| Forest biomass estimation is an essential aspect of quantifying the national or regional forest carbon budget, as well as the changes in the forest ecosystem. Eight national forest continuous inventories (NFCIs) have been accomplished since the 1970s:1973-1976,1977-1981,1984-1988, 1989-1993,1994-1998,1999-2003,2004-2008 and 2009-2013. In the past seven inventories, China’s NFCI focus on the monitoring of forest area and timber volume, but lack of forest biomass and carbon storage estimation. Well-designed and statistically sound NFCI system over long periods, combined with local sample observations, may provide the best data sources for accurately quantifying forest biomass and carbon storage and their dynamics over large areas. This study is focused on nine forest types in the Three Gorges Reservoir Region (TGRR). Based on the field measurements, NFCIs, forestland land map and remote sening data, we discussed the methods of estimating forest biomass and carbon storage in single-trees, plots and even large scales.The results are mainly concluded as follows:1) We developed aboveground biomass equations for six major tree species that compatible with tree volume equations using nonlinear error-in-variable model. Weighted regression model was used to eliminate heteroscedasticity and get final model parameter. The results indicated that biomass equations with diameter at breast height (DBH) for all six major tree species fitted the data well, and could accurately predict the biomass of the felled trees, explaining above 91% of the variation observed, except for Betula (R2 0.801). With the addition of variable tree height (H) and crown width (CW), the adjusted R2 were slightly improved, with a growth rate of 0.8%～4.4%.2) The carbon content factor (CCF) of different components of six major tree species in TGRR were measured by using the dry combustion method. Then weighted biomass regression model was applied to determine the CCFs of whole tree. The results showed that the CCFs of whole tree were 0.5490 for Pinus massoniana,0.5223 for Cunninghamia lanceolata,0.5142 for Cupressus funebris, 0.4914 for Quercus,0.4779 for Betula,0.4471 for populus, respectively. In general, the average CCFs of broad leaved tree species (0.4721) were lower than that of coniferous tree species (0.5285).3) Two methods of estimating the total forest biomass (including both above and belowground biomass) of the TGRR were employed and compared:a variable biomass expansion factor method (VBEFM) and a weighted biomass regression model (WBRM). Firstly, stand biomass was calculated by summing the biomass values of single trees within each plot estimated by the single-tree biomass equations. Then based on VBEFM and the data of 2178 permanent sample plots in the TGRR, we developed the relationships between stand biomass (Mg/ha) and stand volume (m3/ha) for nine forest types. The results showed that the relationships models performed well, with the R2 between 0.911 and 0.978. The linear models between stand biomass and stand volume could be applied to estimate regional forest biomass directly. Besides, we developed the biomass expansion factor (BEF) regression relationships for nine forests groups based on the WBRM. BEF values, reflecting relationships between biomass and tree volume, were calculate for nine forest types in the TGRR. Based on the NFCIs, these BEF values were used to estimate the status and changes of forest biomass in the TGRR during 1999-2013. Results indicated that the total forest biomass significantly increased from 76.108 and 75.923 Tg (1 Tg= 1012 g) in 1999-2003 to 99.905 and 100.418 Tg in 2004-2008, and then went up to 125.752 and 130.643 Tg in 2009-2013 using two methods respectively, with an average annual rate of increase of 3.90% (WBEFM) and 3.95% (WBRM) during 1999-2013. The total forest carbon storage were 52.460 and 52.339 Tg in 1999-2003,73.758 and 74.709 Tg in 2004-2008,84.670 and 84.985 Tg in 2009-2013 using two methods respectively, with an average annual rate of increase of 4.00%(WBEFM) and 4.12% (WBRM) during 1999-2013. The results have verified that it is an effective approach to estimate regional forest biomass using the NFCI data and the volume-derived biomass method, which can be applied to national forest inventories in China. Furthermore, the forest carbon densities were 27.254 and 26.728 Mg/ha (1999-2003).32.661 and 31.846 Mg/ha (2004-2008) and 36.992 and 36.914 Mg/ha (2009-2013) based on VBEFM and WBRM, respectively. In general,the forest carbon density showed a significant linear growth trend from 1999 to 2013.4) Based on the forestland data, the BEF values, and the carbon content factors obtained before, we estimated the forest biomass and carbon storage in TGRR in 2013. The total forest carbon storage were 72.098 and 70.680 Tg using the variable BEF method and WBR model, respectively. Moreover, the spatial distribution characteristics of forest biomass and carbon storage were analyzed. The results showed that the forest biomass and carbon storage were mainly distributed at the slope degree ranging 15～34°, and decreased on both sides of 15～34°, while in the flat(<5°) and steep slope (≥45°), the forest biomass and carbon storage were very less. The forest biomass and carbon storage were evenly distributed on every slope position.5) We estimated the forest carbon storage in the TGRR for three study periods based on NFCls and Landsat TM/ETM+images. The values were 52.276 Tg in 1999-2003,57.738 Tg in 2004-2008, and 68.001 Tg in 2009-2013, respectively, with an average annual rate of increase of 1.90% during 1999-2013. |
PDF Full Text Request