal differences between 2005–2015 and 2090–2100 from a number of parameter perturbation simulations that serve as our ensemble in this example (see Sect. 2.3 for explanations of the parameter perturbations). This enables us to learn whether the different output variables show a similar behavior for the respective parameter perturbation. Then the Pearson correlation coefficients between these changes are calculated and tested by performing a two-sided test at a 5 % significance level, with N = 16, the number of perturbed parameter simulations, and accordingly t_crit = 2.145.

In our example, there is a negative correlation of variable changes evident between “surface air temperature” (A_sat) and “Northern Hemisphere sea ice area” (O_iceareaN). This illustrates that these model output variables show consistent opposite reactions towards the parameter perturbations, i.e., if the perturbation causes surface air temperatures to increase, it also causes northern hemispheric sea ice to decrease. This information is then written into the correlation matrix. By studying the constructed correlation matrix and studying single correlations of changes between model output variables, we can learn about basic processes within the simulated climate system and test whether these agree with our expectations. To simplify the visual analysis of our example we sorted the variables in the matrices according to their strength in correlation of variable changes relative to changes in the commonly used climate change indicator, i.e., “surface air temperature” (A_sat) in the historical scenario.

Figure 2Illustration of the indicator selection process using the example of the correlation matrix for the historical scenario (see Fig. 3 for a more detailed display of the correlation matrix). The correlation matrix, was constructed as explained in Fig. 1 but for the temporal differences between 1850–1860 and 1995–2005. See Sect. 3.2 for a detailed step-by-step description of the evaluation process. Prefixes A, O, L, and F stand for atmosphere, ocean, land, and fluxes, respectively; for a detailed description of the model output variables, see Table A1.

Download

3.2 Step 2: cluster identification and indicator selection

To obtain a set of indicators for the assessment of changes in the system under consideration, we systematically evaluate the previously constructed correlation matrix (see Fig. 2 for an illustration of this procedure). To obtain a comprehensive, nonredundant indicator set, we follow these steps: (1) the first indicator is the variable with the highest number of significant correlations with other variables; (2) all variables with a significant correlation are clustered under this indicator; (3) these clustered variables are then excluded from the selection of the next indicator; (4) the next indicator is again the variable with the highest number of significant correlations with all the remaining variables; (5) this indicator selection procedure is repeated until all variables are clustered and are represented by an indicator. If a variable is not significantly correlated to any of the remaining variables, this variable is considered to be a single indicator. These single indicators are needed for a fully comprehensive assessment, since they show different behavior from all previously selected indicators and hence provide additional information.

In our example, we applied the SCoMaE method to the correlation matrix concerning 46 commonly used variables for the assessment of climatic changes in the historical forcing scenario, simulated by the UVic ESCM (see Sect. 2.1 for details on the simulations). We find that the first indicator for our research question in the historical period is “precipitation over ocean areas” (F_precipO) (Figs. 3 and S5). By following the respective column of F_precipO (17th from the right) in the correlation matrix, we can see that changes in this model output variable are significantly correlated to changes in all variables that are also significantly correlated to changes in “surface air temperature” (A_sat; first from the bottom), with the exception of “mean ocean temperature” (O_temp, 16th from the bottom) but in addition also link changes in global and terrestrial precipitation and evapotranspiration (F_precip, F_precipL and F_evap, F_evapL, respectively; 35th and 37th from the bottom) as well as changes in “surface net upward longwave radiation” (F_uplwr, 40th from the bottom). The changes in these variables due to parameter perturbations are not significantly correlated to changes in “surface air temperature” (A_sat). Hence, based on purely statistical considerations, using “precipitation over ocean” (F_precipO) as an indicator for the research question in the historical period would be preferable to global mean “surface air temperature” (A_sat), the main ad hoc indicator of historical climate change, since it potentially holds more information.

