The paper is interesting. It extends an approach developed by Krouma et al. (2024) to wind and precipitation forecasts.

Major comments

The reason why precipitation and wind speed are treated differently (with different versions of the SWG) is unclear and should be explained. This seems to add an unecessary complication to the manuscript.

A greater care should be used in the figures. Figure 2 mentions "Santander" on the right column, while it should be "Linkoping". The colors for Fig. 3 and 4 (among others) should be consistent across stations. Figure D1 is unreadable. What are the colors for? Figure E2 is very hard to read.

Who is likely to use such systems, which seem to require a complex layer of simulations on top of already complex datasets of ensemble forecasts?

Specific comments

Why are the procedures described in Secs. 3.1.1 and 3.1.2 relevant to extremes?

Please provide a map of the 9 stations (the reader might not be so familiar with European geography).

What do 20Q, 70Q or 90Q mean? This is not stated in the text. If 20Q is the 20th quantile of precipitation, this cannot be considered as extreme.

The "outperform" verb looks like an exageration. Since the SWGs are based on the forecast products, the least that can be achieved is an improved performance. The improvement for extremes is not clearly outlined.

This manuscript discusses statistical methods for the prediction of extreme precipitation and extreme wind and compound events. 

The statistical methods are based on stochastic weather generator (SWG) using analogues from ERA5. The verification focuses on 9 locations in Europe.

Major comments

The verification is based on scores applied to deterministic and probabilistic forecasts.

SEDI and PSS are scores to assess deterministic forecasts. Could the author describe more thoroughly to which forecasts these scores are applied to? This is not clear from the text.

ECMWF forecasts are used as a reference for computing BSS. ECMWF forecasts are not calibrated as illustrated in Figures 6 and 9, despite that it is claimed that they are in line 82. For scores focusing on extreme events, it is crucial to compared calibrated forecasts.

Specific comments

Line 79, "9km". Could you double check the resolution of ECMWF ensemble forecasts in TIGGE over the period 2017-2021? 

Line 82, "bias corrected". Could you indicate how?

Line 85, "local". How do you build a climatology? Which period do you use? Is it local in space and time?

Line 85, "excluding values below 1mm/day". Why? Why not considering higher percentiles instead?

Line 110, "95th percentile of the distribution for the full forecast period". What is forecast period again? How is defined the climatology? Is it a forecast climatology?

Line 121, "we use the Z500 field to estimate daily extreme precipitation". I don't understand this step. Could you please elaborate?

Line 187, "using δ = 5". Could you recall what δ stands for?

Figure 2, P> 95th Q. What does P refer to? Observation, or forecast, or both?

Figure 4, "forecasted" to be replaced by “forecast”.

Line 428, "forecast bias". The definition in the text and Equation A1 do not match.