(1) By reading the introduction and the methods the general impression is that this work combines several existing data and methods (e.g., GCE model GRACE, ERA5 data) without spending much words on the actual contibution of the authors and how the applied methods in the selected context provide an innovative perspective on the multi-hazard theme. In other words, what the work reveals that could not be captured, or well interpreted, or understood, without the multi-hazard analysis? What is the added value? And how these insights help in adaptation? (adaptation is mentioned in the abstract without much results or discussion on that in the body of text).

Thank you for the comment. We will add a paragraph in the Discussion section.

Additionally, we agree with the review that the adaptation part does not have much body in the main text, so this concept will be removed at places where it does not have added value.

2) In the results the description of some of the figures/plots is missing. As if some of the figures does not contribute to the 'story' of the paper. I would suggest to clearly describe and discuss the provided images. Particularly Fig. 4 (not described and poorly readable) and Fig. 8.

Thank you for the comment, Fig. 4 will be updated to enhance its readability, and we will add a description. For Fig. 8 , we will also add a description.

In table 2 Bare rocks are listed as 'affected' by heatwaves, droughts, fires and their combination. Why? are bare rocks vulnerable? I think a bit of discussion and comment on this would be precious for the reader.

Thank you for the comment, we will add a discussion point as to the relevance of mentioning “bare rocks” affected by multi-hazards.

Figure 9. Caption. It seems that also indirect impacts are shown, not only direct.

We appreciate the referee's comment. That is true.  The caption of Figure 9 will be revised as “Direct and indirect impacts on the domestic economy by sectors due to 2018 multi-events in Scandinavia”.

There are some typos, e.g. reference to Fig.16 that does not exist (l. 273), please check the text carefully.

We will remove the typos and will go over the manuscript another time in detail to filter potential other typos.

We thank the reviewer for the comprehensive thoughts on improving the manuscript, specifically on the transparency and robustness of the analysis. We agree that the manuscript should be enriched with additional information on the models used and refining the research objective more. We will address the specific points listed below and we will integrate the comments and feedback into the revised manuscript.

Abstract: While the relevance of multi-hazard risk assessment is well-highlighted, the abstract would benefit from a more explicit statement of the study’s objective and the key takeaways from the results.

We appreciate the reviewer’s comment and we have identified one key sentence on the objective of the manuscript and another summarizing the key findings.

Introduction: The literature review is strong and provides good context for multi-hazard analysis. However, the study’s specific objective remains vague. Additionally, it would be useful to include references to previous research on hazard impacts in Scandinavia and discuss the contexts in which the methods used in this study were developed. How relevant is the macro-economic model to the specific case study?

We appreciate the reviewer’s comment and we will make the study’s specific objective more clear with better defined statements in the introduction. We will also include more previous research and provide more context of the study area.

Methods: The explanation of the methodology (lines 82–92) is fragmented, making it difficult to follow. Consider providing an overview of the rationale behind each methodological step before explaining the individual steps. Additionally, the flowchart could be improved by clearly separating data sources (e.g., ERA5) from methods (e.g., copula modeling).

We appreciate the reviewer’s comment and will add an overview of the rationale between each methodological step. We will look at how we can make the distinction in the flowchart clearer for the reader.

Lines 117–121: The definition of multi-hazard event probabilities is hard to follow. Could you provide a couple of equations to clarify this? Specifically, what does the condition "greater than or less than the 2018 value" mean, and what time periods are being considered?

We will expand this section to explain the method, through which we derive the optimal event definition, in more detail.

What metrics were used to select the best-fit copula for modelling the coupled marginal distributions of precipitation and temperature?

We will add a sentence stating “In this case, we found that the symmetrical Frank copula is the best fit for our data set, by minimizing the Akaike Information Criteria across several copula families (as implemented in the R package VineCopula; Nagler et al., 2024). The copula parameters are determined using the inverse of the Kendall rank correlation coefficient.”

It is unclear how the copula definition of the 2018 event connects to subsequent steps in the methodology. If I understand correctly, the later steps do not consider the probability of (temperature and precipitation) exceedance. Could this connection be clarified?

We have added a sentence stating that the event definition is used as a proportional weight for assessing the economic impact of the 2018 multi-hazard event.

More information on the econometric models is needed. How complete are the confounding variables? It appears that key economic and spatial variables may be missing. Providing details on regression model performance and stepwise variable selection in the Supplement/Appendix would enhance transparency.

We appreciate the reviewer’s comment. We will enclose the estimation results in the Appendix. The variables for selection include variables [dT, dP, dT^2, dP^2, TdT, PdP, dTdP]. Meanwhile, for all the estimations, we will include year effects and country-fixed effects when estimating the imposed response function for agriculture production. We will include both year and month effects for the estimation for the energy production.

