Hyperspectral remote sensing image is susceptible to atmospheric scattering and absorption effect, which causes distortion in the spectral information of the surface recorded by hyperspectral sensor, leading to the fake spectral shape for the surface. The atmospheric effect must be removed in order to get the true surface reflectance. In this study, by following the principle of solar radiation transfer in the atmosphere and the characteristics of hyperspectral remote sensing images, the atmospheric effect within 7 EO-1 Hyperion images between 2001 and 2010 was corrected and the corresponding surface reflectance cube images were obtained. The main contents and results are summarized as follows:
1. A sensitivity analysis for hyperspectral remote sensing atmospheric correction was implemented and a general purpose multi-dimensional look up table was built up according to the sensitivity analysis results. As the look up table can be widely used in atmospheric correction, so the usage and interpolation for multi-dimensional look up table is described in this thesis.
2. Based on the features of EO-1 Hyperion image, a preprocess workflow before atmospheric correction was built up. The image quality was significantly improved after preprocessing and the retrieval error due to the quality of the data was reduced. The preprocess workflow in this thesis is useful as a reference for other hyperspectral remote sensing images.
3. The improved Dark Object method was applied to retrieve aerosol optical thickness and the affected factors, such as aerosol mode, surface elevation and water vapor content, were analyzed. These factors may introduce errors to aerosol optical thickness result. The aerosol optical thickness retrieval result showed good linear relationship with AERONET observed data/MOD04 aerosol optical product with R 2 equaled to 0.79, while root mean square error was found 0.03074.
4. The water vapor absorption and non-absorption channels near 940 nm water absorption region were chosen for water vapor retrieval. The water vapor retrieval result has a good linear relationship with MOD05 water vapor product. The R 2 is 0.973 and RMSE is 0.0824 g/cm 2 . The surface elevation and central band shift in water vapor absorption band has a greater impact on water vapor retrieval result than aerosol optical thickness.
5. An integrated surface reflectance retrieval model was built up under the support of aerosol optical thickness and water vapor retrieval result. The spectral shape of the retrieved surface reflectance is very similar with the theoretical one. The retrieved surface reflectance is comparable with the output of FLAASH software. Comparing to the FLAASH output, the retrieved result has an average correlation coefficient higher than 0.9, average spectral angle less than 0.08 radians and the root mean square error less than 0.02. Both the FLAASH output and retrieved result are run by the same input parameters.
6. The solar induced chlorophyll fluorescence intensity is extracted form EO-1 Hyperion remotely sensed images with the value of 1~8 W/(m 2 ∙sr∙μm), which is also close to the values calculated by other scholars. The NDVI index can be used to indicate and distinguish the Earth’s surface vegetation coverage but it has less relationship with biochemical parameters between vegetation photosynthesis and plant canopy reflectance. Due to the impact of lutein off epoxidation, the surface reflectance values of vegetate area are decreased while the non-vegetation area reflectivity will not be affected by this cause, resulting PRI photochemical index value of the vegetation area is lower than non-vegetation area. The fluorescence intensity of the vegetation area is increased successively in the following images, 2004-10-11, 2004-08-31 and 2010-06-20. The released fluorescence intensity from the golf course grass was found higher than other vegetation area, which might be
related to the type of vegetation and canopy structure.
With the rapid development of spaceborne and airborne sensor and ground imaging spectroscopy, hyperspectral remote sensing image data acquisition has become increasingly frequent and the amount of data is also increasing. How to effectively manage and share a mass of spectral data is a key challenge to the remote sensing scientists. In this paper, hyperspectral data collected by the self-developed ground imaging spectral radiation measurement system (Field Imaging Spectrometer System, FISS) is the main data source to be studyed. In order to establish a spectral database system, the following two things are should to be carried out: First, the overall design of the hyperspectral database, FISS hyperspectral data access technology as well as the display of the spectrum. Combined these with the theory of multi-dimensional database based on three-dimensional spherical system of , we also come up with hyperspectral data query, retrieval and visualization. On the basis of these studies, a data format called *.mdh is proposed for hyperspectral management, and a hyperspectral database system integrated some simple spectral analysis methods and information extraction methods based on image data is devloped. Finally, a spectrum mixing analysis based on weight was presented on this paper.
In this paper, the main results and conclusions are as follows:
1. According to the characteristics of the FISS hyperspectral data and system application requirements, we designed the Hyperspectral data storage model, in which GeoRaster is as the core object. With this storage model, physical database design and detailed design is also been carried out. Research have showed that this model is useful to the the storage and display of hyperspectral database system data, and provides effective support for spectral analysis.
2. Combined with the FISS data storage model, the data storage methods based on the stored procedure and the data acquisition method based on dynamic data table conversion is used to implement the hyperspectral data storage and read. On the basis of these data structure, we proposed a new fomat for hyperspectral data management – *.mdh. This data structure can be used for multi-source data storage, retrieval and three-dimensional spherical system visualization, which is the manifestation of hyperspectral data’s spectral and image.
