They've identified several correlates due to weather (cloud cover, rain, wind, day length), but then I can't find where they correct for these in talking about effect sizes. All of these effects then "add up" to make the nearly 10-minute variation in sleep duration.
Also, sleep isn't only about duration - most sleep scientists would want to know about sleep efficiency too. This is "time asleep / time in bed". If you wake up more when it's cold, then there is a reason for shorter sleep when it's warm. Similarly, you'd ideally correct for air conditioning and factors like this on an individual level, but they don't seem to have this data.
Activity before bed, and the light that gets into your bedroom in the morning are certainly correlated with heat. It's important to investigate how these variables interact.
Also, sleep isn't only about duration - most sleep scientists would want to know about sleep efficiency too. This is "time asleep / time in bed". If you wake up more when it's cold, then there is a reason for shorter sleep when it's warm. Similarly, you'd ideally correct for air conditioning and factors like this on an individual level, but they don't seem to have this data.
Activity before bed, and the light that gets into your bedroom in the morning are certainly correlated with heat. It's important to investigate how these variables interact.