Public Comment: 222 (Kyle Holland)

[ncx-gitbot]

Commenter Organization: EP Carbon

Commenter: Kyle Holland

2021 Deferred Harvest Methodology Section: No Section Indicated

Comment: The methodology is simply crediting the prediction of an imperfect model and fails to account for any uncertainty in those productions. As written in the proposed methodology, it is easy to show the accounting of equation 1 is the product of a model prediction for the quantity of 𝑟 and the carbon stock 𝐶𝑡0 , both estimated quantities. Given equation 1 which is Δ𝐶𝑂2𝑏𝑠𝑙,𝑡 = ∑𝐶𝑡0,𝑖(1 + 𝐺𝑏𝑠𝑙,𝑡,𝑖)(1 −𝑟𝑖 ) 𝑛 𝑖=1 − ∑𝐶𝑡0,𝑖 𝑛 𝑖=1 consider the case of one landowner 𝑛 = 1 to simply the mathematics so that Δ𝐶𝑂2𝑏𝑠𝑙,𝑡 = 𝐶𝑡0(1 + 𝐺𝑏𝑠𝑙,𝑡)(1 − 𝑟)− 𝐶𝑡0 And then further consider the case of the very start of the project crediting where 𝑡 = 0 such that 𝐺𝑏𝑠𝑙,𝑡 = 0 represents no growth, then Δ𝐶𝑂2𝑏𝑠𝑙 = 𝐶𝑡0 (1 − 𝑟)− 𝐶𝑡0 We can show that Δ𝐶𝑂2𝑏𝑠𝑙 = − 𝑟𝐶𝑡0 is simply is the amount of carbon contained in above ground biomass removed in the baseline scenario at time zero: Δ𝐶𝑂2𝑏𝑠𝑙 = 𝐶𝑡0 − 𝑟𝐶𝑡0 − 𝐶𝑡0 Δ𝐶𝑂2𝑏𝑠𝑙 = −𝑟𝐶𝑡0 This shows that the methodology credits the amount of carbon contained in above ground biomass removed in the very first year as the product of the model uncertain model prediction for 𝑟 and the uncertain estimate of 𝐶𝑡0 . The variable 𝑟 appears to represent both “the fraction of total carbon contained in above ground live tree biomass removed in the baseline scenario for property” (page 14) but also the “the fraction of carbon at risk of removal during the activity period,” (page 32) which are not the same. One is theoretical but unknown quantity and the other is an estimate. The authors explicitly show this by assuming 𝑟~𝐵𝑒𝑡𝑎𝑧𝑜𝑖𝑛𝑓(𝛼, 𝛾, 𝜇,𝜙) for the fraction at risk of removal, which is not an absolute known quantity of total carbon contained in above ground live tree biomass removed; rather, it is only a prediction that could be massively imprecise. In fact, Prestemon and Wear (2000) show the uncertainty in related estimates could exceed 100% in some cases. Therefore, it is inappropriate to apply the estimate of 𝑟, which is a random variable, as the truth in calculating Δ𝐶𝑂2𝑏𝑠𝑙,𝑡 . The methodology should account for the potentially massive uncertainty in 𝑟 and conservatively discount the value applied in determining Δ𝐶𝑂2𝑏𝑠𝑙,𝑡 , especially since the entire accounting of carbon is driven by this single quantity. While Appendix A of the methodology gives fancy equations that create the appearance of rigor, ultimately it is presenting an imperfect model and asking for credit without considering the potentially massive uncertainty in the results.

Proposed Change: No Proposed Change

[ncx-gitbot]

NCX response: We agree that accounting for uncertainty is very important in any forest carbon project. Our revised methodology accounts for and requires a deduction associated with the uncertainty of carbon stocks in the project and baseline scenarios.