Figure 3(a) Correlation matrix for the historical scenario. The correlations are calculated between changes in the 46 model output variables for temporal differences between 1850–1860 and 1995–2005 from the results of the perturbed parameter simulations (as in Fig. 1). Correlations significant at a 5 % significance level are marked with crosses. The order of the variables was determined based on their correlation strength to “surface air temperature” (A_sat) in the historical scenario. Prefixes A, O, L, and F stand for atmosphere, ocean, land, and fluxes, respectively; for a detailed description of the model output variables, see Table A1. (b) Indicators as identified from the SCoMaE analysis of the correlation matrix above as illustrated in Fig. 2, ranked by the amount of significant correlations. The indicators are as follows: “precipitation over ocean” (F_precipO), “land surface albedo” (A_albsurL), “ocean surface alkalinity” (O_alksur), “mean ocean temperature” (O_temp), “ocean phytoplankton” (O_phyt), and “ocean overturning” (O_motmax).

Download

“Surface albedo on land” (A_albsurL) is identified as the second indicator. After excluding all variables correlated to changes in “precipitation over ocean” (F_precipO), its changes due to the parameter perturbations are significantly correlated to changes in “net surface downward shortwave radiation” (F_dnswr), “ocean oxygen” (O_o2), and “sea surface salinity” (O_salsur). The third indicator is “ocean surface alkalinity” (O_alksur), which shows the same response to the parameter perturbations as “ocean surface phosphate concentrations” (O_po4sur). When excluding all variables that are clustered under one of the three abovementioned indicators, three variables remain unclustered: “mean ocean temperature” (O_temp), “maximum meridional overturning” (O_motmax), and “ocean phytoplankton” (O_phyt). These variables are hence single indicators, which are needed for a comprehensive assessment of the system under consideration (Fig. 3b).

See Sect. 1 and Figs. S1 and S2 in the Supplement for the results of these analyses for the intermediate–high (RCP4.5) and the business-as-usual (RCP8.5) scenarios, respectively.

Figure 4Indicators identified from the analysis of the RCP4.5 (blue) and RCP8.5 (red) correlation matrices with the precondition to use the historical indicators first. The indicators are as follows: “precipitation over ocean” (F_precipO), “land surface albedo” (A_albsurL), “ocean surface alkalinity” (O_alksur), “mean ocean temperature” (O_temp), “ocean phytoplankton” (O_phyt), “ocean overturning” (O_motmax), “net radiation at the top of the atmosphere” (F_netrad), “ocean surface dissolved inorganic carbon” (O_dicsur), and “downward shortwave radiation” (F_dnswr).

Download

3.3 Step 3 (optional): comparison of indicators for the different forcing scenarios

In order to learn how well the previously identified indicators for one scenario explain a different scenario with changed forcing, we prescribe the use of the previously identified indicator set. The SCoMaE accordingly first uses these indicators and then analyses whether and which additional indicators are needed for a fully comprehensive assessment of the new scenario.

For the example, we prescribed the indicators identified for the historical scenario to assess the intermediate–high (RCP4.5) and the business-as-usual (RCP8.5) emission scenarios (Fig. 4). The results show that if we were to only utilize the indicators from the historical scenario for the assessment of the two RCP scenarios, we would not be able to assess all changes in the climate system as represented by our model: for the RCP4.5 scenario, we would obtain additional information by considering the variables “net top-of-atmosphere radiation” (F_netrad) and “ocean surface heat flux” (F_heat), which are clustered together, and “net upward longwave radiation” (F_uplwr) and “ocean surface salinity” (O_salsur), which form another indicator cluster (Fig. 4).

Note that Earth system variables clustered under the prescribed indicators differ among the different scenarios (compare Figs. 3 and S1): in the historical scenario the indicator “precipitation over ocean” (F_precipO) includes the output variables “net top-of-atmosphere radiation” (F_netrad), “ocean surface heat flux” (F_heat), “ocean surface nitrate” (O_no3sur), “top-of-atmosphere outgoing longwave radiation” (F_outlwr), and “net upward longwave radiation” (F_uplwr), all of which are not included in the “precipitation over ocean” (F_precipO) indicator of the RCP4.5 scenario. Instead, the indicator “precipitation over ocean” (F_precipO) for the RCP4.5 scenario includes “mean ocean temperature” (O_temp), which it is not included for the historical scenario.

