| In recent years, many ecologists and policy makers have paid great attention to searching for ways to solve the problem of increasing concentrations of greenhouse gases, especially carbon dioxide (CO2). Due to the great potential of carbon sequestration, forests are viewed as one way to reduce atmospheric CO2 concentration. Forest carbon accounting refers to estimations of carbon reserve and carbon circulation for different kinds of forests within a fixed timeframe and specific region. Carbon storage and carbon sink capacity are two key points when we do researches on forest carbon accounting. Due to that China is the largest developing country in the world, the problem of greenhouse gas emission is becoming more serious. And the negative effects on ecological environment and social economy are increasingly prominent. The reductions of forest resources in our country have an important influence on the regional and global environment. Therefore, strengthening forest carbon accounting research will be of enormous significance to development of ecology, economy and society for China. Chinese fir (Cunninghamia lanceolata), which is endemic to China, not only provides more merchant woods, but also plays a very important role in the global carbon cycle. To date, there are lots of basic research data about Chinese fir, but the results of carbon accounting for Chinese fir have great differences by reason of various data resources and methods from different researches. It is well known that estimating accurately on forest carbon accounting is very difficult. One appropriate approach for accurate simulation of carbon accounting is to develop high precision models based on amount of biomass (carbon content) research data about Chinese fir. As longs as establishing general models on the basis of some representative samples, we can directly use these models to estimate carbon content of individual tree or carbon storage of forest stand for same or similar species. Moreover, we also can estimate carbon storage of Chinese fir under regional scale by combining these models and forest inventory data. This is a most useful way to research on overall amounts and dynamic changes of carbon storage and carbon sequestration for Chinese fir.In view of this, we collected a large number data about biomass or carbon storage for Chinese fir, including measurement data by us, predecessors’ results from related researches and published national forest inventories. Then we studied on the carbon accounting models, predictions and applications by using Chinese fir in Fujian province as a case considering on different scales. The main objectives of this study included:(1) establishment carbon accounting models for individual tree, forest stand and regional forests; (2) using various dynamic predicting methods to simulating carbon storage and carbon sink capacity for Chinese fir in Fujian province; (3) researching applications of carbon accounting methods for solving some practical problems based on interdisciplinarity.The main results are as follows:1. The results of constructing carbon content compatible models of individual tree indicated that models with two dimensional variables (DH, D2H and D&H) were always superior to those with a single variable (D). The D&H variable combination was found to be the most useful predictor. Of all the approaches, nonlinear adjustment in proportion (NAP) and nonlinear seemingly unrelated regression (NSUR) could ensure predictions compatibility. Simultaneously, we found that the new general model:fi(x,y)=(α·xβ·ε+γ·eη·x·y)θ (where e is the base of natural logarithms, α, β,γ,η,θ are parameters) had better accuracy than others, including the power function, the exponential function and the polynomial function, for predictions, even in model extrapolation.2. From the scale of forest stand, two different methods were used to establish carbon storage models based on stand survey of 400 plots from Fujian Jinsen Forestry Limited Company:(1) The formulas of compatible model with volume are: simultaneous equations between In M=h1+d1 In N+g1 In(D(?) H(?)) (stand volume equation) and C-a2(Db2Hc2)Nd2(stand carbon storage equation) with constraint condition b1=b2 and c1= c2 (where by C is carbon storage of forest stand, M is stand volume, N is stand density, D is stand average diameter,H is stand average height, b1, c1, di, gi, hi, a2, b2, c2, d2 are parameters). Based on these models, we could make the variable density yield table of carbon storage under different site indexes for carbon storage prediction; (2) Machine learning modeling methods, including Back Propagation Artificial Neural Network (BP-ANN) and Support Vector Machine (SVM), were introduced to simulating stand carbon storage. We set age, average diameter, average height, density and site index as input vectors, and carbon storage as output vectors to construct models. The results illustrated that the determination coefficients (R2) of two samples (modeling samples and validation samples) were both larger than 0.94, which indicated that the predictions of stand