Major comments

In my view, there is currently one major flaw in the authors' work. This is relevant to the absence of any attempt to relate changes seen/projected in fire weather to changes in weather. For instance, what is driving the deterioration of fire weather conditions in the UK? Is it the wind? Is it temperature? I strongly advise the authors to try to identify the drivers of the increases seen in the indices. This would improve significantly the added-value of their work. For instance, they could use statistical methods, such as GAMs or neural networks, to attribute changes seen in the indices to changes in the basic meteorological variables. 

Minor comments

Sect. 1.2: I think that the title of this section is misleading. By reading "Causes of wildfire", I expected to read more on ignition sources (e.g., arson, lightning). Instead, this section focuses on what we often call the fire environment (fuels, weather, topography) and how it influences wildfire activity. Consider revising the title (e.g., "Drivers of wildfire activity"?) 

L65-69: This paragraph does not realy fit in this place. I would suggest adding a new section describing the concept of fire danger in more detail. I believe this is necessary, since this study focuses on fire danger. 

L155-156: A reference and some notes on how the thresholds were adapted for the UK would be highly welcome. 

L155-166: I believe that some references would enhance the manuscript, along with maybe more detailes on how the Canadian FWI system was adapted to the pyric environment of the UK. Simply saying, for eaxmple, that FFMC is a good indicator for the UK in spring does not really say much to the interested reader. More information would be appreciated. 

Sect. 2.3: It is not really clear if FWI was computed from the ERA5 data or the authors employed the readily available ERA5-based FWI dataset. If the first option was followed, why the use of the readily-available ERA5 FWI dataset was excluded? 

Figures 1-3: Burnt area should be reported in ha. 

Technical corrections

L132: "was calculated". 

L268: Revise the sentence. 

The authors reconstructed the historical fire weather in the UK and project the future changes at 2 °C and 4 °C global warming. Annual and seasonal variations in fire weather and spatial distribution of fire weather were analyzed. These were done by analyzing two fire weather indexes: ISI for spring fires and FWI for summer fires, which were calculated using historical and future climate data. The authors showed the historical patterns and projected that future “very high” fire weather in the UK would largely increase, especially in summer and in a larger warming scenario.

Major comments

In general, this paper was clearly and well written and provided a detailed analysis of the past and future trends in fire weather in the UK, which was not much focused on in the past. This is nice, of course. But as the authors claimed in the manuscript, fire weather does not represent actual fire events (i.e., differences in historical observed burned area and fire weather). If the authors would like to link more their results to policy suggestions on fires in the UK (this should be an important part for a regional study), more discussions about other drivers of “real” fire events will help, for example, how ignition patterns, fuel amount, land-use change, and fragmentations, suppression activities will change in the future and possible effects on fire events. I understand this is not your main focus, but I think including a separate paragraph in the Discussion regarding these other drivers will improve the broad accessibility of the paper.

Minor comments

L224: What do you mean by “annual cycle”? multiple-year-mean for every month? 

L293-295: Replication with figure legend, delete.

L364: Is this the 20-year mean value for panel (a)?