Atmospheric Carbon Dioxide Retrieval Based On Satellite Hyperspectal Infrared Data

As one of the most important greenhouse gases?GHGs?,atmospheric carbon dioxide?CO₂?has been kept increasing since Industrial Revolution,and the concentration has exceeded 400 ppm since 2005.Scientific research has confirmed that the atmosphere greenhouse effect is caused by anthropogenic emission of GHGs,which is also the major driving factor for global climate change.Therefore,monitoring CO₂ in near real-time is of significant importance to study the characteristics of carbon sources and sinks.Consequently,it will help the government to make polies in order to reduce the carbon emissions.In recent years,with the successful launches of satellite infrared instruments,like AIRS?Atmospheric Infrared Sounder?,IASI?Infrared Atmospheric Sounding Interferometer?,CrIS?Cross-track Infrared Sounder?,GOSAT?Greenhouse gases Observing SATellite?and OCO-2?Orbiting Carbon Observatory-2?,it has been of great reality to monitoring CO₂ at global scale.However,as a weak absorption gas,improvement of the accuracy of CO₂ concentration retrieval is still a difficulty and hotspot.In this thesis,the relevant research has been conducted and the results as well as the conclusions can be summarized as follows:1)Evaluation of satellite infrared hyperspectral global CO₂ retrieval product.Firstly,the observation of CO₂ from AIRS is compared with the measurements from 132 WDCGG?World Data Center for Greenhouse Gases?sites worldwide.It was found that the locations and altitudes of the sites affected the quality evaluation of AIRS CO₂ data.Since AIRS is sensitive to mid-latitude CO₂,the sites located in north hemisphere,which are affected by near-surface CO₂,are not suitable to validate AIRS CO₂ product.In addition,after the comparison with the average concentration of CO₂ retrieved by GOSAT and CO₂ profiles from NUCAPS CrIS,it was concluded that the mid-tropospheric CO₂ concentration was higher than that of near-surface.So,the sensitive height should be considered before using different satellite CO₂ products.2)The sensitivity study of CO₂ using infrared channels.Based on the fast radiative transfer model,RTTOV?Radiative Transfer for TOVS?,the sensitivity study on CO₂ sensitive channels and its influencing factors was conducted by using profile disturbance and information content methods.It was found that the uncertainty of the temperature profile significantly influenced the CO₂ retrieval,while the influence of the uncertainty of the humidity profile can be neglected;the change of the mid-tropospheric temperature significantly affected the spatial distribution of CO₂ information content;Moreover,the instrument noise showed a more significant influence than the spectral resolution:the capability of CrIS?1305 channels?to retrieve CO₂ is better than that of AIRS?2378 channels?and IASI?8461 channels?.3)Fast retrieval algorithm study based on satellite hyperspectral infrared data.The Dual-Regression?DR?retrieval and linear regression retrieval algorithm were conducted for temperature profile and CO₂ time series retrieval based on satellite radiance observations and climate dataset simulated radiances,respectively.After the comparison with both NASA operational products and NCEP GDAS?National Centers for Environmental Prediction Global Data Assimilation System?reanalysis,it was found that DR algorithm can fast and efficiently perform comparable temperature profile and CO₂ time series retrievals.Furthermore,based on climate dataset simulated radiances,the linear regression retrieval algorithm was used under clear-sky condition.It was found that an optimized training dataset used by linear regression algorithm can greatly improve the accuracy of temperature profile and CO₂ retrievals.