This is the result of one full year of data gathering and analysis on how the ambient temperature affects the efficiency (a.k.a. fuel economy) of my Bolt EV. Specifics of the data used are explained in my blog post.
I picked out the trips with the similar characteristics and grouped them into two categories: expressway and intercity driving. For the expressway data, 36 samples were used, covering 9,036.7 km (5,615.1 miles). Intercity data had 72 samples worth 1,593.1 km (989.9 miles). My Bolt EV went around for an excess of 27,000 km (16,777 miles) total during the period examined, so the samples represent a significant portion of it.
As I have to start and finish the trip within the cities, some parts of the trip did take place in the slower downtown roads. But I made sure to select only the cases where this is less than 25% of the trip distance. Also, traffic congestion would warp the comparison and so I filtered out the cases where the average trip speed was less than 70 km/h for expressway samples and 40 km/h for intercity ones.
Crucially, I have completely avoided using heaters all this time - deep apologies to my family - in order to rule out its effects on the data. I did have to use air conditioning sparingly from time to time, but made sure that it used less than 1% of the total power consumption.
I have observed that the battery conditioning never kicked in, and the usable battery capacity remained stable throughout the year (around 56 kWh), with smooth, slow degradation occurring regardless of the season.
Considering all this, the variance of the car's efficiency can be largely attributed to the difference in the ambient temperature and their correlation can be plotted as a graph you saw on top.