The differences between the correlation matrices for the RCP8.5 scenario compared to the historical scenario are even larger (compare Figs. 3 and S2). For the RCP8.5 scenario, 8 out of 46 considered variables would not be included if we applied the indicators identified for the historical scenario. Instead we need three additional indicators for the assessment of the system under consideration, namely “net top-of-atmosphere radiation” (F_netrad), “ocean surface dissolved inorganic carbon” (O_dicsur), and “net surface downward shortwave radiation” (F_dnswr) (Fig. 4). Note that six of the eight remaining variables that were initially included in the first indicator cluster for the historical scenario, namely “precipitation over ocean” (F_precipO), are no longer significantly correlated to it for the RCP8.5 scenario.

These differences in the correlation matrices for the different forcing scenarios indicate changes in prevailing correlations between Earth system variables with the imposed climate forcing. This illustrates that a reevaluation of the indicators chosen may be needed for a comprehensive assessment of different climate strategies yielding different climate states.

Figure 5(a) Correlation matrix for all three scenarios, merging the significance information from all three scenarios. Colors indicate in which scenario the changes in variables due to parameter perturbations showed a significant correlation; see color bar for an explanation. The crosses mark combinations of variables where the correlation of variable changes is significant at a 5 % significance level in all three scenarios. For details on the model output variables under consideration, see Table A1. (b) Indicators as identified from the analysis based on the correlation matrix above against the number of significant correlations (blue) and with the condition that “surface air temperature” (A_sat) is prescribed as the first indicator (red). The indicators are as follows: “atmospheric carbon content” (A_co2), “precipitation over land” (F_precipL), “atmosphere-to-ocean carbon flux” (F_carba2o), “net top-of-atmosphere radiation” (F_netrad), “net surface downward shortwave radiation” (F_dnswr), “atmosphere-to-land carbon flux” (F_carba2l), “ocean surface nitrate” (O_no3sur), “top-of-atmosphere outgoing longwave radiation” (F_outlwr), “ocean oxygen” (O_o2), “ocean surface alkalinity” (O_alksur), “ocean phytoplankton” (O_phyt), “ocean surface salinity” (O_salsur), “sea surface phosphate” (O_po4sur), “ocean overturning” (O_motmax), “precipitation over ocean” (F_precipO), “ocean carbon” (O_totcarb), and “surface net upward longwave radiation” (F_uplwr).

Download

3.4 Step 4 (optional): evaluation of a common correlation matrix

To advance this analysis such that changes in correlation matrices from different forcing scenarios can be taken into account, it is possible to create a correlation matrix representing only those correlations that are significant in all forcing scenarios; this is defined as a common correlation matrix. Applying the SCoMaE method to such a common correlation matrix identifies an indicator set that can be used to assess and also compare multiple scenarios and which hence differs from the previously identified sets for the individual correlation matrices.

To obtain a common indicator set for the three example forcing scenarios (historical, RCP4.5, and RCP8.5), we construct a correlation matrix in which only correlations of variable changes that are significant in all these scenarios are considered (Fig. 5). Furthermore the color shading indicates in which of the scenarios the correlations between variable changes were found to be significant.

A first visual evaluation of the common correlation matrix shows more reddish than bluish shading, which indicates that the correlation patterns for the historical and RCP4.5 scenarios are more similar than for the historical and RCP8.5 scenarios (Fig. 5). This means that for a lower future emission scenario, the indicators from the historical scenario are more suitable than for a higher future emission scenario. This is true with the exception of the terrestrial and oceanic carbon fluxes (F_carba2l and F_carba2o, respectively). These two fluxes are perturbed by the land-use scheme implemented in the RCP4.5 scenario, since this scenario includes a high amount of afforestation and reforestation. Furthermore, greenish shading shows correlations of variable changes that are significant only in the RCP scenarios, indicating that those correlations of variable changes depend on the increasing anthropogenic (mainly CO₂) forcing, included only in these scenarios.

The first indicator obtained from the common SCoMaE analysis is “atmospheric CO₂” (A_co2), which was also found to be the first indicator in the RCP8.5 scenario (Figs. S7 and S8). Its changes are significantly correlated to changes in 27 other output variables in all three scenarios. This indicates that these correlations of variable changes are robust throughout the different strength of CO₂ forcing in the three scenarios. The fact that “atmospheric CO₂” (A_co2) is the first indicator with a large number of correlated variables suggests its ability to reflect the changes in other Earth system variables with regard to the parameter perturbations, such as changes in temperatures, carbon fluxes, and moisture fluxes over the ocean. This can possibly be explained by the fact that the changes in these variables are sensitive to the imposed CO₂ forcing, which in turn is reflected in the atmospheric carbon concentration.

