A method and system for predicting aboveground biomass of woody plant

Abstract

The invention discloses a method and a system for predicting the aboveground biomass of woody plants. The method firstly obtains input vectors (stem length, leaf number and root number of woody plants) and output vectors (fresh weight of stem and leaf and dry weight of stem and leaf of woody plants) of the BP neural network; constructing a BP neural network model for predicting aboveground biomassaccording to the input vector and the output vector; then the BP neural network model is trained in cycles according to several training samples, and the trained BP neural network model is generated,which can be directly used to predict the aboveground biomass of woody plants (fresh weight and dry weight of stem and leaf). The trained BP neural network model selects the phenotypic characteristics of woody plants (stem length, leaf number and root number) as independent variables, thus reducing the complexity and acquisition time of sample data acquisition, and eliminating the need of consuming a large amount of manpower and material resources; because of the close relationship between phenotypic characteristics and aboveground biomass, the prediction results have high accuracy.

Description

technical field [0001] The invention relates to the technical field of aboveground biomass prediction, in particular to a method and system for aboveground biomass prediction of woody plants. Background technique [0002] Biomass is the energy basis and source of nutrients for the operation of the ecosystem, and it is also an important manifestation of the productivity of the entire ecosystem (BROWNS, LUGOAE.1984). At the same time, it is also an important indicator reflecting the strength of the community or ecosystem function (Zhao Tongqian, Ouyang Zhiyun, Zheng Hua.etal2004). In the study of vegetation ecology, biomass is not only an important parameter to characterize the quantitative characteristics of plant communities, but also an important indicator to reflect the primary productivity of plant communities, and it is also the main embodiment of the ability of the ecosystem to obtain energy. Has a very important influence (Yu Wantai, Yu Yongqiang. 2001). The biomass ...

AI Technical Summary

Problems solved by technology

These two methods are commonly used prediction methods, but they may have large errors when solving complex nonlinear problems

The average biomass method is based on the biomass per unit area of ​​various types of forest plots measured in the field, and then multiplied by the area of ​​​​the type of forest to calculate the regional forest biomass. System biomass estimation has been widely used (DixonRK, BrownS, HoughtonRA, et al.1994), but the realization of the average biomass method requires a lot of manpower and material resources

The optical remote sensing method starts from the physiological process of photosynthesis, that is, the formation of vegetation productivity. According to the absorption, reflection, transmission of solar radiation by plants and the transmission of radiation in the vegetation canopy and the atmosphere, combined with the ecological impact factors of vegetation productivity, in the Establish a scientific and complete mathematical model and its analytical formula between the information received by the satellite to invert the remote sensing information and environmental factors to estimate the biomass (SpencerRD, GreenMA, BlggsPH.1997); the cost of remote sensing method is high and human interference is large, The acquisition speed of change data is slow, and multiple receptions will cause errors to accumulate

It can be seen that the current domestic and foreign biomass research methods generally have technical problems such as large measurement errors, low efficiency, labor and material resources, and high measurement costs.

Images