I think you're right, they do note this as a limitation of their study due to insufficient data: > The great unanswered question faced when exploring cryptodam

I think you're right, they do note this as a limitation of their study due to insufficient data: > The great unanswered question faced when exploring cryptodamages is that while we can identify select geographic hotspots of production we currently do not know in the aggregate where the electricity used in cryptocurrency mining is physically produced. This is because we, like Krause and Tolaymat, do not know the physical locations of cryptocurrency miners, whether individuals, groups or aggregates in, say, a region or country. There is considerable evidence of concentration of mining operations in particular locations, typically where reliable electricity is cheaply available, though, precise data are lacking. In the US, perhaps the most well-known concentration is the Mid-Columbia Basin area in central and eastern Washington State, where cheap electricity is produced by hydropower along the Columbia River, however, mining in other US locations also occurs. There is also evidence of large mining camps in China. With time and emergent research there may be improved information about the amounts of electricity devoted to mining cryptocurrencies for particular locations or regions, but it is currently not available. It would be interesting to see the results of their analysis if recalculated using the more granular CCAF survey data.