The second indicator is “precipitation over land” (F_precipL), which is clustered with “terrestrial evapotranspiration” (F_evapL) and “net upward longwave radiation” (F_uplwr) (Fig. S6). This cluster accordingly represents changes in terrestrial moisture fluxes and the resulting surface upward fluxes of longwave radiation. The latter relates to the surface air temperature, which on land is strongly influenced by the amount of evapotranspiration, and the resulting evaporative cooling. Note that the fact that terrestrial moisture fluxes are clustered under a separate indicator suggests a different sensitivity of these variables to the perturbed parameters. Since these three variables show significant correlations of variable changes to each other in all three scenarios, one could use any of them as the indicator for this cluster. The same is true of the next indicators and their clusters, which are “air-to-sea carbon flux” (F_carba2o) and “soil respiration” (L_soilresp); “net top-of-atmosphere radiation” (F_netrad) and the “ocean surface heat flux” (F_heat); and “net surface downward shortwave radiation” (F_dnswr) and the “land surface albedo” (A_albsurL).

The remaining single indicators are “air-to-land carbon flux” (F_carba2l), “ocean surface nitrate” (O_no3sur), “top-of-atmosphere outgoing longwave radiation” (F_outlwr), “ocean oxygen” (O_o2), “ocean surface alkalinity” (O_alksur), “ocean phytoplankton” (O_phyt), “sea surface salinity” (O_salsur), “ocean surface phosphate” (O_po4sur), and “maximum ocean meridional overturning” (O_motmax).

3.5 Step 5 (optional): including expert judgment

If stakeholders or experts were to inform the indicator selection process, it would be possible to prescribe indicators and then use the SCoMaE analysis to identify additional uncorrelated variables that are needed to obtain a comprehensive assessment of the system. Also, instead of using global mean time series, one could look at time series of regions or already processed variables, such as heat stress or cumulative emissions. This approach in combination with the SCoMaE analysis enables us to learn about variables which have previously been disregarded but potentially provide new information about the system or to learn which of the indicators previously considered actually provide redundant information.

How would the common indicator set from our example change if we were to include the condition that surface air temperature should be the first indicator, instead of atmospheric CO₂?

Prescribing “surface air temperature” (A_sat) as the first indicator for the common correlation matrix leads to the replacement of “precipitation over land” (F_precipL) by “precipitation over ocean” (F_precipO) as the second indicator (Fig. 5b); its change with the parameter perturbations is correlated with 12 variables that are clustered under this indicator. Almost all of these variables were initially clustered under “atmospheric CO₂” (A_co2) but are not significantly correlated to changes in “surface air temperature” (A_sat). These variables mainly describe global and oceanic moisture fluxes, as well as carbon fluxes or reservoirs on land: “precipitation over the ocean” (F_precipO), “global evaporation” (F_evap), “global precipitation” (F_precip), “vegetation net primary productivity” (L_vegnpp), “leaf area index” (L_veglai), “vegetation carbon” (L_vegcarb), and the “surface upward sensible heat flux” (F_upsens). The only exception to this behavior is “total ocean carbon” (O_totcarb), which in turn becomes a single indicator. In addition the second indicator, “precipitation over the ocean” (F_precipO), now incorporates the previously identified clusters of the second and third indicators, namely the clusters of “precipitation over land” (F_precipL) and the “air-to-sea carbon flux” (F_carba2o). Only “net upward longwave radiation” (F_uplwr), which was also clustered under “precipitation over land” (F_precipL) becomes a single indicator, remaining unclustered when “surface air temperature” (A_sat) is prescribed as the primary indicator. In turn, “air-to-land carbon flux” (F_carba2l), which was a single indicator in the default SCoMaE analysis, is now clustered under “surface air temperature” (A_sat).

The third and fourth indicators are “net top-of-atmosphere radiation” (F_netrad) and “net surface downward shortwave radiation” (F_dnswr), which were found with the same underlying clusters in the default analysis (compare Figs. S8 and S9). Finally, eight of the nine previously identified single indicators remain unclustered and hence are still single indicators.

