Reviewer 1 

We thank the reviewer for taking the time to review our manuscript and for their constructive comments on our paper. We respond to each point below, where the reviewer comments are in regular text and our responses are in italics. 

Abstract  

In the abstract, the authors indicate that the temperature assessment is performed in the Kelvin unit (e.g., line 10), but the analysis in the paper seems to be in degree C.   

Corrected 

1. Introduction:  

The research question is well-defined and relevant.  

• Minor typos:  

o Line 28 – space after "risk".  

Corrected 

o Line 32 – change RCMS to RCMs.  

Corrected 

• Line 34 – add McGregor 2005: CCAM: Geometric aspects and dynamical formulation. Technical Paper 70. CSIRO Atmospheric Research, Melbourne. 

Added  

• The authors should mention and describe "the ESCI prototype" as they utilise it later in the discussion section to show BARPA-R's improvements. 

Added  

1. Data and Methods:  

• The authors could provide equations for the indices and metrics calculation.  

We have added the equations for the metrics near line 180 

• Minor typos:  

• Line 91 – “above ground level (AGL)” with capital first letters.  

Corrected 

• Line 131 – Remove the comma before "to a regular grid".  

Corrected 

1. Performance Evaluation:  

1. Mean State   

• Figure 2: The authors should improve labelling and caption – "TX, TN, and PRCPTOT" labels are misleading. One could assume these are indices, but they are referring to mean Tasmax as TX, mean Tasmin as TN and total precip as PRCPTOT; these need to be explained in the figure caption and in the running text (line 170). Also, indicate in the caption that the digits in the plots represent the averaged bias.  

corrected 

• Line 174 – The authors refer to a reduced diurnal range, which, I assume, is the difference between maximum and minimum temperature; they could elaborate on how they read this from the plots? Is that the case when Tasmin shows warm bias and Tasmax cold bias over a specific region? This could be explained in the running text; it's a bit confusing to me, and it might not be clear to everyone. Describing the Tasmin and Tasmax patterns separately (and referring to them as daily Tasmax/Tasmin) may be clearer than referring to "diurnal" patterns.   

We have added the following text immediately prior to this discussion: "When temperature biases show a decrease in maximum temperatures coupled to an increase in minimum temperatures in the same season, this can be interpreted as an underestimation of the diurnal temperature range." 

• Line 183: Is a reference missing "NWP (cite)"?  

Added Hudson et al 2017 

• Line 190: I assume with "daily maximums", the authors mean "diurnal maximum"?  

Removed all references to tasmax and tasmin as 'diurnal max/min' following your later recommendation 

• Line 190: It would be good to have some equations. 

We have added the equations for the metrics near line 180 

• Line 204-205: It would be good to have some equations.  

We have added the equations for the metrics near line 180 

• Line 205: Please indicate what SDII stands for and how it is calculated.  

Clarified to read: "SDII - the Simple Daily (precipitation) Intensity Index, which is calculated as the average precipitation rate across all days with at least 1 mm of precipitation." 

• Figures 3 and 4 – Please improve the caption: indicate what the indices abbreviations stand for, e.g., "number of summer days, SU, (with daily maximum temperatures exceeding 25 degrees C)", and so on.  

Corrected 

• Figures 5 and 6 – Add units for "Delta".  

Added 

1. Trend   

The agreement between BARPA-R and AGCD for temperature-related indices is encouraging, though the discrepancies for minimum temperature-based indices need further investigation. The challenges in simulating extreme rainfall events noted in the warm season are essential findings, and the potential reasons behind this should be explored further, which is probably done in the paper's discussion section?  

We have added the text to the discussion: both BARPA-R and ERA5 underestimate the intensification trend of wet-day precipitation (SDII) observed in AGCD. This result is consistent with global studies of atmospheric models which parameterise convection, and has been found to be rectified by the explicit representation of atmospheric convection (Fowler et al 2021, Lee et al 2022, Luu et al 2020). This is particularly true for sub-daily extreme rainfall, which has not been evaluated in this paper. Further downscaling of climate projections and regional reanalysis to convection permitting length-scales over the Australian region are therefore necessary for the assessment of changes in high intensity, short duration rainfall (Wasko et al, 2023).  

