Water content of Korean pine samples was studied using near infrared spectrascopy (NIR) combinded with parial least squares regression (PLS) analysis. Seven typical wave lengths were selceted to characterize the wood water content of Korean pine. Prediction models of wood water content were developed based on the original NIR spectra at seven varied wave lengths. The best model was showed between wavelengths from 1000 to 2100 nm with correlation coefficient (R) of 0.9871, standard error of calibration (SEC) and root mean square error of calibration (RMSEC) of 0.0337 and 0.0334, respectively. The correlation coefficient（R） for model validation was 0.9817 with standard error of prediction (SEP) and root mean square error of prediction (RMSEP) of 0.0470 and 0.0465, respectively. This study indicated that NIR is a useful tool for fast prediction of wood water content of Korean pine. The modeling proficiency could be improved significantly for some certain wave length. It could be a research focus for future related studies.

Topsoil porosity (TSP) is an important parameter for the research of soil physics, agriculture and environmental protection. However, the traditional method for measuring porosity is time consuming. Conversely, a series of new methods measuring soil surface roughness (SSR) are increased and become more and more quickly. Some researchers propose to predict TSP by SSR. In this study, two fields cultivated by different tillage type were investigated under natural condition during four years (2006-2009). The results of this study show that (i) both of soil roughness and porosity are decreased over time; (ii) there are strong correlation between soil porosity and roughness effected by rainfall; (iii) after introduce the index of accumulative mean rainfall (AMR), a model of multiple linear regression for presenting the correlation among SSR, TSP and rainfall was built using sampling data of 2006-2009 with R2>0.7.

Underground mining has caused drastic disturbances to regional ecosystems and soil nutrients. Understanding the 3D spatial distribution of soil organic matter in coal arable land is crucial for agricultural production and environmental management. However, little research has been done on the three-dimensional modeling of soil organic matter. In this study, 3D kriging interpolation method and 3D stochastic simulation method were used to develop the 3D model of soil organic matter , and the root-mean-square error (RMSE) and mean error (ME) were used as evaluation indexes to compare the simulation accuracy of the two methods. Results showed that the spatial distribution of soil organic matter obtained by using 3D kriging interpolation method is relatively smooth, which reduce the difference of spatial data; while the spatial distribution of soil organic matter obtained by using 3D stochastic simulation method is relatively discrete and highlights the volatility of spatial distribution of raw data, the RMSE obtained by 3D kriging interpolation method and 3D stochastic simulation method respectively is 2.7711 g/kg and 1.8369 g/kg. The prediction accuracy of organic matter interpolation obtained by 3D stochastic simulation method is higher than that by 3D kriging interpolation method; so the 3D stochastic simulation method can reflect the spatial distribution characteristics of soil organic matter more realistically, and more suitable for 3D modeling of soil organic matter. According to the 3D modeling of soil organic matter, the content of soil organic matter has obvious spatial difference in different soil depth（0-20 cm、20-40 cm、40-60 cm） and decreases with the increase of soil depth; The result also showed that the content of soil organic matter decreased rapidly from the upper slope to the middle slope, and gradually increased from the middle slope to the bottom, so the soil organic matter content was obviously lost in the middle slope. This result may provide useful data for land reclamation and ecological reconstruction in coal mining subsidence area.

Model for predicting wood density of Larch was established using near-infrared spectroscopy (NIR) combined with support vector machine (SVM). A hundred and seventeen Larch samples were used in the study. Wood density of samples was measured according to standard test methods for physical and mechanical properties of wood. Support vector machines for regression (SVR) was used for model building. Radial basis function (RBF) was used as kernel function to establish a model for predicting wood density. For the train set, the coefficient of determination (R2) and the mean square error (MSE) were 0.8504 and 0.6460×10-3, while the R2 and MSE was 0.8520 and 0.4451×10-3, respectively, for the test set. Results showed that using SVM in near-infrared spectroscopy calibration could significantly improve the model performance in order to rapidly and accurately predict wood density.

Application of BP neural network and NIRS for larch wood density prediction was investigated in this paper. The original spectra were collected and pretreated with the first derivative and 9 point smoothing. Eleven typical wave lengths were selected as BP network inputs to establish prediction model for wood density. Model was validated using cross-validation approach. The prediction correlation coefficient (R) is 0.916 while the root mean square error of prediction (RMSEP) is 0.0221. The results showed that using BP neural network in near-infrared spectroscopy calibration could significantly improve the model performance in order to rapidly and accurately predict wood density.