Summary
US emissions are estimated to be in the range [110, 190] Tg C yr-1, with the caveat that the upper limit might still be a reach. The midpoint of this interval is roughly 2.5 times that of Canada
Canadian emissions are estimated at 56.6 Tg C yr-1.
Estimate for US logging-related emissions for 2005
I used three primary sources to bound an estimate for logging-related emissions of CO2 for 2005:
1. Heath and Birdsey (1993), “Carbon trends of productive temperate forests of the coterminous US”
2. Turner et al. (1995), “A carbon budget for forests of the conterminous US”
3. EPA (2007), LULUCF & Annex 3
First, these estimates do not include: (1) emissions associated with harvesting, transporting, or processing harvested wood; (2) partial offsets from energy production; and (3) offsets from substitution of more energy intensive building materials (e.g., cement).
Heath and Birdsey (1993)
To be consistent with the estimates from Turner et al. (1995) as well as the estimates from the EPA report, I thought the numbers presented from Heath and Birdsey should include carbon stored in wood products and landfills. The results still indicate that much greater carbon sequestration can be achieved with the no-harvest option, but including HWP storage does dampen the effect. Secondly, as the model predicts net sequestration, I would avoid calling the annual change “CO2 emissions.”
Changes in pools (Tg yr-1) for US timberlands under alternate harvest scenarios.
2010 2030 2050 2070
Normal harvest 73 71 37 -63
HWP, normal 78 82 79 66
Total, normal 151 153 116 3
No harvest 552 297 261 169
Difference 401 144 145 166
Comparison of endpoints under alternate harvesting scenarios
1987 2070
Carbon in forest, normal harvest 31,890 34,460
Carbon in storage*, normal harvest 741 6,829
Carbon in forest, no harvest 31,890 59,130
* Numbers actually provided for year 1990, but I assume a similar quantity in 1987 for purposes of estimation.
Comparison of annualized sequestration
C change in forest, 1987-2070, normal harvest +2,570
C change in storage, normal harvest +6,088
Total C change, normal harvest +8,658
C change per year, in CO2, normal harvest +382 Mt CO2
C change**, 1987-2070, no harvest +27,240
C change per year, in CO2, no harvest +1,203 Mt CO2
**Assuming no substantial change in storage pools under the no harvest option, though we would expect some if not most of the initial storage to decay over the period of interest (1987-2070).
Assuming the difference between annual CO2 uptake rates under normal and no harvest scenarios is a fair estimate for logging-related emissions, these results suggest that logging could annually release as much as 821 Mt CO2, or 224 Tg C. This estimate appears high, which could stem from any number of reasons, including outdated assumptions about carbon flux and product usage. For instance they assume 20% loss of initial soil C after harvest across the US
Another approach is to use representative values presented in the conclusions of the article. Heath and Birdsey estimate that absent harvest the forest C pool would increase at a rate of 328 Tg C yr-1. Under the normal harvest scenario forests continue to sequester additional carbon but at a decreasing rate, with an average value of 60 Tg C yr-1. Additionally, carbon sequestered in wood products and landfills adds on average 75 Tg C yr-1. Expected annual net sequestration from switching to a no harvest scenario is therefore 193 Tg C yr-1. This is likely a high estimate given current sequestration is much higher than 60 Tg C yr-1; the EPA estimate for net sequestration by US forests in 2005 was 190.6 Tg C.
Theoretically the difference between scenarios represents annual harvest-related emissions. The uncertainty surrounding this number is likely high, as it represents average values projected into the future.
Turner et al. (1995)
Turner et al. (1995) estimated that harvest related activities can annually transfer 266 Tg C out of the live tree biomass pool. Of the 266 Tg C yr-1, 124 Tg C yr-1 is a reduction in growing stock, with the remaining 142 Tg C yr-1 non-growing stock partitioned between harvest emissions (72 Tg C yr-1) and transfer to the woody debris pool (70 Tg C yr-1). Additional releases from rapid changes in the forest floor and understory following harvest amounted to 33 Tg C yr-1. Thus their estimate for on-site harvest related emissions was 105 Tg C yr-1. Of the removed growing stock, 50-75% is stored in long-term forest products, and the rest is expected to be released to the atmosphere within 5 years. If half is released and we assume a constant decay rate, an additional 12.4 Tg C would be released per year, for a total of (105 + 12.4) = 117.4 Tg C yr-1. Scaling this up by 7% to account for the larger harvest value in the US
EPA (2007)
It is important to distinguish between: (1) emissions associated with harvests that occurred in 2005 and (2) emissions associated with harvests that occurred prior to and including 2005. The later quantity represents the balance between inputs and outputs, and requires information on previous harvest quantities and decomposition rates. Because the US EPA report only provides net changes in these pools over time, it is difficult to estimate the relative contributions of recruitment into and decomposition from various pools. To do so would require much more information than is provided in the EPA documents (contacting Smith or Heath might be the place to start). I therefore estimate the first quantity – emissions associated with harvests that actually occurred in 2005. Total emissions would of course be higher, but this presents a reasonable lower bound.
Total HWP emissions can be calculated as the total harvest less any increase in HWP storage. From Table A-200, column 5, total harvest in 2005 was 132.9 Tg C, and the net increase in storage was 28.2 Tg C (A-200, columns 2A + 2B, or A-198). Total HWP emissions are therefore 104.7 Tg C (A-200, column 7). This emissions estimate does not include releases from on-site carbon that was transferred to new pools after harvest, e.g., slash.
Tree carbon corresponds to 35-48% of total on-site carbon (Harmon, personal communication). During harvest, approximately 65% of the tree is removed, or 23-31% of total on-site carbon. Of this quantity harvested, between 50-75% ends up in long-term forest products, or 11-23% of total carbon. If 132.9 Tg C represents roughly 30% of total on-site carbon, then total on-site carbon can be estimated as 443 Tg C. Assuming half of that total carbon is in stable soil pools, 222 Tg C of labile carbon was on-site prior to harvest. After removal of 132.9 Tg C from harvest, 89.1 Tg C is left on site to decompose over time. Assumptions about transfers into various pools and decomposition rates influence estimates for the quantity that would be released within a year of harvest. Slash decays within 5-30 years, sometimes longer (Harmon, personal communication).
Below I vary the percentage of remaining on-site biomass that decays within a year and estimate slash-related emissions, as well as total harvest-related emissions.
Slash Total
5% 4.5 Tg C 109.2 Tg C
10% 8.9 Tg C 113.6 Tg C
15% 13.4 Tg C 118.1 Tg C
20% 17.8 Tg C 122.5 Tg C
25% 22.3 Tg C 127 Tg C
Estimate for Canadian logging-related emissions for 2005
My primary source here was a document the Canadian government sent to IPCC (FCCC/SBSTA/2005/MISC.9), in particular Figure 3 (Production Approach) on page 9.
Total harvest (fuelwood, firewood, IRW): 50,107 Gg C
Increase in reservoir of long-lived products consumed in Canada
Estimate HWP emissions: 37,323 Gg C.
Adding slash (19,273 Gg C), brings total harvest-related emissions to: 50,107 (harvest) + 19,273 (slash) - 12,784 (HWP storage) = 56,596 Gg C.
Breakdown
Annual HWP Contribution (annual change in carbon stock associated with HWP):
12,784 Gg C (12.8 Tg C)
Annual HWP-related emissions:
37,323 Gg C (37.3 Tg C)
Estimated total harvest-related emissions (including HWP storage):
56,596 Gg C (56.6 Tg C)
