Retrieval Method Of Air Water Vapor Content Based On Microwave Radiometer

Atmospheric science is a new discipline to explore the various physical phenomena,the physical and chemical conditions,and the processes and regularities of the atmosphere motion which occur at the inside of the earth's atmosphere.The atmospheric remote sensing has attracted a lots of attentions in the field of atmospheric probing due to the rapid development and update of the modern instruments and equipment.The emerging technology which is employed to exploit atmosphere is inherent difference in the aspects of the detecting principle and technology,the observing and analyzing methods as compared to the traditional probing methods and possesses also the particular efficiency.The ground-based microwave radiometer,a modern apparatus of meteorological observation,can receive the electromagnetic waves from outside space with few energy depletion,high sensitivity,good motility,and high detecting precision.Moreover,the observation data can be obtained successively by picking out different group of frequencies,resulting in the drawbacks of the traditional probing apparatus regarding the large time span in the daytime,the high-value consumption for each exploration,and the complicated process of obtaining the atmospheric parameters are remedied.From the aspects of guiding the artificial rainfall and the medium and small scale numerical weather forecast in a short period of time,the ground-based microwaves radiometer has a large application prospects because it can measure the change of the liquid water in the cloud and the water vapor content in the atmosphere rapidly and precisely,leading to an accurate weather forecast.The precision of the ground-based microwave radiometer to measure the data from the atmosphere depends not only on the inherent calibration of the hardware of the employed instruments depends but also on the adopted inversion algorithm to some extent.Two group sounding data below eight kilometers at 8 and 20 o'clock of Beijing in every day are recorded for each July from 2009 to 2014 in Zhengzhou,and then the each obtained data are analyzed to ascertain whether there exists clouds or not at that time.On the basis of the relative humidity threshold value method,the clouds may be there when the estimated relative humidity is higher than or equal to 85%,otherwise there are cloudless.Meanwhile,the height of the cloud is calculated.If the estimated relative humidity is higher than 95%,the density of liquid water in the clouds is estimated to be 0.5g/m3.For those sounding data that the relative humidity in the range of 85% and 95%,the density of liquid water in the cloud can be revealed on the basis of the perfect linear relationship between their liquid water densities in the cloud from zero to zero point five.There are 84 samples meeting the demands to be selected after analyzing all of the produced rounding data in detail,and then the bi-channel brightness temperature are calculated by carrying out the ground-based microwave remote sensing equations.However,the radiation brightness temperature received by the frequency channel of the ground-based microwave radiometer is not a perfect linear correlation to the LWP or the IWV,though the LWP and the IWV are two important physical descriptor to describe the weather and climate.Fortunately,the artificial neural network has significant performance to solve the nonlinear question.The bi-channel brightness temperatures are obtained from analog computation,and then they and the temperature humidity pressure at the same time were further regarded as the input sample vectors,using the calculated IWV and LWP based on the same sounding data as the output vector to test BP neural network,it is found that the total amount of atmospheric water vapor and liquid water in cloud obtained from the analog computation are in good agreement with that got from the sample values,indicating a good linear correlation between the two group data.Furthermore,the IWP and LWP are measured simultaneously using the ground-based microwave radiometer,and the precipitation intensity in the rainfall area are calculated by remote sensing.The experimental results demonstrated that the evolution processes and changing processes of the IWV and LWP in the local area can be monitored accurately by employing the ground-based microwave radiometer to proceed to measure continuously.This is of great importance to choose the suitable occasion to implement the weather modification.The LWP value obtained from the microwave radiometer can be used to predict the rainfall and its amounts.In summary,the ground-based bi-channel microwave radiometer BP neural network algorithms are of important guiding significance to improve the speed and precision of the radiometer.