Working Paper: CEPR ID: DP16267
Authors: Hector Calvo Pardo; Jose Olmo; Tullio Mancini
Abstract: Building on an economic model of rational Bitcoin mining, we measure the carbon footprint of Bitcoin mining power consumption using feedforward neural networks. After reviewing the literature on deep learning methods, we find associated carbon footprints of 3.8038,23.8313 and 19.83472 MtCOe for 2017, 2018 and 2019, which conform with recent estimates, lie within the economic model bounds while delivering much narrower confidence intervals, and yet raise alarming concerns, given recent evidence from climate-weather integrated models. We demonstrate how machine learning methods can contribute to non-for-profit pressing societal issues, like global warming, where data complexity and availability can be overcome.
Keywords: Machine Learning; Carbon Footprint; Cryptocurrencies; Nowcasting; Feed Forward Neural Networks; Climate Change
JEL Codes: Q47; Q54; C45; C55; F55; F64
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Electricity consumption (L94) | Carbon intensity of bitcoin mining (L71) |
| Carbon intensity of bitcoin mining (L71) | Carbon footprint of bitcoin mining (L71) |
| Electricity consumption (L94) | Carbon footprint of bitcoin mining (L71) |
| Bitcoin mining (L72) | Carbon emissions (Q54) |
| Carbon emissions (Q54) | Climate change (Q54) |
| Social cost of carbon associated with bitcoin mining (Q52) | Environmental concerns (Q56) |