Although the total number of indicators has not changed, the identified clusters and their meaning differ: in the default analysis, the first indicator represented changes in temperatures, carbon fluxes, and global and oceanic moisture fluxes. If “surface air temperature” (A_sat) is prescribed, the global and oceanic moisture fluxes are moved to the second cluster, which in addition incorporates some Earth system variables from the previously identified second and third indicators. This is one example showing how the SCoMaE method allows for the inclusion of expert judgment or preconditions, is able to account for changes in correlation patterns, and allows one to determine which indicators are needed for a comprehensive and nonredundant assessment. (For more discussions, see Sect. 2 and Fig. S3 in the Supplement.)

4.1 Discussion of the results from the example

4.2 Discussion of the SCoMaE method

The construction of an individual or a common correlation matrix can be a useful tool for assessing the state of complex systems. Individual correlation matrices allow one to obtain an initial overview of relationships between the different system variables, whereas a common correlation matrix shows how changes in the state of a system, imposed by, e.g., varying forcing scenarios, influence these relationships. The SCoMaE method then allows us to cluster the variables, based on statistical considerations, to obtain a nonredundant indicator set to guide more detailed analysis.

However, in order for this to be useful one must carefully select, what information to include in the correlation matrix, which in turn strongly depends on the given research question. This can be illustrated by the implicit choices made for our example case study, where we regarded correlations of variable changes in globally averaged model output variables given various parameter perturbations. The first choice in this case study was to use global aggregates of the model output. However, if the research focus were set on, e.g., regional phenomena, the correlations for the matrix could also be constructed either between regional aggregates or based on the correlation strength for a given spatial pattern.

The second choice for the case study, was to regard correlations between changes in model output variables based on their reaction to a parameter perturbation under changing climate forcing. Instead of using model output, it is also possible to further process the data and calculate derivatives of the model output variables, such as heat stress or cumulative time series. On another note, using a model with higher internal variability, it would also be possible to regard temporal correlations of Earth system variables over a chosen time period. In contrast to the purely process-based parameter perturbations that we regarded in the case study, this would hold information about the timescales and temporal development of the model output variables, which, in turn, could indicate common underlying processes in the model. Additionally, if the considered time series showed higher internal variability, it might be conceivable to apply a specific temporal filter to the data before calculating the correlation matrix. This could allow the distinction between important processes on different timescales, from daily and seasonal to interannual or decadal.

In the following we want to discuss the contribution of the SCoMaE method to achieve the three characteristics for indicator selection as introduced by Radermacher (2005). Constructing a correlation matrix enables scientists to comprehensively identify correlations in complex systems, such as the Earth system, both simulated and observed. The application of SCoMaE allows one to identify scientifically consistent sets of indicators, which are independent and hence do not provide redundant information, to be used in a science-led assessment. This method represents a bottom–up, natural-science perspective on indicator selection. It thereby tackles one of the three characteristics discussed by Radermacher (2005), namely that of scientific consistency.

In our example the SCoMaE method is based on model data and hence does not account for information about the statistical measurability of the identified indicators. This makes it difficult to directly translate a model-based indicator set to a “real-world” application. This is the case, for example, for the first indicator in the historical scenario: “precipitation over ocean” (F_precipO). The lack of long-term historical precipitation measurements over the ocean (New et al., 2001) would prevent this indicator from being used in a real-world application. It is, however, noteworthy that there is value in the knowledge that this variable could hold information about other Earth system variables, and hence it might be worth improving the observational system.

The third characteristic mentioned by Radermacher (2005) is the political relevance of indicators. Since SCoMaE is a bottom–up, science-led approach for indicator selection, it at first does not take into account political, ethical, and economical considerations. However, these considerations as well as measurability constraints can be included in the analysis. Prescribing a certain indicator, e.g., “surface air temperature” (A_sat) in Sect. 3.5, SCoMaE allows us to include expert judgment and enables us to identify the remaining indicators needed for a fully comprehensive assessment. An iterative learning process for which indicators are societally relevant and scientifically consistent as proposed by Oschlies et al. (2017) and Singh et al. (2015) would hence allow the SCoMaE method to identify scientifically meaningful, measurable, and politically relevant indicators sets.