1. Interannual Variability   

The section focuses on three major modes of interannual climate variability, especially during the Austral Spring, which is highly relevant to understanding the performance of BARPA-R in capturing significant climate drivers in the Australian region. The results suggest that, overall, BARPA-R performs well in simulating the teleconnections associated with ENSO, IOD, and SAM.  

The authors could consider exploring the sensitivity of BARPA-R to different model configurations or parameters, which may help improve its performance in capturing interannual temperature variability associated with ENSO.  

Unfortunately, this suggested analysis is beyond the scope of this paper. We expect the response to ENSO to be highly dependant on the driving models in future GCM downscaling experiments. 

• Line 251 – what is meant by "diurnal maximum temperatures"? Is it daily Tasmax, or the range between Tasmax and Tasmin? Maybe the authors could avoid using the term "diurnal", as it might confuse some.  

Removed all references to tasmax and tasmin as 'diurnal max/min'. We have retained the term 'diurnal temperature range' to refer to tasmax-tasmin in section 3.1 

1. 10-meter Winds  

The section comprehensively evaluates BARPA-R's performance in simulating 10-metre wind speeds. The improvement in high percentile wind speeds compared to ERA5 (reference) is a positive finding, indicating that BARPA-R better captures extreme wind events in many regions. However, the persistent underestimation of calm weather conditions suggests that there is still room for refinement. The authors could explore whether model parameterisation or resolution adjustments can address this issue?  

This is an instrumental issue: the stations do not measure wind speeds between 0 and 2 m/s and instead record these low wind speeds as 0 m/s. To avoid confusion, we have removed these low percentiles from figure 8. 

• Line 262 – remove space between 1 and 0 "of 1 0m AGL".  

Corrected 

• Figure 8. – Is it possible to make this plot larger? The labels are hard to read.  

Done 

1. Process Evaluation:  

1. Circulation  

This section shows that the four identified large-scale circulation features are present in both BARPA-R and ERA5 (reference), and their seasonal cycles match. Identifying biases and discussing their potential causes contribute to a better understanding of model limitations and areas for improvement. How can the biases related to the monsoon westerlies and the SPCZ be addressed?  

This paper aims to document the key biases and limitations of BARPA-R. Resolving these biases would require considerable model development and is currently out of scope. Ocean coupling may help with the SPCZ biases. 

We have added the text: 

This westerly wind bias has been linked by Martin et al 2019 to errors in the representation of convection and precipitation over the Maritime Continent and the western/central equatorial Indian Ocean. Systematic rainfall biases in the maritime continent region are common due to the complex, multiscale nature of convection in this region. 

A reduced southerly bias in the location of the SPCZ has been documented in the ACCESS-S1 seasonal forecast system (Beischer et al 2021), suggesting that ocean coupling may improve the representation of the SPCZ.  

• Line 301 – Remove TODO "(not shown, TODO)"?  

Corrected 

• Figure 9 – Change the labels "barpa" and "era5" to upper case. And add colour bar labels.   

done 

• Tabel 1 – Captions for tables are placed above the table.  

Corrected 

1. Weather Systems   

The authors investigate BARPA-R's performance in simulating key large-scale weather systems that influence Australia, including tropical and extra-tropical cyclones and Australian Northwest Cloud-Bands (NWCBs). While extra-tropical cyclones are generally well represented, biases and decoupling of interannual variability are observed for tropical cyclones and NWCBs. Addressing these issues and further investigating the causes would enhance the overall understanding of BARPA-R's capabilities in simulating important weather systems in the Australasian context.  

• Line 325 – What are these direct observational products? I think they are listed in the caption of Figure 10, but they need to be mentioned here as well.  

This information is given in the next paragraph. This text has been improved as follows for clarity:  

"In this analysis, some direct observational products are available, and these are used as references where possible. For weather systems where no direct observation is available, ERA5 is used as the reference. Identification algorithms and reference datasets are described below." 

• Line 325 – The authors say that Obs are replacing ERA5, but as they mention in the caption of Figure 10, ERA5 is a part of "Obs". Please clarify.  

