The paper investigates the wind resource around south-east Australia during peak electricity demand associated with abnormally cold and hot days. For this purpose the authors use three years of consumption data and 29 years of reanalysis data. They estimate the high demand days using a random forest model trained for which both consumption and reanalysis data was available. They conclude that particular weather patterns are associated with peak demand days. 

The paper is structured well, is well written and provides the necessary amount of information, only the quality of the figures could be improved as well as the captions. The conclusions are clear and provide a nice overview of the potential of offshore wind being able to provide the necessary power supply on peak days. 

More detailed comments are given in the attached PDF. 

This study examines the wind resource across southeastern Australia during high electricity demand days, including both warm and cold days. The analysis combines three years of energy consumption data with 29 years of reanalysis data.

The study concludes that wind resources and energy consumption data align well on both warm and cold high-demand days. 

Overall, I find the study being well-structured and a significant step toward making more informed decisions about offshore wind farm placement. That said, there is room for improvement, and my specific comments are outlined below. Also, as the authors noted at some point, I would be interested in seeing how higher spatial and temporal resolution data—capable of capturing complex topography, coastline features, and daily flow variations—could complement the findings presented in this study.

More specific comments and suggestions:

• Line 34: “This emphasizes that the variation in wind power is a combination of synoptic-scale processes and local-scale processes”
  
  How are these local processes represented in the dataset?
• Line 97: “some errors in the direction of the perturbations around areas of complex coastline.”
  
  Are these errors reflected in the patterns of any variables used in this study? If so, how?
• I recommend including a high-resolution regional map that shows both topography and coastline features, so readers can better understand the regional characteristics influencing the findings.
• Line 149: “False Alarm Rate (FAR) and Probability of Detection (POD)”
  
  Please include a reference or justification for selecting these two metrics.
• Line 151: “The predictors for the best performing model for high-demand hot days were CDI, Tmax, Tav, RH, WS, and DOW, where all predictors were significant at the 5% level except for RH, and Tav, while the predictors for the best performing model for high-demand cold days were HDI, Tmax, RH, Tmaxlag1, and DOW, where all predictors were significant at the 5% level.”
  
  It would be helpful to include the significance level  for each variable in the final models. Which variables had the strongest statistical significance?
• Line 156: “Although there was some variation in the order of the most important predictors…”
  
  Again, please include the significance levels for each variable. If there is a variation in the order of predictor importance, where might this variation stem from? How do you explain it?
• Line 161: There appears to be a typo in Table 1 (should “Hot80 RF” be “Cold80 RF”?). Additionally, how do you explain the performance of the Cold80 LR model?
• Line 183: “Average wind capacity factors”
  
  Please clarify how this factor is calculated or assessed.
• Figures 8, 9, and 10: The plots should have clear titles or labels indicating which represent Hot80 and which represent Cold80 days.
• Line 281: “While not examined in this work, we also note that certain synoptic patterns may be associated with concurrent hazards…”
  
  While this is an important point, it feels somewhat disconnected from the rest of the discussion. Consider linking it more clearly to the implications for demand/management during extreme weather events.