Cyclone generation Algorithm including a THERmodynamic module for Integrated National damage Assessment (CATHERINA 1.0) compatible with Coupled Model Intercomparison Project (CMIP) climate data

Le Guenedal, Théo; Drobinski, Philippe; Tankov, Peter

doi:https://doi.org/10.5194/gmd-15-8001-2022

Articles | Volume 15, issue 21

https://doi.org/10.5194/gmd-15-8001-2022

© Author(s) 2022. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/gmd-15-8001-2022

© Author(s) 2022. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 15, issue 21

Model description paper

|

07 Nov 2022

Model description paper |

| 07 Nov 2022

Cyclone generation Algorithm including a THERmodynamic module for Integrated National damage Assessment (CATHERINA 1.0) compatible with Coupled Model Intercomparison Project (CMIP) climate data

Théo Le Guenedal, Philippe Drobinski, and Peter Tankov

Download

Final revised paper (published on 07 Nov 2022)
Preprint (discussion started on 02 Dec 2021)

Interactive discussion

Status: closed

RC1:
'Comment on gmd-2021-384', Anonymous Referee #1, 13 Jan 2022

“Cyclone generation Algorithm including a THERmodynamic module for Integrated National damage Assessment (CATHERINA 1.0) compatible with CMIP climate data” describes CATHERINA which is a model that generates national tropical cyclone damages from monthly climate model data. CATHERINA takes the relative humidity, sea surface temperature, sea level pressure, and tropopause temperature from CMIP5 models. It then uses ERA5 reanalysis data to get more spatially refined estimates of these climate measures. The future estimates of each climate model are adjusted for the historic bias in the CMIP5 data. The model assumes the frequency of tropical cyclone origins does not change and the random path of tracks roughly conforms with past tracks. What changes is the power of each storm which is assumed to increase at an increasing rate with sea surface temperature depending on the ocean basin. They use data from Eberenz 2019 that uses night lights, population and national borders to measure how physical assets vary across space within a country. They then take a damage function from Eberenz 2020 that is calibrated to measure the tropical cyclone damage in each region. The result is an estimated annual damage cost for each country from tropical cyclones.

It is not clear that this tropical cyclone model leads to accurate forecasts of storms. Changes in wind patterns can have no effect in this model. Is the cyclone model in this paper as accurate as the models developed by Emanuel? Or is this model a step backwards?

The estimates of the effect of each hurricane are crude. The model assumes that all damage is from wind whereas only 40% of cyclone damage is wind related. Another 40% of cyclone damage is from storm surge. But storm surge strikes largely just the coastline. The remaining 20% of damage is from excess precipitation which often falls far from where the cyclone strikes land.

The model depends a great deal on the damage function. But it is not clear how this damage function was estimated.

The estimates of how national assets are distributed across space are crude. Light times population is not going to allocate national assets carefully. I am specifically concerned about how well they model the assets near the coast.

The model appears to assume the spatial distribution of assets are fixed within a country.

The paper does allow national assets to change over time, but they do not describe how this is done.

There is no effort to measure adaptation by the country being hit or how that might change over time.

The initial forecasts of windspeed from the climate models are very inaccurate. The corrections appear to matter a great deal. However, these corrections have been made are on the historic data. So once they adjust historic data to actual historic outcomes, they do fine. But how well the model predicts future wind speeds is unclear.

Figure 19 suggests the model predicts a small probability of very large damage but an expected value that is quite small. What explains this large tail to the distribution of damage? Is this simply the probability of a large storm striking a large coastal city? What is the expected value of damage?

Why does going from historic (1980-2020) to RCP2.5 lead to more damage than going from RCP2.5 to RCP8.5? Going from historic temperature to RCP2.5 is a 1C increase whereas going from RCP2.5 to RCP8.5 is going from 2C to 5.4C? Given the assumption that wind speed increases more rapidly as sea surface temperature rises, this outcome is hard to understand.

How much confidence do the authors have that they understand the relative damage caused by tropical cyclones at the end of the century across countries? How much of this is simply assuming the same distribution as today?

It is not likely that anyone could design adaptation measures from this study given the crudeness of both the tropical cyclone predictions as well as the damage predictions. Is there any reliable prediction of a change in tropical cyclone outcomes from current outcomes other than they will get uniformly more powerful?

