The whole section on future climate projections under different scenarios is superbly explained. The information is well delivered and backed up by relevant peer-reviewed literature.

This is an accurate description of event attribution.

The Forest Fire Danger Index (FFDI) does indeed point to worsening conditions. It's worth noting though that it's very hard to characterise bushfire risk using a single index and the FFDI wasn't designed with a changing climate in mind. In general, there is a great deal more confidence in projections for extreme weather like heatwaves than there is for bushfire weather.

This is a great way of communicating a more nuanced change in the climate.

The information is clearly displayed and the source of the data is given.

This article uses relevant information on extremes to help show readers the impact of climate change through information less abstract than the global temperature.

The report focusses on possible scenarios very much on the extreme end of what could happen but then claims there's a "high likelihood" of human civilisation ending. These two statements don't fit together.

This is hyperbole. The scenario constructed in this report does not have a "high likelihood" of occurring in part because it requires a confluence of circumstances coming together. While it's certainly true that climate change will be damaging to society and the environment and many of the consequences will be severe this does not equate to a high likelihood of civilisation coming to an end.

Projections of extreme events such as these are very difficult to make and vary greatly between different climate models.

The article points to several regional records or near-records for heat extremes that occurred near the start of the year.

The increase in the number of warm temperature records has been documented for many regions of the world (e.g. the US: https://agupubs.onlinelibrary.wiley.com/doi/abs/10.1029/2009GL040736) and can be attributed to human-caused climate change: https://agupubs.onlinelibrary.wiley.com/doi/full/10.1002/2017EF000611.

Now, according to the NASA GISS record, 2018 is likely to be the 4th warmest year on record https://twitter.com/ClimateOfGavin/status/1063106429530333184

This is an excellent summary of why the 20th century average is a suitable baseline

The article very quickly and correctly attributes the cause of the warmth to human-cause greenhouse gas emissions

Without a full analysis of this particular event we can't explicitly blame global warming for it. It is likely that the weather patterns that are bringing this heat have not changed in their frequency too much, but that the temperatures are a little warmer than they would have been under analogous weather patterns in the past.

We are experiencing more heat records in general due to human-caused climate change and this has been shown for many regions of the world.

Recent research shows that even under the Paris Agreement global warming limits, hot extremes like these will become more common: https://www.nature.com/articles/s41558-018-0191-0 Other research has also demonstrated this to be true across Europe, e.g. https://www.nature.com/articles/nclimate2468

Here and in other places in the article there are unhelpful exaggerations and hyperbole that provide a slightly misleading representation of this event.

This record has not been accepted by the UK Met Office, so the record high temperature for Scotland remains 32.9°C in August 2003

A reference here would have been useful as it's unclear exactly where this number comes from. We know that human-caused climate change can increase both the moisture content of the atmosphere, as described by the Clausius-Clapeyron relation, and the energetics of the storm, as there is increased latent heat release. Both of these climate change-influenced drivers have the potential to cause increased precipitation.

This is an intriguing statement. While it's certainly true that major hurricanes could develop in the Gulf of Mexico and strike places like Houston, it is unclear still whether such a strong system could develop in the absence of climate change. This would be very difficult to answer.

This storm was quite remarkable in its intensification. It does seem likely that the warmth of the surface and sub-surface waters, primarily linked to variability but also with a small climate change contribution, enhanced the storm.

This is a reasonable statement. As tropical cyclones are such complex events - the confluence of many factors coming together - it is difficult to estimate the climate change role in such events. However, it is likely that human-caused climate change has worsened aspects of this event as described later in the article.

There is a risk that some people might confuse the term "Paris pledges" with the 1.5- and 2-degree targets that came out of the Paris Agreement.

The article doesn't hide the fact that there is fairly high uncertainty in these projections. This helps the reader understand the confidence in the projected changes.

The map is based on the medium-emissions trajectory of RCP4.5, which is roughly in line with the combined current pledges from the Paris Agreement. It would be useful if the article noted that this trajectory is not in line with either the 1.5- or 2-degree global warming targets from the Paris Agreement.

This is both cherry-picking of the data (starting 1998) and not correct anyway http://www.realclimate.org/index.php/archives/2017/01/there-was-no-pause/

Last year the annual global temperature was about 1.1C above a late-19th century baseline but only about 0.1C of this was due to the El Nino, the other 1C is due to human influences.

So while the record occurred in 2016 due to El Nino, most of the anomaly was due to climate change.

Satellite temperature datasets have fewer records than in situ-based global temperature datasets largely because they are considerably more noisy. A lower signal-to-noise ratio results in fewer records.

This is a statement that very few climate scientists would disagree with.

The El Niño itself waned quite a long time ago with neutral conditions by mid-2016. The effects of the El Niño persisted through the year though.

This statement is referring to rapid attribution analyses that were conducted by the World Weather Attribution group (https://wwa.climatecentral.org/).

We found that the Arctic heat of late 2016 was made much more likely by human-induced climate change (although it remains a very unusual event in the current climate).

In another analysis we found the February heatwave in New South Wales was at least twice as likely to occur because of climate change.

This is an important qualification. We expect in a stationary climate that we would have some periods with more record hot temperatures than record lows and vice versa. Over an extended period of time, in a stationary climate, this ratio would average to be one. It's because this pattern of more record hot than record cold temperatures has persisted over recent decades that allows us to link this trend to climate change.

This paper is the first I'm aware of to highlight this effect over the US: http://onlinelibrary.wiley.com/doi/10.1029/2009GL040736/abstract

No direct link with climate change has been drawn here and this is good. For these daily heat records attribution to human-caused climate change is a lot harder than for extremes that occur on longer or larger scales. Indeed, in parts of India, the increase in aerosols due to pollution has counteracted the warming from greenhouse gas emissions so that there is virtually no trend in extreme high temperatures.

The use of quotation marks here helps to highlight that Pruitt's views are counter to the consensus of climate scientists.

A rapid attribution analysis has found that the extreme temperatures in the Arctic of late 2016 would be extremely unlikely without human-induced climate change https://www.washingtonpost.com/news/energy-environment/wp/2016/12/28/2016s-super-warm-arctic-winter-extremely-unlikely-without-climate-change-scientists-say/?utm_term=.a86459c4802f

This is a really important point, especially as 2017 is likely to not break the 2016 record. We expect year-to-year variability, so we may not break the 2016 record for a few years, but the overall trend is upwards.

This is an accurate description of how extreme event attribution studies are conducted.

This is true and is also consistent with findings from other areas of the world. For example, Sophie Lewis led work that showed in Australia there have been 12 times as many hot records compared to cold records since 2000.

While it seems intuitive that a general warming would lead to longer heat waves, this isn't necessarily the case. In locations where weather systems cause abrupt endings to heat waves (for example, cold fronts passing over Melbourne can reduce temperatures by more than 20 degrees Celsius in summer) then the general warming of the climate by roughly one degree Celsius isn't necessarily causing heat waves to last longer. In general, the duration of a heat wave is often related to how long a weather system like a blocking high sits in the same location, and for many areas we have not seen significant changes.

This is certainly true. Over the last few years we've seen more and more attribution statements related to heatwaves (where there is higher confidence in the climate change link compared with other extremes, such as heavy rain events) and scientists have become more confident in making this connection.