| The reflection spectrometry in solar reflection has been widely applied to retrieve the composition content on lunar surface. The observation band range of the lunar satellites are ultraviolet, visible and near-infrared.Theenvironment of lunar surface can greatly affect the minerals identification and component retrieval,including the decreasing of reflectance, spectral band contrast and less obvious spectral feature. Therefore, the mineralogical identifications obtained from reflect spectroscopy are approximately confined to 5% of the lunar surface withlow maturity of soil. Compared with reflected thereflection spectrometry, emission spectrometry not only sensitive to silicate minerals, but it also very sensitive to roughness of lunar surface and thermophysical properties of lunar materials, like thermal inertia, granularity, density and rock’s fraction. The lunar’s atmosphere has little influence on the emissivity feature of silicate minerals, thus providing us a new way to study the component and physical properties of lunar surface.The ways to study lunar infrared remote sensing are including ground-based optical telescope, fire balloon, landing space missions on lunar surface, laboratory measurements and On-orbit satellite measurements. The study on emissivity feature of laboratory lunar soil and the lunar emissivity inversion are the two most important parts. By using the detected method of thermal remote sensing and theoretical basis of blackbody radiation equation and radiation transfer equation, the lunar emissivity inversion and scale effect are studied based on the LRO Diviner thermal infrared data through analyzing the emissivity feature of laboratory lunar soil and the lunar emissivity inversion method.The materiale missivity is built on the homogeneous isothermal surfaces. However, the surface of spatial observation is uneven, and surface temperature could be retrieved from the energy balance of the ideal surface. Different kinds of spectra database in mixed pixel are developed in order to study the radiance relationship between specific pixel and black body with same temperature, which explain the scale effect of pixel emissivity when the apparent emissivity is determined.Emissivityspectra characteristics of lunar samples could be useful indexes to assess the compositions of lunar rock and mineral, which mainly includes Christiansen Feature(CF) and Reststrahlen Bands(RB). The study designed the measurementexperiment of lunar soil simulants to obtain the emissivity spectra, and the experiment conducted in natural laboratoryenvironment, the nitrogen cold background environmentand vacuum environment by using three measurement methods, respectively. The maturity, mineral composition, chemical composition, particle size and sample temperature are analyzed using thethermal infrared emissivity spectra of Apollo samples. This result determined CF value is the indicative character of mineral contents and element contents in the lunar surface.The processing method of the Diviner infrared data was proposed and applied.Error bar charts of brightness temperature of Diviner Channels 3 to 9 from January to June2012 werebuilt to analyzethe stabilityof the Diviner data. Based on the distribution of thermal infrared data in day and night time and ‘bad pixel’ and ‘bad line’ phenomenon, the study provide the diviner data processing methods, such as numerical fitting of bright temperature, elimination of singularities, piecewise fit and latitudinal direction correction. The Brightness temperature data of channel 6 of the Diviner has been simulated by the diviner data processing methods.Thereby the distribution of high spatial resolution of full coverage of the bright temperature were obtained and evaluated by the time interval brighttemperature.The results shows that the diviner data processing methods are feasible and reliable.The scale effectof the emissivity has been studied using retrieved the emissivity on the lunar surface. The basic of the emissivitynormalization method is that the real temperature value of the pixels is replaced by the results of quadratic fit of the maximum of three channels bright temperature values in 8μm channel in Diviner. The CF value calculated by the quadratic fit of emissivity of three channels. The analysis of single pixel of emissivity value proves the existence of scale effect and discover a new space effect of emissivity.Based on the two different emissivity data of Diviner, the influences that the pixel scale effect have on the inversion results of emissivity are evaluated and the two scale effects are eliminated, which would provide reliable emissivity data for component inversion on lunar surface.Correlation relationsbetween emissivity spectral feature and component content of 18 kinds of Apollo samples with different particle sizes in LSCC are analyzed and lunar surface componentemissivity retrieval models are built.Correlation coefficients between CF values and Ilmenite, Plagioclase, Pigeonite and Mg-Clinopyroxene are 0.5825, 0.5352, 0.4483 and 0.887, respectively. And correlation coefficients are 0.5539、0.6404、0.5992、0.7834、0.6836、0.482、0.75、0.7097 and 0.5306 compared CF values with Si O2、Ti O2、Al2O3、Cr2O3、Mg O、Ca O、Mn O、Fe O and SO2. With removed scale effectof CF image, the contents of four minerals and nine elements are derived and inversion results are evaluated using composition content of Apollo15 samples. Mean square deviations(MSD) between inversion values and measured values of four minerals are 0.33、15.35、5.28 and 1.58, respectively. And seven element MSDs are 2.25、0.92、2.63、2.80、1.56、0.05 and 2.34 compared inversion values with measured valuesof Si O2、Ti O2、Al2O3、Mg O、Ca O、Mn O and Fe O. The results show that the contents of lunar surface minerals and elements can be derived based on CF values, and compared with mineral inversion result, inversion accuracies of element contents are higher, and the study provides a new method for inversion of lunar components. |
PDF Full Text Request