Citation: https://doi.org/10.5194/gmd-2021-384-RC1
- AC1:
  'Reply on RC1', Théo Le Guenedal, 10 Feb 2022
  Thank you very much for your comments and very relevant questions. Before giving detailed answers to each of your comments, we would like to emphasize the aims and scope of our paper. Our methodology aims to provide country-level estimates for the future damage from tropical cyclones, consistent with climate model projections, with mainly economic and financial applications in mind. Examples of applications are estimating the impact of cyclones on the creditworthiness of government debt; providing a physical risk module for integrated assessment models; stress testing the resiliency of the financial system at a country level and at a global level to physical risk etc. Given this aggregate country-level analysis objective, our model is certainly a simplification compared to state-of-the-art cyclone dynamics models, does not aim for precise prediction of individual cyclone tracks, and does not integrate a bottom-up description of damage to individual assets. At the same time, our paper improves earlier studies of cyclone damage on the aggregate level (for example, compared to Mendelsohn et al. (2012): our model uses several climate scenarios with state-of-the-art bias correction, uses SSPs to project population and GDP, is based on precise asset value distribution etc.) We will include these clarifications in the introduction of the revised version of the manuscript. For reviewer’s convenience, we attach a supplement with further technical elements and figures.
  Q1 - Cyclone description accuracy It is not clear that this tropical cyclone model leads to accurate forecasts of storms. Changes in wind patterns can have no effect in this model. Is the cyclone model in this paper as accurate as the models developed by Emanuel? Or is this model a step backwards?
  We reiterate that building precise cyclone forecasts is not the aim of CATHERINA. We propose an algorithm designed to assess the future cost distribution for country-level damage assessment. The cyclone intensification process used is inspired from STORM model (Bloemendaal et al., 2020), which includes a single climate variable, and extended following Holland (1997) and Emanuel (1988) to encompass 2 more variables. We found that this extension provides statistically significant instrumental variables in the description of tropical cyclone intensification, which is the aim of the algorithm. Another step forward in our methodology is the use of state-of-the-art bias correction module for integrating climate model projections.
  Consequently, even if some thermodynamical processes have been simplified in this approach, our approach is still a step forward with respect to state-of-the-art in the context of integrated assessment models (IAM) for climate impact analysis. Indeed, our approach can integrate any CMIP simulation with limited set of available variables (only few vertical levels, some only available at monthly time scale, with some variables not always available). The adaptability of our algorithm to any CMIP exercise and simulation comes with a constraint implying necessary simplifications. Our approach combined with a bias correction module makes our algorithm easy to implement, more sophisticated in terms of processes included with respect to existing IAM and bias-corrected. We will emphasize this message in the section 3.2.3 of revised version of the manuscript.
  Q2 - Damage and sub-perils The estimates of the effect of each hurricane are crude. The model assumes that all damage is from wind whereas only 40% of cyclone damage is wind related. Another 40% of cyclone damage is from storm surge. But storm surge strikes largely just the coastline. The remaining 20% of damage is from excess precipitation which often falls far from where the cyclone strikes land.
  The model does not distinguish sub-perils, associated with key thermodynamical processes of cyclones (heavy precipitation, storm surge and associated flooding, strong winds) but instead uses a statistical relationship to estimate the global damage induced by a cyclone from a proxy variable given by the maximum wind speed (which is the proxy used in Saffir-Simpson Hurricane wind scale to define the intensity of a cyclone). We will include these elements in the third section (3.2.1) of the paper. The damage function is fitted on multiple events from the total damage reported in the global disaster database EM-DAT (Guha-Sapir et al., 2018). This database, used in most studies on the topic, accounts for the total reported damage (sum over all sub-perils) and does not distinguish damages from sub-perils. So, despite the relevance of the reviewer’s comment, for our application (see our comment in the introduction of our reply, which are now included in the manuscript introduction to clarify the context of our study), distinguishing the sub-perils generating the impacts is not needed.
  Q3 - Damage function calibration The model depends a great deal on the damage function. But it is not clear how this damage function was estimated.
  We use region-specific damage functions from Eberenz, Lüthi, et al. (2020). This method uses a parametric function following Emanuel (2011). The parameters are estimated for each region with machine learning techniques from the reported damage estimates in the International Disaster Database (EM-DAT) Guha-Sapir et al. (2018) crossed with cyclone tracks (IBTrACS), and geographic and socio-economic information along these tracks.
  We reiterate the main step of the optimization performed by Eberenz, Lüthi, et al. (2020) and Lüthi (2019) to define the regional damage functions. The authors first defined the event damage ratio (EDR) as a fraction error between normalized reported (NRD) and simulations (SED) for each cyclone and the total damage ratio (TDR) is defined in each region summing over events. For each event, there is a value for vh allowing to optimally calibrate the explicit damage function described in Emanuel (2011). Then, the authors proposed two complementary optimization methodologies to find the value of vh maximizing the prediction of the regional damages Eberenz, Lüthi, et al. (2020):
  Root mean square fraction (RMSF), minimizing the spread of the event damage ratios (EDR) – defined as the ratio of simulated damage vs. reported damage.
  
  Total damage ratio (TDR), finding the value of vh, such as the ratio of total simulated damage – obtained summing over event damage – and total reported damage tends to 1.
  