3. The hyperspectral database system, in which the Fiss data is the main data source is deveoped. This system have implemented the data import, query and visualization. Using the database system we attempt to built a FISS sample database and integrate some simple spectral analysis method and hyperspectral information extraction module in it. Finally, a weight spectral unmixing method based on a FISS spectrum sample library is proposed, and it is supportive for further data filtering.
The process of generation, migration and transformation of non-point source pollution is simulated by means of the model of non-point source pollution，which is of theoretical significance and practical value to control and management of non-point source pollution and the protection to water sources. Based on the methodology of estimation of non-point source pollution load in the large - scale basins and the spatial datasets of Yuecheng reservoir basin, this paper studies on the coupling of spatial data and the model of non-point source pollution and proposes a method of estimation of non-point source pollution load based on pixel. The study concentrates mainly on two aspects: For inversion of the key environmental parameters, the research finishes the land use and cover classification, the inversion of vegetation coverage and the simulation of soil erosion. For estimation of non-point source pollution load, the research couples spatial data to the model of non-point pollution, accomplishes estimation of non-point source pollution based on pixel, and
estimates the non-point pollution load in Yuecheng reservoir basin. Main contents and results are summarized as follows:
1. According to the underlying surface characteristics of semi-arid areas, the decision tree classification rules based on object-oriented were established in this study. The results of land use and cover classification in Yuecheng reservoir basin showed that the overall accuracy reaches 86.95% and the coefficient of KAPPA reaches 0.8518 which meets the requirements of estimating the non-point source pollution. The vegetation coverage was inverted based on binary pixel model, and the results showed that the high vegetation coverage accounting for more than 50% of the entire basin in Yuecheng reservoir basin. Soil erosion was simulated Soil erosion was calculated based on USLE, and the results showed that the erosion modulus less than 5t/hm2 • a accounting for 73%, and erosion modulus between 5t/hm2 • a and 25 t/hm2 • a accounting for 26% of the entire basin. In general, the situation of soil erosion is not serious in Yuecheng reservoir basin.
2. The research is focus on coupling spatial data and the model of non-point III pollution, accomplishes estimation of non-point source pollution based on pixel, and the Model validation results showed that the relative error of TN, TP, NH 4+ and CODis respectively 10.11%、37.37%、28.89% and 9.76%.
3. The non-point source pollution load is estimated by the method of estimation of non-point source pollution load in the large-scale basins based on pixel in the Yuecheng reservoir basin, and the results showed that the load of nitrogen,phosphorus, ammonia nitrogen and COD which is in the dissolved state respectively is 1369.05t, 134.9t, 482.28t and1241.79t, and the load of nitrogen and phosphorus which is in the adsorbed state is respectively 531.44t and 210.45t.
The determination of the coastline position is of great significance to integrated coastal zone management，and the use of remote sensing technology to extract coastline attracts more and more attention. Recently, two problems which need to be solved when using remote sensing to extract coastline are as follows: 1) The accuracy of coastline extraction cannot be guaranteed by applying single algorithm，because there are many coastal types and different types of coasts consist of different characters. 2) Due to the influence of tide and other factors, the boundary of land and water, extracted from remote sensing image directly, is not the real coastline. To obtain the real coastline, tidal correction is needed to move instantaneous waterline to MHWS (Mean High Water Springs)
The existing researches are mostly focused on the extraction of instantaneous waterline，but lacking of tidal correction and verification of accuracy. This paper presented a method combining coastline extraction with coastal type and tidal correction. MNF rotation，MNDWI，morphology and edge detectionwere applied to SPOT-4 and SPOT-5 data acquired in Qinhuangdao coastal zone to extract instantaneous waterline.
Based on the accurate extraction of the instantaneous water line, this paper also adopted the methods of using tidal data, and DEM data to calculate the slope of shoal, respectively. Moreover, the verification of accuracy of coastline extraction was achieved by the GPS data obtained in the same period. The results show that the accuracy of coastline extraction using two methods is better than the uncorrected, as a result of comparing the mean distance between GPS point and coastline extracted. The mean distance by using tidal data and DEM data is 3.48m and 6.82m respectively, which is shorter than that without tidal correction, 9.54m. Besides, the distance between the coastline extracted by using tidal data and eighty two percent of GPS points is shorter than 5m (half of a pixel), making this method more accurate.
Finally, this paper analyzed the coastline change by using multi-temporal sandy coastline data of Beidaihe which was extracted by this paper’s method. The results show that the distance of coastline moving landward is 3.30mduring the period of 2006-2009, and the distance of coastline moving seaward is 3.02m during the period of 2009-2011.