[Researches of soil normalized difference water index (NDWI) of Yongding River based on multispectral remote sensing technology combined with genetic algorithm].

Basin soil type, moisture content and vegetation cover index are important factors affecting the basin water of Yongding River, using traditional sampling method to investigate soil moisture and the watershed soil type not only consuming a lot of manpower and material resources but also causing experimental error because of the instrument and other objective factors. This article selecting the Yongding River Basin-Beijing section as the study area, using total station instruments to survey field sampling and determination 34 plots, combined with 6 TM image data from 1978 to 2009 to extract soil information and the relationship between region's soil type, soil moisture and remote sensing factors. Using genetic algorithms normalization to select key factors which influenced NDWI, which is based on the green band and near-infrared bands normalized ratio index, usually used to extract water information in the image. In order to accurate screening and factors related to soil moisture, using genetic algorithms preferred characteristics, accelerate the convergence by controlling the number of iterations to filter key factor. Using multiple regression method to establish NDWI inversion model, which analysis the accuracy of model is 0.987, also use the species outside edges tree to meet accuracy test, which arrived that soil available nitrogen, phosphorus and potassium content and longitude correlation is not obvious, but a positive correlation with latitude and soil, inner precision researched 87.6% when the number of iterations to achieve optimal model calculation Maxgen. Models between NDWI and vegetation cover, topography, climate ect, through remote sensing and field survey methods could calculate the NDWI values compared with the traditional values, arrived the average relative error E is -0.021%, suits accord P reached 87.54%. The establishment of this model will be provide better practical and theoretical basis to the research and analysis of the watershed soil moisture and organic of Yongding River.

[Forest fire division by using MODIS data based on the temporal-spatial variation law].

Forest fires are harmful to the ecological environment, which have induced global attention. In the present paper fire activities extracted from MODIS and burned areas were compared, and it was found that the wave band of 8-9 extracted from MOD14A1 was useful for fire monitoring, and the data accorded with field investigation with goodness of fit reaching up to 0. 83. Through combining this wave band and the relative data to make the time and space analysis of the forest fires for 11 years, from 2000 to 2010, the study showed that the fire occurred most frequently in the spring, the autumn took the second place, and in the summer there was almost no fire occurrence unless drought. Through the analysis of the research area, the burned areas of the coniferous forest and temperate mixed forest were 53.68% and 44%, respectively, while the grassland was only 2.32%. Da Hinggan Ling region was the main combustion area, the burned areas were 64.7% and that for Xiao Hinggan Ling was about 23.49%, while those for other areas were less than 5%. The majority of forest land of burned areas has a gentle slope (< or =5 percent), and is in the middle altitude between 200 and 500 m. So, using satellite remote sensing to analyze the time series of burned areas in forests would make the relationship between the fire activities, climate change, topography and vegetation type clear and it is also helpful to predicting the risk level of the fire areas.

[Study on artificial neural network combined with multispectral remote sensing imagery for forest site evaluation].

Multispectral remote sensing data containing rich site information are not fully used by the classic site quality evaluation system, as it merely adopts artificial ground survey data. In order to establish a more effective site quality evaluation system, a neural network model which combined remote sensing spectra factors with site factors and site index relations was established and used to study the sublot site quality evaluation in the Wangyedian Forest Farm in Inner Mongolia Province, Chifeng City. Based on the improved back propagation artificial neural network (BPANN), this model combined multispectral remote sensing data with sublot survey data, and took larch as example, Through training data set sensitivity analysis weak or irrelevant factor was excluded, the size of neural network was simplified, and the efficiency of network training was improved. This optimal site index prediction model had an accuracy up to 95.36%, which was 9.83% higher than that of the neural network model based on classic sublot survey data, and this shows that using multi-spectral remote sensing and small class survey data to determine the status of larch index prediction model has the highest predictive accuracy. The results fully indicate the effectiveness and superiority of this method.

[Advances in the research on hyperspectral remote sensing in biodiversity and conservation].

With the species reduction and the habitat destruction becoming serious increasingly, the biodiversity conservation has become one of the hottest topics. Remote sensing, the science of non-contact collection information, has the function of corresponding estimates of biodiversity, building model between species diversity relationship and mapping the index of biodiversity, which has been used widely in the field of biodiversity conservation. The present paper discussed the application of hyperspectral technology to the biodiversity conservation from two aspects, remote sensors and remote sensing techniques, and after, enumerated successful applications for emphasis. All these had a certain reference value in the development of biodiversity conservation.

[The inversion processing of vegetation biomass along Yongding River based on multispectral information].

Researching on vegetation biomass using the traditional measurement method is time-consuming and hard sledding, and prediction precision of biomass is always not good because of uncertain influencing factors. The present article aims at the current situation of Hebei-Beijing reach along Yongding River, using the Thematic Mapper data in this place on 20th July 2009 as source data, with the 30 meters Digital Elevation Model data in Beijing and other auxiliary information, meanwhile through field observation data, to find out the possible functional relationship along vegetation biomass and remote sensing image factor. The authros sorted out the vegetation biomass and remote sensing image factor on the sample plot, then set up an inverse model through multiple linear regression analysis, and analyzed the precision of inverse model. After calculating the measured value and predicted value, the authors got the global relative error is -0.025%, the average relative error is -0.016%, and the general predictive precision is 84.56%. The establishment of this model is able to investigate eco-environmental factors on large range timely, quickly and accurately, also can provide the experimental base for the eco-environmental survey on river basin, and make the foundation for the problem diagnosis of ecological environment and the research on ecosystem degradation mechanism of Yongding River.