carbon storage were accurate.3. Based on Chinese fir biomass data from predecessors’results, we tried to introduce Geographically Weighted Regression (GWR) model to estimate stand carbon storage in regional scale. The results revealed that the fitting results by GWR were better than those by Ordinary Linear Regression (OLR) model for some conifer species besides Chinese fir, including Masson pine(Pinus massoniana Lamb.), Dahurian larch (Larix gmelinii) and Chinese pine(Pinus tabuliformis Carriere). These results showed that it was more suitable to use GWR model to estimate stand carbon storage in regional scale because the fitting results were closer to observation values when geographic coordinates of each plot were introduced as one model variable.4. Based on measurement data on carbon content of 54 individual trees of Chinese fir by us, we first proposed carbon ratio models for average carbon ratio of Chinese fir in Fujian province by using 18 kinds of commonly used models in forestry researches. The results revealed that there were some relationships between D (diameter at breast height) and average carbon ratio. The best fitting model was Peal-Read model, and quadratic polynomial function took second place. The R2 of these two models were higher than 0.95 under different data processing methods. These results showed that the average carbon ratio would be decreased with age (D) increasing, but in some growth phase (in general, in over-mature phase for Chinese fir), the average carbon ratio would have been increased slightly. On basis of this regular, we obtained different average carbon ratios for different age groups of Chinese fir in Fujian province:young forest age (0.5388), half-mature forest (0.5095), near-mature forest (0.4865), mature forest (0.4840) and over-mature forest (0.4867), respectively. Then we combined these average carbon ratios and Biomass Expansion Factor (BEF) to calculate carbon storage by three Volume Derived Biomass methods, including IPCC, BEF function and regression modeling. Finally, the range of average carbon storage was from 38706.15×104t to 45413.37×104t for Chinese fir in Fujian province during 2009-2013.5. Based on the Eighth National Forest Resources Inventory of China (2009 ～2013) and the Forestry Statistical Yearbooks of Fujian province (2014), CO2FIX model was used to simulate trends of carbon sink capacity of Chinese fir in Fujian province in the next 60 years (2016～2075). The results indicated that, considering influences of some practical factors, the total amount of carbon sink capacity of Chinese fir would be between from 1.6×108t to 2.1×108t by 2075, which accounts for 35%-50% of carbon storage of Chinese fir at present.6. Economic benefits and optimal rotation periods were analyzed in consideration of carbon sequestration and timber based on survey data of sample plots for a Chinese plantation in northwest of Fujian province. The stand volume and carbon storage were predicted by one-dimensional (ARIMA) time series model and multi-dimensional (CAR) time series model. The results indicated that the accuracies of ARIMA models and CAR models for stand volume and carbon storage were high. So the results predicted by these models were acceptable and effective. And the economic benefits were calculated by using average net present value (ANPV) and land expectation value (LEV). So the main conclusions included:(1) Under the present Chinese fir management environment, forest managers do not need to change the optimal rotation periods to obtain maximum total economic benefits including carbon sequestration and timber. It is beneficial to enhance the enthusiasm of Chinese fir operation; (2) Choosing high quality site to plant Chinese fir will increase economic benefits; (3) The higher carbon price, the more economic benefits in consideration of carbon sequestration and timber; (4) High interest rate is disadvantageous to manage Chinese fir.7. According to Chinese fir biomass data from predecessors’results, the functional relationships between biomass, community growth and annual litter fall were established for Chinese fir in Fujian province. The results showed that the net production of Chinese fir in Fujian province was 16.2022×106t/a and the average net primary productivity was 11.9406 t/(a·hm2) during 2009～2013. Based on carbon ratio models of Chinese fir in Fujian province, the amount of carbon sink capacity was 8.2299×106 t/a and the average carbon sink capacity was 6.07 t/(a·hm2). And the proportions of the total carbon sink amount for young age forest, half-mature forest, near-mature forest, mature forest and over-mature forest were 33.65%,23.38%, 22.10%、18.56% and 2.41%, respectively. Therefore, the amount of carbon sink by Chinese fir in Fujian province equaled 8.37% of average annual amount of carbon emissions from fuel energy consumption during 2009～2013, and the equaling rate of different age forests of Chinese fir were 2.82%(young age forest),1.95% (half-mature forest),1.85%(near-mature forest),1.55%(mature forest) and 0.20% (over-mature forest). |
PDF Full Text Request