Ecological Prediction Study On The Biomass Of Eucalyptus Based On 3PG Model

Forests are the largest terrestrial ecosystem carbon pool, an accurate assessment of regional forest carbon storage and carbon balance is an estimate of future CO2 concentrations in the atmosphere to predict climate change and its critical impact on terrestrial ecosystems. Our country is the largest plantation area in the world, where the eucalyptus plantation area is the third largest in the world. Eucalyptus plays an important role in mitigating the supply and demand of wood and forest products, maintain economic and social development. Prediction of eucalyptus plantation ecological conduct is the basis of biomass plantations and carbon sequestration potential effect on the quantitative assessment of regional scale. How to accurately predict large spatial scales (regional scale) and time scales (10 years,20 years and 30 years) the evolution of the trend according to small spatial scales (individual tree, kind of scale) and short time scales (1 year) research combined effect relationship between different spatial and temporal scales? This is related to the development of major national policies, industrial layout, forestry construction of major and critical strategic issues.Biomass Model is an effective and relatively accurate method to estimate forest biomass, but the prediction accuracy of the model will be affected by errors. It is generally believed that the error comes mainly from three aspects:data input error, the error system error and model structure. In this paper, we chose Nanjing County, Zhangzhou City, as the study area, based on the data in 2003 and 2009 two Type II data and survey data obtained under field parsing eucalyptus plantation wood obtained. Model eucalyptus biomass analog value and the real the difference between the observed value of the spatial distribution. Use 3PG model to analyze the impact of the mechanism and the mechanism of space environmental factors, biomass distribution differences in diagnostic framework constructed herein, improved model for the structure of the regression equation so that the improved model can be more accurate analog value, and provide support for the development of regional biomass spatial and temporal scales. The results show that:for Nanjing County, Zhangzhou City, March 2003 23 compartments, the space factor, notably terrain elevation biomass distribution differences enormous impact, and soil factor on relatively small; analog value than the initial model real observations up 30.39% through improved error model predictions after dropped 0.96%. In terms of spatial and temporal scales expand, 2003 The results will be applied to improve after 2009 1517 compartments, deviations from the model simulation of 38.36% prior to 1 down 2.54%.diagnostic framework to analyze environmental factors’impact on the difference of biomass’distribution for the mechanism and add the improving function to the model for more accurate biomass prediction, which can provide basis for the scaling transformation. The result shows that for the 323 stands in 2003, topographic factors affect the difference of biomass between the simulation and observation most, the soil factors affect relative less. The original biomass simulation is 30.39% more than the observation, after the optimization the new simulation is 5.77%. Applied to the 1517 stand plots in 2009, the error between the simulation and the observation changes from 38.36% to 11.39%.This paper constructed diagnostic framework includes:Geographic detector regression analysis,3PG model, genetic algorithm and computer programming. Geography regression detector is a new statistical method, which explores the intrinsic link between objectives and environmental factors, and not a priori assumptions and limiting conditions. The software is based on the spatial pattern of variable consistency, including the risk detection, factor detection, ecology and interaction detection probe.3-PG model is based on a physiological process of photosynthesis, while a variety of factors to consider in turn weather, site conditions, management measures and physiological characteristics of trees, etc., stand growth prediction model. It has been widely used in the world’s countries and regions. In this paper, we explore the use of geographic differences in situ probe return value biomass and analog value between the spatial distribution model to quantify the contribution of different environmental factors of the differences, and to study the mechanism and the mechanism of different factors on the difference and thus provide a theoretical basis for improving the model. Genetic algorithms are inspired by the biological theory of evolution and genetics by theory and developed, is a kind of simulation of natural biological evolution process and mechanism to solve the problem of self-organization and adaptive artificial intelligence technology, it is a natural biological reference random search algorithm selection and natural genetic mechanisms. In this paper, the genetic algorithm to explore the specific contribution of space environmental factors, biomass distribution differences, environmental factors in turn will add to the improved model to obtain more accurate biomass analog value.Further research should focus on the following:(1) to be distinguished according to whether the eucalyptus plantation suffered the impact of human activities in order to develop a more appropriate management measures for different types of forests; (2) the study of the genetic algorithm, We will take into account the higher-order functions for better calibration results; (3) in order to improve the model, the model can not only consider rewriting some modules, but can also add a priori explore the underlying mechanism to get on some of the original module it did not have.