  We will clarify the calibration of these functions in section 4.3. In particular, we will review the approach of Eberenz, Lüthi, et al. (2020) to find the values of vh (c.f. technical supplement).
  Q4 - National Asset The estimates of how national assets are distributed across space are crude. Light times population is not going to allocate national assets carefully. I am specifically concerned about how well they model the assets near the coast.
  We chose to build our model based on state-of-the-art estimates, in such a way that the methodology is uniform country-wise. This dataset (Eberenz, Stocker, et al., 2020) is also used for the calibration of damage functions in Eberenz, Lüthi, et al. (2020) (discussed in Q3). Therefore, the use of this data allows to estimate the exposure in a consistent manner. To verify the accuracy of estimation, a back-test has been performed (Section 4.4). As we mention in the beginning, the only way to improve the estimates of asset value distribution would be to use the actual asset distribution from asset-level databases, but such databases are not yet available at the global scale. We will add this explanation when introducing the exposure dataset section 2.4.
  Q5 - Spatial distribution dynamics The model appears to assume the spatial distribution of assets are fixed within a country.
  The model assumes that the spatial distribution varies with population changes proposed in the Socio-Economic Data Application Center dataset presented by Jones and O’Neill (2017, 2020). In particular, the spatial distribution of the population is different in varying shared socioeconomic pathways (SSPs). These projections are available with a one-eigth degree resolution. Figure 1 (in the supplement) represents this multiplicative factor in the SSP2 (1a), SSP3 (1b), SSP4 (1c) and SSP5 (1d) in 2100. The revised version will include a subsection to better describe the exposure dynamics (spatial and temporal) lacking in the current manuscript (c.f. technical supplement).
  Q6 – National asset dynamics The paper does allow national assets to change over time, but they do not describe how this is done.
  To estimate future exposures along the cyclone track in each scenario, we use the downscaled estimation for the exposed wealth and the coefficients representing the change between the current state and the future scenario. We use the most granular projections of GDP per capita variation curves (Figure 2 – Data Source : https://tntcat.iiasa.ac.at/). Binding the two (regional GDP per capita and local population) we build a dynamic projection of exposure factor. Similarly as for Q5, the asset-exposure dynamics will be further detailed in the final version of the paper (c.f. technical supplement).
  Q7 - Adaptation There is no effort to measure adaptation by the country being hit or how that might change over time.
  Indeed, we left this question for further research. Supposing that adaptation increases with time alone would not be a relevant hypothesis. However, this question could be one of the direct applications of the model. For example, measuring the investment costs required to shift the value of vh or vt – and thus reduce the risk of future damage – can be a research question derived from this model simulations. In the revised paper, we will present more clearly the possible application of the model integrating the adaptation scenario, changing the values for the vulnerability parameter (vh and vt) in the section 4.2.
  Q8 - Bias control The initial forecasts of windspeed from the climate models are very inaccurate. The corrections appear to matter a great deal. However, these corrections have been made are on the historic data. So once they adjust historic data to actual historic outcomes, they do fine. But how well the model predicts future wind speeds is unclear.
  Our bias correction approach is the standard in the climate community (see http://ccafs-climate.org/bias_correction/)[1]. We do not have reanalysis data for the future. Therefore, there is no ’reference’ value to evaluate the prediction of the model. This is why we control the bias using the past distributions, where we can compare climate models and reanalysis and assume that errors between the two are similarly distributed in the future. We reiterate that this assumption is relatively classical in the climate community and we will integrate these precisions in the paper in section 5.1.
  Q9 - Results: quantiles vs. expected Figure 19 suggests the model predicts a small probability of very large damage but an expected value that is quite small. What explains this large tail to the distribution of damage? Is this simply the probability of a large storm striking a large coastal city? What is the expected value of damage?
  We ran the 7 models over 300 representative years to obtain these distributions. There is an effect due to certain large coastal cities exposure for the ’very unlikely’ band (between 95 to 99 percentile) of annual damages. However, given the scale observed more than one city have been hit by storms. Because the aim of the model was also to stress test the resiliency of the financial and economic systems, looking at the expected value of damage was less interesting that studying the quantile value especially in the context of events with large tail risk. Coronese et al. (2019)[2] investigating the increase of economic damage due to extreme natural disasters supports this thesis showing that the impact of climate change is particularly striking for extreme events (See for example, Coronese et al. 2019, Figure 2A). The table containing the expected value of damage after bias correction is in the technical appendix. The revised version will integrate this summary table with the expected value of the damage in the section 5.2 as well as the precisions above to explain the focus on quantiles in the visualization.
  Q9 - Results: SSP vs. RCP components Why does going from historic (1980-2020) to RCP2.5 lead to more damage than going from RCP2.5 to RCP8.5? Going from historic temperature to RCP2.5 is a 1C increase whereas going from RCP2.5 to RCP8.5 is going from 2C to 5.4C? Given the assumption that wind speed increases more rapidly as sea surface temperature rises, this outcome is hard to understand
  Socio-economic change leads to wider differences than climate change, and this was expected (cf. Mendelsohn et al. (2012), Figure 3 for example). The explanation for this is contained in the dynamics of (i) GDP and (ii) population in SSPs. In the revised version we add further explanation about this result including more references to discuss the results of our simulations.
  Q10 - Results: Countries damages How much confidence do the authors have that they understand the relative damage caused by tropical cyclones at the end of the century across countries? How much of this is simply assuming the same distribution as today?
  Thank you for this very interesting question. We can see in Figure 4 (20 in the paper) that the distribution across countries is different from one SSP to another. For example, we have sensibly the same distribution in SSP2 and SSP5 with a higher expected damage in SSP5 because of the growth hypothesis this scenario relies on. However, SSP3 (rocky road) or SSP4 (inequality) are distributed differently. The scenario emphasizing inequalities –and its interpretation by scientists in terms of (i) socioeconomic developments (Riahi et al., 2017) and (ii) population distribution (Jones & O’Neill, 2017) – increases damage concentration in the United-States. On the other hand, the rocky-road scenario, linked to higher and more rural population, lower GDP and national rivalry sees the damage more equally distributed on other nations. We integrate this precision in the final version.
  Q11 - Overall critics It is not likely that anyone could design adaptation measures from this study given the crudeness of both the tropical cyclone predictions as well as the damage predictions. Is there any reliable prediction of a change in tropical cyclone outcomes from current outcomes other than they will get uniformly more powerful?
  The current dataset – with low resolution data, and maybe not entirely sufficient number or realizations – might not be accurate enough to calibrate properly adaptation measures. However, we believe that the framework presented here is perfectly adapted to project a dense set of trajectories, compute expected and damage percentile over the next decades and therefore measure the investment required to either adapt, mitigate or include a migration factor in global economic modeling in the next fifty years. This work also reflects a practical exercise not carried out until now, which makes it possible to cross-reference the latest data sets developed, putting into perspective both the socio-economic and climatic development hypotheses, and to carry out a bottom-up, rather than top-down, damage calculation. The conclusion of the revised manuscript will mention the limits of the current application and better explain the scope of applicability of the model.
  [1] Navarro-Racines, C., Tarapues, J., Thornton, P., Jarvis, A., and Ramirez-Villegas, J. 2020. High-resolution and bias-corrected CMIP5 projections for climate change impact assessments. Sci Data 7, 7, doi:10.1038/s41597-019-0343-8
  [2] Coronese, M., Lamperti, F., Keller, K., Chiaromonte, F., & Roventini, A. (2019). Evidence for sharp increase in the economic damages of extreme natural disasters. Proceedings of the National Academy of Sciences, 116(43), 21450-21455, https://www.pnas.org/content/116/43/21450
  
