Research On The Real-time Solar Radiation Model Of The Lunar Surface And Its Geological Applications Based On Multisource Data

Because of the unique space location,surface environment and energy potentiality,the Moon has become the priority target of human deep space exploration,and is also the strategic high ground for space development and scientific and technological innovation of major aerospace countries and space organizations.The lunar atmosphere is very thin,and its internal heat source is also weak,the radiation from the Sun is the most important energy source of the Moon.With the development of lunar exploration plans,detailed understanding and analysis of the solar radiation is not only the prerequisite for human activities on the lunar surface,but also of great significance for the establishment of thermal physical models and the study of thermal parameters with remote sensing data.Therefore,it is of great significance to quantitatively analyze the lunar surface solar radiation and apply it in the lunar scientific research.The quantitative simulation of solar irradiance and solar radiation energy is helpful to understand the illumination information of the Moon.Until now,the study of solar irradiance is only a parameter in the temperature model,neither considering the time-varying difference brought by the real-time position of the Sun and the Moon,nor analyzing its spatial distribution characteristics on the lunar surface.However,the research of solar energy is mainly expressed by the illumination rate,without considering the factors such as the surface albedo and the solarirradiance variations.Therefore,the research on the mathematical model and distribution characteristics of solar radiation energy received by the lunar surface also needs to be supplemented.Improving and enriching the analysis methods of surface solar radiation provide an important guarantee for further understanding of the energy conditions and thermal characteristics of the lunar surface.The application of solar radiation information on the lunar surface is also an important topic.The solar irradiance and other physical parameters at the time of interest can eliminate the influence of the observation time on the surface temperature in the simulation model.In addition,the diffuse reflection energy from complex topography is one of the important factors affecting the surface temperature.The current temperature models only simplified this factor.In order to obtain the surface temperature more accurately,this complex process must be discussed.Remote sensing thermal data is an important basis for the study of the geological information of the Moon.Combined with the surface temperature information,the study of its geological application will greatly enrich our cognition of the Moon.Its study methods and results are of great geological significance.Chang'E Lunar Microwave Sounder(CELMS)data is closely related to the temperature of lunar regolith,which is an important data source to study the parameters of shallow lunar regolith,and greatly influenced by the parameters of lunar regolith.Because of the deficiency of original data,the research of CELMS data is greatly limited.Previous spatial interpolation methods can not provide the non-linear relationship between the lunar regolith parameters and the brightness temperature(TB),so it is a useful supplement to obtain accurate and detailed CELMS TB distribution based on the lunar regolith parameters.Under above background,this paper constructs a real-time geometricmodel to analyze the solar radiation on the lunar surface using multisource data,which is based on the astrometry,blackbody radiation law and radiation transfer equation.According to these quantitative results,the surface temperature model of the lunar dayside is improved,and its spatial-temporal variation characteristics and geological applications are studied.In addition,combined with the parameters of lunar regolith obtained from other remote sensing data,a noon microwave TB prediction model is constructed based on machine learning method to obtain a more detailed TB distribution of the Moon.The main research results of this paper are as follows:(1)A real-time solar radiation model of the lunar surface is constructed to quantitatively analyze the spatial-temporal distribution characteristics of the effective solar irradiance and the solar radiation energy received.Based on the DE430 ephemeris,1/64° LOLA(Lunar Orbiter Laser Altimeter)topography data and Clementine UVVIS data,this paper analyzes the real-time position relationship between the Sun and the Moon,further considering the slope,lunar libration,shadowing of the topography and surface albedo.The temporal and spatial variation of the effective solar irradiance and the received solar radiation energy on the lunar surface are analyzed quantitatively.The results well show the energy difference caused by latitude effect,surface topography,surface albedo,illumination rate,and the angle between the lunar equatorial plane and the ecliptic plane and lunar physical libration.(2)An improved surface temperature model of the lunar dayside.In this paper,we further consider the effects of lunar libration and diffuse energy of complex topography,and improve the lunar surface daytime temperature model based on the physical parameters of the time of interest.Taking Sinus Iridum as the study area,the surface temperature distributions of all the lunar unit surface with local time at 6:00,9:00,12:00,15:00 and 17:30 are simulated using 1/64° LOLA data and Apollo-15in-situ data,and the results show the great variations of the lunar surface temperature during the daytime and the temperature difference due to the topography and latitude effect.The diffuse reflection energy on the surface temperature at the complex topography areas shouldn't be ignored.The temperature caused diffuse reflection energy is the smallest and highest at noon and dusk,respectively.Their highest temperatures are 10.2 K and 36.3 K,respectively.Furthermore,the noon simulation has the best accuracy,and the results near the night show the variation complexity of the lunar surface temperature.(3)The geological application of lunar surface temperature based on the spectral emissivity method.Based on the lunar surface temperature model in this paper,the processing method of spectral emissivity map is improved.We calculate the surface temperature at observation time for all of the lunar unit surfaces in the field of view of remote sensing data,and their mean values are selected as the physical temperature of lunar regolith.The CELMS data and Diviner(The Diviner Lunar Radiometer Experiment)TIR(thermal infrared)data are two feasible data strongly related to temperatures,and their geological applications are evaluated by the estimated spectral emissivity maps.The distribution features of spectral emissivity hint good agreements with the previous geological interpretation results,which shows the advantages and feasibility in distinguishing the thermal radiation differences of the lunar regolith.Microwave data can indicate the thermal characteristics of the lunar regolith in the deep layer,which is less affected by the outer space.Moreover,in order to verify the effect of observation time on remote sensing thermal data,this paper analyzes the change of daily maximum temperature in a lunar day from 2000 to 2040,which proves that the temperature data collected by the probes in orbit are heavily affectedby the observation time,that is,when using the remote sensing thermal data from long-term observation,the thermal features caused by observation time should be analyzed carefully.(4)The prediction model of noon microwave TB based on BPNN(Back Propagation Neural Network)method.The distribution of microwave TB with high spatial resolution is an important method to study the thermal characteristics of lunar regolith.In order to reveal the effect of the surface temperature,the material composition of the lunar regolith and the unbalanced spatial distribution of the CELMS data,this paper attempts to use the BPNN method to describe this complex relationship.Moreover,the noon period is selected because of the greatest effect of solar radiation on TB.We choose the Fe O abundance,Ti O2 abundance,effective solar irradiance,solar radiation energy received,lunar surface albedo and surface roughness as the input parameters,and the TB values of each channel as the output parameters,a pixel-based 6-13-1 three-layer noon microwave TB prediction model is constructed.Taking Mare Nectaris as the study area,the TB distribution maps with the spatial resolution of 0.25°× 0.25 ° are obtained to analyze the microwave thermal radiation features at the four channels.The results indicate that the lunar regolith composition is inhomogeneous at different depths,and the thermal evolution of the Mare Nectaris is complex.The accuracy analyzing results show that the coefficients of determination of this model at3.0GHz,7.8GHz,19.35 GHz and 37.0GHz are 0.950,0.953,0.934 and 0.877,respectively;the regression coefficients of all data are 0.973,0.976,0.967 and 0.934,respectively.That is,the prediction results at 7.8GHz are the best,those at 37.0GHz are not as good as other channels,which shows that it is feasible to predict the lunar surface TB data at noon based on the BPNN method,and the inversion results provide an important reference for processing the CELMS data.This paper is of great significance to the understanding of the thermal environment of the lunar surface and the research of remote-sensing thermal data.It also provides an important reference for the analysis and application of solar radiation of other celestial bodies without atmosphere.Furthermore,it is helpful to reveal more geological issues related to solar radiation on the lunar surface.