Please see correction above. ERA5 is used where no direct observation is available. 

• Line 339 – The first letters in upper case "outgoing longwave radiation (OLR)"  

Corrected 

• Line 344 – Please mention that ERA5 is part of "Obs"? Because you say in line 325 that Obs are replacing ERA5.  

Replaced 'ERA5' with 'the ERA5-based reference' 

• Tabel 2 – Captions for tables are placed above the table.  

Corrected 

• Figure 10 – Please add labels to the colour bars.  

done 

• Figure 10 – This part of the captions is not clear: "NWCBs (red lines; contour interval: 1 event/season, first contour: 2 events/season)"  

corrected 

Lagged Temperature-Precipitation Relationship:  

In this section, the authors calculate Spearman's ranked correlations with a lag time of ±10 days between diurnal temperature and precipitation outputs, focusing on DJF and JJA over different NRM clusters, and demonstrate that BARPA-R's capability to capture multivariate relationships, particularly between temperature and precipitation, as compared to observational and reanalysis datasets. However, there are regional variations that may have implications for climate impact assessments.  

How could these discrepancies be addressed? And what are the practical implications of these findings for climate impact assessments?  

We have added the text 

The skilful representation of the relationship between temperature and precipitation has implications for the interpretation of climate risk assessments of compound weather events. For example, hot and dry conditions may lead enhanced bushfire risk, while hot and humid conditions can lead to enhanced heat stress on humans and livestock. In regions where models struggle to represent the correct temperature-precipitation relationship, simulation of compound events may be adversely affected. Improvements to the representation of atmospheric convection, either through improved parametrisation or explicit simulation, may improve the minimum temperature – precipitation relationship in the northern Australian wet season. 

Line 370 – Maybe use the term daily temperature-precipitation relationship instead of "diurnal"; it can be confusing.  

Removed all references to diurnal  

• Figure 11 – Please add a legend.  

Corrected

• Figure 12 – Caption typo: minimum temperature instead of maximum temperature  

Corrected 

• Line 376 – Add the abbreviation "precip" after rainfall, as you start using it in the paragraph below.  

Replaced each occurrence of 'precip' and 'rainfall' with 'precipitation' 

• Line 392 – Remove the dot after "(Figure 12 f-h)."  

This is meant to be there: end of a sentence 

• Line 398 – Remove "between the two variables", it’s a repeat.  

corrected 

6. Discussion and Conclusion:  

Overall, the discussion provides a comprehensive summary of the key findings and implications of the study. It effectively communicates the strengths and weaknesses of BARPA-R, highlights improvements over previous RCMs, and underscores the importance of hazard analysis and ongoing work in downscaling CMIP6 GCMs. The conclusions drawn are well-supported by the study's results and contribute valuable insights to the field of regional climate modelling in Australia. The conclusion highlights the significance of the study and suggests avenues for future research.  

The authors could consider providing a more detailed discussion of the potential reasons behind BARPA-R's limitations, and the specific discrepancies observed between BARPA-R and the reference data and recommend potential strategies to address these limitations, e.g., exploring the sensitivity of BARPA-R to different model configurations or parameters.  

We have added the text:   

Many of the biases and limitations in BARPA-R identified by this study are common biases of the MetUM. These include the overall wet bias (Hudson et al 2017), the overestimation of the monsoon westerlies (Martin et al 2021) and the reduced diurnal temperature range during winter (Su et al 2023), Future development of BARPA-R will take advantage of the ongoing MetUM model development, such as the inclusion of the CoMorph convection scheme and updates to the Jules land surface model, with the potential of improving model shortcomings going forward. 

• Line 423 – Why is the cold bias referred to as "1K" when the analysis was in degrees Celsius?  

Corrected 

• Line 429 – The authors mention the ESCI prototype for the first time here. This needs to be referenced and briefly described in the introduction, too.   

Text added as follows in the introduction: "This model is a continuation of prototype work developed for the Energy Sector Climate Information (ESCI) project, documented by \cite{Su2019} and hereon referred to as ESCI-BARPA." 

• Line 431 – Add "C" after -1.1 degrees.