  Citation: https://doi.org/10.5194/gmd-2021-384-AC1
RC2:
'Comment on gmd-2021-384', Anonymous Referee #2, 30 Mar 2022
Overview:

The paper presents a novel approach to evaluate damage from tropical cyclones at a national scale, and allowing for the influence of climate change on both TC intensity and economic exposure. The system is composed of TC track and intensity generation model, nationally-applicable damage functions and bias-correction techniques for key climate variables. These components have potential to provide a reliable approach to estimating future economic losses from TCs that incorporates changes to both the physical phenomena driving damage, and the local asset values impacted by TCs. In a topic that is receiving significant attention in the literature (e.g. Jewson 2021, 2022; Steptoe et al. 2022), this approach provides another view on evaluating TC-related risk, further expanding our understanding of epistemic uncertainties.

Detailed comments:

Table 1: the selection of GCMs used should be justified. This could be through reference to model performance literature for key parameters, a specific evaluation process or perhaps simply availability of required variables for the analysis (though I note the variables used are available for all CMIP5 models).

Section 2.2: the MSLP from ERA-5 is sampled 500 km from the centre of the cyclone. Is the same done for the other variables? Since the data are sampled from monthly means, it's possible the sampled values may not accurately represent the conditions at the time of TC passage (especially relevant for variables with sharp gradients such as SST).

Section 3.2: Given the literature of TC track generation methods, comparison with common metrics is encouraged. Specifically, as landfall is critical to reliable performance of a damage model, it would be helpful to present a comparison of the observed and simulated landfall rates (see for example Hall and Jewson, 2007; Lee et al., 2018; Arthur, 2021). This would strengthen the quality of the track generation results significantly.