What are the uncertainties in the macro-economic model? How well do the model’s predictions align with observed GDP changes in the affected regions?

We appreciate the reviewer’s comment. The CGE model used in this study, like all macroeconomic models, has uncertainties that can be summarized in the following aspects. 

First, the GRACE model relies on input-output tables, which represent fixed sectoral interdependencies. However, during hazards, these relationships can change dynamically due to shifts in production structures, substitution effects, and market adjustments. In this study, we focus on the yearly impact of the 2018 multi-hazard. Given that major structural adjustments within a single year are rare, the model's static representation of sectoral interdependencies is a reasonable assumption for short-term analysis.

Second, uncertainty arises from the GRACE model's parameters. The CGE model relies on key parameters, such as substitution elasticities between capital, labor, and natural resources. These parameters are based on values from the literature and previous studies, as a common practice in CGE modeling. The uncertainties of model outcomes to these elasticities exist. However we use baseline dataset for model calibration, to ensure that the model replicates key features of the regional economy as accurately as possible in the base year. 

Third, there may be uncertainty in the damage function estimation. The economic impacts of heat, drought, and fires are incorporated through productivity loss functions. These functions are necessarily simplified representations of complex biophysical and economic processes. While they provide useful estimates, they introduce uncertainty in quantifying the exact magnitude of sectoral impacts.

Regarding the comparison with observed economic changes, particularly GDP, we find it is challenging as real-world GDP outcomes are influenced by multiple overlapping factors beyond the specific multi-hazard event in this study. Observed changes in GDP, production, and prices reflect market equilibrium adjustments that include autonomous adaptation to multiple disturbances. Our study isolates the impact of the 2018 multi-hazard event, making direct GDP comparisons complex. 

While the change in the GDP level may not provide a clear validation of the model, we find that the growth of real GDP in the Scandinavian region did slow down from 2.4% in 2017 to 1.4% in 2018, based on the World Bank (2025) dataset. The reduced growth rate potentially reflects the economic consequences of the 2018 events on the aggregate level.

Furthermore, we also find consistent evidence in observed price trends. According to Statistics Sweden (2025) and the Swedish Forest Agency (2025), the average real price of sawlogs and pulpwood increased by 5.7% and 11.5%, respectively, in 2018 (real price adjustments based on CPI from Statistics Sweden). In Norway, the real price of timber products increased by 12% in 2018, based on data from Statistics Norway (2025). In Finland, the real price of timber products increased by 4% in 2018 (Statistics Finland, 2025). Similarly, the real price of electricity grows notably by 50% in Norway, 12% in Sweden and 5% in Finland.

Our study finds a consistent directional change in the prices of forestry products and electricity. However, the magnitudes observed in empirical data are somewhat larger than our CGE model evaluation. Several factors may explain this : 

1), the CGE model provides average changes across the agriculture, energy, and forestry sectors for Scandinavia as a whole, potentially smoothing out regional price variations.

2) The empirical price trajectory may reflect broader market dynamics, where supply-chain disruptions and energy market responses extend beyond Scandinavia, amplifying price changes beyond what the model captures at a regional level.

3) The CGE model represents an equilibrium state based on rational economic behaviour, but in real-world markets, short-term speculative behaviour can drive prices higher. For instance, following a supply shock due to heat, drought and fires, traders and businesses might anticipate further shortages of production of energy, agriculture and energy products, leading to price spikes that overshoot equilibrium predictions. Thus, while GDP comparisons remain challenging due to confounding factors, the model successfully captures the qualitative patterns of price adjustments observed in the market.

Results and Discussion: The discussion on the contrast between single- and multi-hazard modeling in the introduction is an interesting point. It would be valuable to illustrate how modeling multi-hazards adds insights compared to single-hazard models. Does it bring the predictions closer to ground truth with respect to impact assessments?

We thank the reviewer for this comment. We agree and will more explicitly illustrate, argue and discuss how modelling multi-hazards adds insights compared to single-hazard analyses.

The manuscript does a good job of highlighting the limitations. However, it remains unclear how these limitations quantitatively affect the estimates. How far do these uncertainties impact the impact estimates? Incorporating quantitative or qualitative validation would significantly strengthen the manuscript’s credibility and help readers assess the reliability of the findings.

Thank you for the comment. We intend to strengthen the credibility of findings by incorporating qualitative validation, particularly focused on the observed price trends, as discussed in the question of “What are the uncertainties in the macro-economic model? How well do the model’s predictions align with observed GDP changes in the affected regions?” 

We will also add paragraphs in the Discussion section.