Retrieval Of Martian Water Vapor Column Based On Planetary Fourier Spectrometer Data

Martian atmosphere and climate have always been important topics for Mars remote sensing research.Water vapor,as an important composition of Martian atmosphere and a controlling factor on Martian climate,has gained increasing attention from researchers especially with the observing of water vapor for the first time in 1977.Although the absolute amount of water vapor is small in Martian atmosphere,it has drastic seasonal dynamics caused by temperature,pressure,and dust storms.Water vapor can act as condensed form in Martian surface-atmosphere interactions,thus plays an important role in affecting Martian atmospheric circulation.The distribution of water vapor,especially its vertical distribution,can affect other atmospheric processes through photochemical reaction and cloud radiation created by water vapor.Currently,infrared band observations obtained by satellites in Martian orbits are the major sources of data for Martian water vapor research.Water vapor absorption bands in near-infrared(1.35?m),middle-infrared(2.6?m),and thermal-infrared(20-50?m)regions are used to derive water vapor content.Among them the thermal-infrared bands are least affected by dust aerosol which exists year round.However,temperature profile is needed for thermal-infrared based water vapor retrieval.One of the challenge for Martian water vapor research lies on the limited spatiotemporally continuous observations for studying the water cycle.The Planetary Fourier Spectrometer(PFS)entered Martian orbit in 2004 and is still operating until now.It covers the thermal-infrared bands sensitive to water vapor absorption.However,the low signal to noise ratio of PFS makes it hard to get high accuracy water vapor retrievals which limits its application on analyzing long-term dynamics of water vapor.The PFS observations cover the polar cap of the Mars and it changes the time of observing which makes it possible to be used to investigate daily dynamics of water vapor.Thus,it is necessary to improve the water vapor retrieval algorithm for PFS to get more data for Martian water vapor research.This study investigates the water vapor retrieval algorithms for PFS observations for the first time in China.The interferogram data was calibrated to get radiance.The impacts of various atmosphere and surface parameters on water vapor retrieval were then analyzed within the water vapor sensitive spectrum region.Based on that result,the surface temperature and emissivity were calculated.Then,the atmosphere temperature profile was retrieved with 667 cm^-1 as center channel.Finally,the water vapor column content was retrieved.1)The pre-processing and calibration methods of the PFS data were improved to increase the accuracy of calibration.First,Hanning function was used which is the most stable and efficient apodization function to reduce side-lobe interference and the noise level in data.However,the spectral resolution of the data was also reduced from 1.3cm^-1to 2 cm^-1because of the apodization.Second,considering efficiency and effectiveness,the absolute value based method was used to correct the phase error.Finally,the digital value after Fourier Transform was calibrated to get the radiance.The impact of instrument temperature on calibration was considered and incorporated through generating temperature dependent calibration parameters which increased the accuracy of calibration.2)The impacts of various atmosphere and surface parameters on PFS based water vapor retrieval were analyzed to select the water vapor retrieval channels.First,the impacts of temperature,pressure,and other atmospheric composites on water vapor line absorption were evaluated.Considering the possible differences caused by season and region,the influences of atmospheric temperature,pressure,and vertical distribution of water vapor on radiance were simulated for high,medium,and low latitude regions in northern and southern hemisphere in summer and winter seasons.Results showed that the impacts of atmospheric temperature and pressure on PFS-LWC water vapor retrieval channel is positively correlated with water vapor content.The higher the water vapor content,the larger the influences.Lastly,the impacts of surface temperature and emissivity on radiance were evaluated.Although the true surface temperature is missing,the contribution of the surface could be accurately calculated through adjusting the surface emissivity based on the water vapor absorption channel with similar surface temperature.3)Martian water vapor column content was retrieved.The prior knowledge for retrieval was obtained from European Mars Climate Database(EMCD).Surface pressure was calculated based on elevation and time of PFS observation.Planet model in LBLRTM was used as forward model.The gravity and atmospheric molecular mass was adjusted to Mars'.Since signal-to-noise rate of single PFS observation was not high enough to retrieve water vapor,several PFS observations were merged into one spectra to derive water vapor.Surface temperature and emissivity were estimated based on merged spectra.Although individual surface temperature and emissivity estimation was not accurate,surface effects can be removed exactly when jointly using them in retrieval.Using the absorption band for CO₂ at 15?m,atmosphere temperature profile under 50 km was retrieved using Bayesian optimization at a 5 km vertical resolution.Finally,water vapor column content was retrieved using Levenberg-Marquardt method.Results showed that water vapor can be calculated correctly under different atmospheric conditions.4)PFS spectral observations were divided into 5°*2.5°(Longitude of sun*Latitude of Mars)grid to generate one spectra in each cell.Then five Martian years water vapor column content were retrieved.Seasonal and latitude variations of Martian water vapor were analyzed.Results showed that Martian water vapor has a clear seasonal cycle.This study also compared PFS results with water vapor from EMCD and retrieval results of other popular sensors(SPICAM and TES)to verify its accuracy.Comparisons showed that they were mostly agreed in spatial and temporal patterns.The absolute values of water vapor column content were also similar.