Eq 3 - note that most best track data used wind pressure relations (WPRs) to determine Pc. Typically the work flow involves determining the Dvorak T number, converting this to a sustained wind speed, followed by regionally-specific WPR to determine Pc. The conversion back to wind speed from reported Pc using a single WPR will introduce errors, as an array of WPRs are used to operationally estimate Pc, not only between basins but within basins as well (e.g. Harper, 2002; Courtney and Knaff, 2009; Courtney and Burton, 2018; Courtney et al. 2021).

Eq 10 describes the dominant control on the maximum intensity of TCs (maximum pressure drop - MDP). This is tied only to SSTs. The model uses maximum potential intensity (MPI) to control the depression dynamics (i.e. intensification rates). The formulation of MPI is directly applicable to the problem of estimating the maximum intensity, accounting for factors beyond SST alone that control maximum intensity. This suggests using SST as the only predictor of the MDP is deficient.

Further, Chen et al. (2021) suggest rapid intensification is dependent on dynamical (e.g. upper divergence and wind shear) as well as thermodynamical factors. While the difference between Pc and MPI is a factor in predicting rapid intensification, and the dynamical factors are probably accounted for by the random innovation (Eq. 12), these other dynamical factors should be acknowledged.

Apply CDF-t to model variables, then evaluate MPI - I suggest comparing quantiles of ERA5 MPI against the bias corrected CMIP MPI values to demonstrate the effect of bias correction. Q-Q plots would be an effective way to do this. One risk with this approach is that correcting individual variables may lead to unrealistic combinations when evaluating MPI - e.g. extremely low tropopause temperatures in combination with very high SSTs that lead to unrealistic lapse rates and therefore unrealistically large MPI. Two solutions present themselves: 1) apply the bias correction methods to calculated MPI or (2) consider the joint distributions of variables when evaluating the bias corrections.

The distributions of SST presented in Figure 16 do not appear representative of SSTs sampled in the vicinity of TCs, and is inconsistent with the distribution shown in Figure 10. SSTs of 26C (299K) are typically considered a lower bound for TC formation (Gray, 1979), but median values from the ERA5 are well below that - for example based on Figure 16 the median SST for the South Pacific basin along synthetic tracks is 290-292K, for the Western Pacific 295K. Only the N Indian basin has a median SST near 300K. This suggests that the synthetic tracks are traversing areas not typically covered by TCs, or occurring at the wrong time of year for the respective basin leading to the unusual SST distribution.

Completely absent is any discussion on TC rates in the projections. Comprehensive literature reviews and expert elicitations indicate a global decline in TC frequency (albeit with generally low-medium confidence) (Knutson et al. 2020). Changes in TC rates will have a significant impact on the annualised losses. This is an important component that should be addressed.

In parallel, there is no discussion on changes in track behaviour. Observed trends in TC translation speed (Kossin, 2018) and poleward migration of maximum intensity (Kossin et al., 2014) should be considered in projections of TC activity. This has profound implications for TC-related risk in key marginal areas (e.g. Bruyere et al., 2020) where vulnerabilities are high, but present-day frequency of TCs is low.

Section 5.2: Consideration of SSPs in determining the effects on damage is novel, but the explanation is very limited. Given growth of exposure is constrained in existing high exposure regions, regional growth may not be in areas exposed to TC impacts.

The description of the implementation of projections of local physical asset value dynamics is very limited, but probably the most novel part of the connected modelling system. There should be a more substantial discussion on how the SSP definitions are used to modify asset values.

Technical comments

Page 4, footnote 2: Please use the full reference for the Copernicus Climate data store (Hersbach et al., 2020)

Figures need to be larger to be legible.

Figure 3: Recommend plotting each track with the same vertical scale (on the pressure and wind axes respectively) - i.e. use a scale of 0-75 m/s for wind speed and 880 - 1025 hPa for pressure on all panels. This will aid intercomparison of the time histories

Line 135: Please include an equation label

Page 7 - footnote: References to World Bank (2019b) and Credit Suisse Research Institute (2017) in the footnote of page 7 are not included in the bibliography.

Page 11: Footnote 13 should be in the body of the manuscript, as this is a key difference between James and Mason (2005) and the implementation in the current study. Following on from this, Tables A2, A4 and A6 reflect basin-wide fits, while the tracking method uses 5-by-5 degree grid. It may be more appropriate to provide maps of the relevant coefficients on the grids in the Appendix rather than the tables.

Page 17, line 289: Does "This study" refer to the current manuscript, or to the previously referenced Unawa et al. (2000). Please clarify.

Page 20, line 333: Change "non-EOCD" to "non-OECD"

Page 24, line 376: suggest changing "unbiased" to "bias-corrected"

Figure A8: Add units to the horizontal axes of the plot, or indicate what the values are in the caption. Ideally, each of the sub-plots should also use the same horizontal scale to aid comparison

References:

Arthur, W. C., 2021: A statistical–parametric model of tropical cyclones for hazard assessment. Nat. Hazards Earth Syst. Sci., 21, 893–916, https://doi.org/10.5194/nhess-21-893-2021.

Bruyère, C., and Coauthors, 2020: Severe Weather in a Changing Climate. Insurance Australia Group, https://www.iag.com.au/sites/default/files/Documents/Climate%20action/Severe-weather-in-a-changing-climate-2nd-Edition.pdf

Chen, Y., S. Gao, X. Li, and X. Shen, 2021: Key Environmental Factors for Rapid Intensification of the South China Sea Tropical Cyclones. Frontiers in Earth Science, 8.

Courtney, J. B., and A. D. Burton, 2018: Joint Industry Project for Objective Tropical Cyclone Reanalysis: Final Report. Bureau of Meteorology, Australia.

——, G. R. Foley, J. L. van Burgel, B. Trewin, A. D. Burton, J. Callaghan, and N. E. Davidson, 2021: Revisions to the Australian tropical cyclone best track database. Journal of Southern Hemisphere Earth Systems Science, 71, 25, https://doi.org/10.1071/ES21011.

Courtney, J., and J. A. Knaff, 2009: Adapting the Knaff and Zehr wind-pressure relationship for operational use in Tropical Cyclone Warning Centres. Australian Meteorological and Oceanographic Journal, 58, 167.

Gray, W. M., 1979: Hurricanes: Their formation, structure, and likely role in the tropical circulation. Meteorology over the tropical oceans, D.B. Shaw, Ed., Royal Meteorological Society, 155–218.

Hall, T. M., and S. Jewson, 2007: Statistical modelling of North Atlantic tropical cyclone tracks. Tellus A, 59, 486–498.

Harper, B. A., 2002: Tropical Cyclone Parameter Estimation in the Australian Region: Wind Pressure Relationships and Related Issues for Engineering Planning and Design. Systems Engineering Australia, Pty. Ltd.

Hersbach, H., and Coauthors, 2020: The ERA5 global reanalysis. Quarterly Journal of the Royal Meteorological Society, 146, 1999–2049, https://doi.org/10.1002/qj.3803.

James, M. K., and L. B. Mason, 2005: Synthetic Tropical Cyclone Database. Journal of Waterway, Port, Coastal, and Ocean Engineering, 131, 181–192, https://doi.org/10.1061/(ASCE)0733-950X(2005)131:4(181).

Jewson, S., 2021: Conversion of the Knutson et al. Tropical Cyclone Climate Change Projections to Risk Model Baselines. Journal of Applied Meteorology and Climatology, 60, 1517–1530, https://doi.org/10.1175/JAMC-D-21-0102.1.

——, 2022: Application of uncertain hurricane climate change projections to catastrophe risk models. Stoch Environ Res Risk Assess, https://doi.org/10.1007/s00477-022-02198-y.

Knutson, T. R., and Coauthors, 2020: Tropical Cyclones and Climate Change Assessment: Part II: Projected Response to Anthropogenic Warming. Bulletin of American Meteorological Society, 101, E303–E322, https://doi.org/10.1175/bams-d-18-0194.1.

Kossin, J. P., 2018: A global slowdown of tropical-cyclone translation speed. Nature, 558, 104–107.

Kossin, J. P., K. A. Emanuel, and G. A. Vecchi, 2014: The poleward migration of the location of tropical cyclone maximum intensity. Nature, 509, 349–352.

Lee, C.-Y., M. K. Tippett, A. H. Sobel, and S. J. Camargo, 2016: Rapid intensification and the bimodal distribution of tropical cyclone intensity. Nature Communications, 7, 10625, https://doi.org/10.1038/ncomms10625.

Steptoe, H., C. Souch, and J. Slingo, 2022: Advances in numerical weather prediction, data science, and open-source software herald a paradigm shift in catastrophe risk modeling and insurance underwriting. Risk Management and Insurance Review, n/a, https://doi.org/10.1111/rmir.12199
Citation: https://doi.org/10.5194/gmd-2021-384-RC2
- AC2: 'Reply on RC2', Théo Le Guenedal, 15 Apr 2022
  
  Dear Referee,
  Thank you very much for your thorough review of our manuscript and for your detailed comments that are all very relevant, constructive and will allow us to improve our initial submission.
  We provide a detailed answer to each of your comments in the attached document.
  Kind regards,
  Théo Le Guenedal
  On behalf of all manuscript authors
  
  Citation: https://doi.org/10.5194/gmd-2021-384-AC2
RC3:
'Comment on gmd-2021-384', Anonymous Referee #3, 07 Apr 2022
General Comments

This modeling system provides a much-needed bridge between global climate models and climate risk assessment. It represents an important contribution to what is an active area of research. What is presented is not a new model in itself, but rather a new combination of existing models, connecting across dynamical climate models, a statistical tropical cyclone track model, a bias correction procedure, exposure datasets, and damage functions. I congratulate the authors on producing this end-to-end modeling system that connects concepts across multiple disciplines. This is no easy task. The work is a substantial technical contribution and should be well received by the risk modeling community.

The work is well motivated with strong reference to most of the key prior studies. I also appreciate that the code and some data are publicly available and easily accessible. This supports reproducibility of the work. The methods are generally valid, though I do have requests to elaborate further on the data and methods (see specific comments below). The presentation quality is generally good although the manuscript would benefit from attention to grammar. I’ve called out a small number of errors at the end of this review, but there are many other spelling and grammatical errors that need attention.

The subject matter is appropriate for GMD and is worth being published after my comments below have been addressed.

Specific Comments

My expertise is in climate and tropical cyclone modeling so my specific comes from that background.

Line 17: I disagree that we are lacking tools to assess impacts of future TCs. See for example Geiger et al. (2021)

Geiger, T., Gütschow, J., Bresch, D.N., Emanuel, K. and Frieler, K., 2021. Double benefit of limiting global warming for tropical cyclone exposure. Nature Climate Change, 11(10), pp.861-866.

Line 7 and Line 390: I disagree with the claim that the framework is ‘a simple solution’. The framework requires expertise across multiple disciplines.

Line 32-34: It seems odd to make this assertion in the introduction without any supporting evidence. I suggest reframing this statement as a hypothesis to be tested.

This is perhaps my most important comment. I don’t think the difference between your TC model and STORM is made clear enough. STORM appears to use the same SST-pressure drop relationship as you do, and STORM also uses MPI (calculated using the Bister and Emanuel formulation) to limit TC intensification. I don’t understand what is new in your TC intensity formulation. Please clarify exactly what is new in the text. Is it the use of local MPI and SST along the synthetic tracks?

On a related note, the paper highlights the importance of this new representation of the thermodynamic influence, and makes claims on lines 43-45 that is it better, but this has not been demonstrated. Is it possible (if not too onerous) to run projections with and without this new representation of thermodynamic influence to demonstrate its importance.

It’s not clear to me how you calculate local SST and MPI along the synthetic tracks. If I am correct, the synthetic track generation samples from the IBTrACS record. If so, how do you assign a calendar year to each synthetic track to extract SST and MPI (from either ERA5 or CMIP)? If it’s a random year then the environment might not necessarily be favorable for the synthetic TC (i.e., too cool SST or low MPI).

ERA5 is still too coarse resolution to capture the most intense TCs. I suggest on Line 110 to change to ‘better resolves than climate models’.

Line 110-113: Your method to use data away from the storm center is fine but I don’t think it’s necessary. You are using monthly data that should smooth out the influence of TCs. This is just a comment – I’m not suggesting to make a change.

Line 117: I note that ERA5 is now available back to 1950, but is considered preliminary.

Line 122: Please be more descriptive of what you mean rather than the ambiguous term ‘erratic’.

I’m not sure what I learned from Fig. 3. I think this can be removed.

Section 5: I think it would be useful to remind readers that you are keeping TC frequency and genesis distribution constant.

Line 278-279: Please further explain why you wait 3 steps before applying the decay.

Technical Corrections

Fig 1: Correct ‘Tranform’ to ‘Transform’

I don’t see a reference in the text to Figure 5.

Figure 8: Please explain the distinction between the red shading vs. the red tracks.

Line 81: Please correct ‘AOCGM’ to ‘AOGCM’ and expand the acronym.

Line 273: Correct ‘Algorithm 1’ to ‘Figure 1’.

The reference to Figure 18 should be to Figure 17. If I am correct, then I’m also not seeing a reference in the text to Figure 18.
Citation: https://doi.org/10.5194/gmd-2021-384-RC3
- AC3: 'Reply on RC3', Théo Le Guenedal, 15 Apr 2022
  
  Dear Referee,
  Thank you very much for your thorough review of our manuscript. We are grateful for your overview, which highlights the relevance of our contribution and complexity of the exercise.
  Your comments are all very welcomed and will allow us to improve our initial submission.
  In the attached document, we provide a detailed answer to each of your comments.
  Kind regards,
  Théo Le Guenedal
  On behalf of all manuscript authors
  
  Citation: https://doi.org/10.5194/gmd-2021-384-AC3

Peer review completion

AR: Author's response | RR: Referee report | ED: Editor decision | EF: Editorial file upload

AR by Théo Le Guenedal on behalf of the Authors (09 Jun 2022) Author's response Author's tracked changes Manuscript

ED: Referee Nomination & Report Request started (19 Jun 2022) by Jinkyu Hong

RR by Anonymous Referee #1 (28 Jun 2022)

Suggestions for revision or reasons for rejection

“Cyclone generation Algorithm including a THERmodynamic module for Integrated National damage Assessment (CATHERINA1.0) compatible with CMIP climate data” contains two distinct efforts. On the one hand, the model forecasts tropical cyclone climatology directly from GCM outputs. On the other hand, the model predicts damage in future regions based on these simulated future cyclones.

The modelers have gone to great effort to reproduce the historic climatology of global tropical cyclones over the last 40 years. The aggregate difference between their simulations of the last 40 years and the historic cyclones is small. There is no similar way to test whether the model captures the future behavior of cyclones in new climate conditions. What you can see is that the future cyclone results are uncertain. There is no clear statistical difference between cyclone characteristics under RCP2.6, RCP4.5, and RCP8.5 by the end of the century. But this may be because the true difference is small not necessarily because the model is flawed.

The modeling of the damage from the cyclones does not seem nearly as careful as the modeling of the cyclones themselves. The paper makes an internally inconsistent assumption that emissions are independent of economic outcomes. However, that is not possible- just as it is not possible for solar radiation to increase without warming ocean temperatures. Without the 3% per year growth of GDP in SSP5, one cannot generate enough emissions to see RCP8.5.

The paper assumes that the aggregate damage from a cyclone increases with the cube of wind speed. But empirical research has shown that aggregate damage increases faster than the cube (Nordhaus 2010). Further, aggregate damage is more accurately predicted by minimum pressure than maximum wind speed (Mendelsohn et al. 2012).

It is not totally clear, but it appears that the authors have assumed damage will increase proportionally with GDP and population over time. The paper makes no mention of adaptation to tropical cyclones. However, it is quite clear that all countries but the USA have adapted to tropical cyclones (Bakkensen and Mendelsohn 2019). Damage has increased only slightly with GDP and population density in almost every country. Only in the USA have damages increased proportionally with GDP. Even in the USA, is it reasonable to assume that this will continue to 2100?

The link between simulated tracks and expected future damage is not clear in the paper. The damage depends on the strength of the cyclone, whether it strikes at low or high tide, and whether it hits a major metropolitan area or not. Are there enough simulated cyclones to get the proportion of these different outcomes correct for each country?

The authors never explain how CATHERINA1.0 can possibly lead to an estimate of the risk to sovereign debt or other major macroeconomic outcomes.

Most of the future results in the paper do not come from a change in the cyclones themselves but rather in the damage each cyclone causes. This is entirely because the authors are assuming there will be a lot more in harm’s way in the future. It is possible but not at all likely that the world will be this foolish.

Bakkensen, L. and R. Mendelsohn. 2016. “Risk and Adaptation: Evidence from Global Hurricane Damages and Fatalities” Journal Association Environmental Resource Economics 3(3): 555-587.
Mendelsohn, R., K. Emanuel, S. Chonabayashi, and L. Bakkensen. 2012. “The Impact of Climate Change on Global Tropical Cyclone Damage” Nature Climate Change 2: 205-209.
Nordhaus, W.D., 2010. The economics of hurricanes and implications of global warming. Climate Change Economics, 1(01): pp.1-20.

Hide

RR by Anonymous Referee #3 (14 Jul 2022)

ED: Reconsider after major revisions (26 Jul 2022) by Jinkyu Hong

AR by Théo Le Guenedal on behalf of the Authors (06 Sep 2022) Author's response Author's tracked changes Manuscript

ED: Referee Nomination & Report Request started (07 Sep 2022) by Jinkyu Hong

RR by Anonymous Referee #3 (16 Sep 2022)

ED: Publish subject to minor revisions (review by editor) (18 Sep 2022) by Jinkyu Hong

AR by Théo Le Guenedal on behalf of the Authors (26 Sep 2022) Author's response Author's tracked changes Manuscript

ED: Publish subject to technical corrections (28 Sep 2022) by Jinkyu Hong

AR by Théo Le Guenedal on behalf of the Authors (29 Sep 2022) Manuscript

Post-review adjustments

AA: Author's adjustment | EA: Editor approval

AA by Théo Le Guenedal on behalf of the Authors (26 Oct 2022) Author's adjustment Manuscript

EA: Adjustments approved (30 Oct 2022) by Jinkyu Hong

Short summary

The CATHERINA model produces simulations of cyclone-related annualized damage costs at a country level from climate data and open-source socioeconomic indicators. The framework couples statistical and physical modeling of tropical cyclones to bridge the gap between general circulation and integrated assessment models providing a precise description of tropical-cyclone-related damages.